Cantarella owns two registered trade marks for ORO in Australia in respect of coffee. TM 829098 and TM 1583290.
The earlier of the two, the 098 mark, was filed on 24 March 2000. However, Cantarella’s evidence was that it first began using ORO as a trade mark for its coffee products “at least” from 20 August 1996.
Having successfully sued Modena and fended off the ownership and distinctiveness challenges to its registrations, Cantarella sued Lavazza for infringement of these trade marks by selling in Australia LAVAZZA qualità ORO coffee in packaging such as:
At trial, Yates J found this packaging involved use of ORO as a trade mark but did not infringe Cantarella’s trade marks because they were invalidly registered as Cantarella was not the owner:
The registration of a trade mark may be opposed on the ground that the applicant is not the owner of the trade mark.
Who used the trade mark first
The key fight in the appeal, as at trial, was whether Caffè Molinari’s products had used “ORO” as a trade mark in Australia before Cantarella’s first use on 20 August 1996.
In the Modena case,[1] evidence from Molinari’s senior executive only showed use after that date. In this case, however, Yates J accepted evidence from two different Molinari employees and two witnesses from CMS / Saeco that CMS / Saeco, a coffee business in Coburg Victoria, had in fact imported shipments of the Molinari product before Cantarella’s first use. So, in contrast to the Modena case, Cantarella was not the owner of the trade mark ORO in Australia[2] and its trade marks were invalidly registered.
For the most part, this conclusion turned on Yates J’s assessment of the creditworthiness of the witnesses which, despite some question marks, the Full Court considered it was not appropriate to disturb.
The Full Court rejected Cantarella’s argument that the finding about the date of Molinari’s first use in Modena was conclusive. It was a factor, but needed to be weighed in the balance with all the other relevant evidence.
Cantarella attacked the evidence from the Molinari witnesses who advanced data from Molinari’s records and set out in Excel spreadsheets. The issue here was that the invoices and shipping documents were no longer available for the sales before the priority date of Cantarella’s use. One of the Molinari witnesses also admitted that the spreadsheet might record “pro forma” invoices. Apparently, these were what we might call quotes, sent to customers and which only became sales if the customers accepted them. The Molinari witnesses were unable to say whether the relevant entries were sales or “pro forma” invoices.
The Full Court, however, pointed out that “pro forma” invoices of this kind would still count as use of the trade mark as offers to sell. (Also, there was the evidence of the CMS / Saeco witnesses that orders were placed and product in the packaging was received.)
Did Molinari use ORO as a trade mark in Australia
Next, Cantarella challenged Yates J’s ruling that ORO was used as a trade mark on the Molinari 3 kg packaging:
Molinari’s Miscela Di Caffè ORO composite usage
Cantarella argued that this was use only of a composite mark “Miscela di Caffè ORO”.
As with the finding about when Molinari first used ORO in Australia, Cantarella was unsuccessful in challenging Yates J’s acceptance of the evidence that Molinari’s 3kg packaging in 1995 and 1996 took the form of the 2007 packaging.
On that basis, the Full Court dismissed Cantarella’s appeal on this ground, explaining at [93]:
We have had regard to the primary judge’s finding, applying Modena, that the word ORO is inherently distinctive and capable of acting as a badge of origin when used in relation to coffee. We have had regard to the prominence given to the word ORO both in the context of the packaging as a whole, and relative to the words MISCELA DI CAFFÈ in the much smaller font. We have also had regard to the appearance of what Cantarella calls the composite expression as it appears on both the front and back of the 3kg packaging, where the relevant words are differently arranged. In each case the word ORO appears much more prominently than the words MISCELA DI CAFFÈ. We consider the word ORO simpliciter acts as a badge of origin. We therefore agree with the primary judge’s finding on this topic.
Use as part of a composite mark
To succeed on the s 58 ground, Lavazza had to show not just that Molinari was the first user of ORO as a trade mark in Australia, but also that it had used it continuously and not abandoned it.
During the Modena litigation, howeve, Molinari had changed its packaging to use the expression “QUALITÀ ORO” – which Cantarella had not challenged. Thus, the evidence showed a progression of uses:
Caffè Molinari’s QUALITA ORO branding vs the Miscela di Caffè ORO branding
Cantarella contended these were use of the composite mark “QUALITÀ ORO” only, not ORO alone.
Lavazza did not try to argue that “QUALITÀ ORO” was substantially identical to ORO (and so might invoke s 7(1)).[3] Rather, it argued these later forms of packaging were use of ORO itself.
At [113], the Full Court explained:
The fact that the mark in question may appear as a component of a larger mark does not preclude a finding that it has been used as a trade mark. Whether or not there has been such use will depend on the circumstances and the overall impression conveyed.
Then, the Full Court accepted Molinari had used ORO itself as a trade mark in the later forms of packaging. Their Honours noted this depended on matters of impression about which minds might reasonably differ but “on balance” agreed with Yates J’s finding explaining at [114]:
In the case of both the 6077 and 6021A/6021E products, the word ORO is given prominence by the use of a larger font and, in the case of 6021A/6021E, ORO occupies a central position on the pack separated from the word QUALITÀ by the horizontal line immediately above the word ORO. We also note that the word ORO is given prominence across the relevant product range including on packaged coffee products 1525E, 6091 (as depicted in J [201], being the December 2003 1kg packaging) and 6071. The use across the range is consistent with the use of ORO as a sub-brand within the Molinari range of packaged coffee.
No exercise of the discretion against removal
The Full Court then rejected Cantarella’s appeal from Yates J’s refusal to exercise the discretion under s 88 not to remove the registrations.
At [141], the Full Court recorded the long established proposition that, once a ground for cancellation is established, the registration should be cancelled unless sufficient reason appears for leaving it on the Register. The trade mark owner having the onus to show sufficient reason.
As this was a discretionary judgment, this required Cantarella to show legal error. The Full Court considered Cantarella failed to demonstrate any error and so this ground failed.
The main reason advanced by Cantarella was its reputation arising from its long use. However, the Full Court at [148] agreed with Yates J that this was not a matter of significant weight.
Assuming Cantarella had a strong reputation, it still retained the ability to prevent misrepresentation through passing off and the prohibition on misleading or deceptive conduct. And so, there was no compelling necessity for the marks to remain on the Register.
The unresolved issue
Cantarella sought to argue for the first time on appeal that it should be considered a joint owner of ORO at common law as a result of its long use as an honest concurrent user.
That is, Cantarella sought to throw into question the relationship between s 58 and s 44(3).
In McCormick, Kenny J had ruled that the new “plain English” version of the Act which put “honest concurrent user” in Part 4 of the Act dealing with grounds of rejection and “ownership” in Part 5 dealing with grounds of opposition meant that s 58 was in effect in a separate sphere from honest concurrent user and independent and superior to that ground. So that, even if an applicant established it was an honest concurrent user, it still could not be validly registered in the face of a successful s 58 opposition.
The Full Court refused leave for Cantarella to argue this for the first time on appeal.[4] At [134], however, their Honours did say:
The preferable solution to this difficulty may be to read s 58 and s 44(3) together to form a harmonious legislative scheme in which s 58 is read subject to s 44(3): Project Blue Sky Inc v Australian Broadcasting Authority (1998) 194 CLR 355 at [69] – [71] per McHugh, Gummow, Kirby and Hayne JJ citing (inter alia) Ross v The Queen (1979) 141 CLR 432 at 440 per Gibbs J. That would involve recognising that s 58 does not apply in the circumstances where s 44(3) has been satisfied. But it is not necessary to resolve this issue in the present case because even if s 58 operates subject to s 44(3), the requirements of s 44(3) were not shown to have been met in this case. (emphasis supplied)
The problem for Cantarella was that it had not led any evidence explaining how it had come to adopt ORO and so of honest concurrent user would not be available.
I guess we shall have to wait and see whether a special leave application is forthcoming and, if it does, its fate.
Modena was Molinari’s exclusive distributor in Australia. The High Court’s decision, of course, laid down the law in Australia to determine when a mark has no inherent capacity to distinguish and held ORO was inherently distinctive for coffee. Consequently, Modena’s defence to infringement failed. ?
The owner being the first person either to use the mark in Australia as a trade mark for the relevant goods or services or to apply to register it with the intention of using it as a trade mark here: Re Hicks’s Trade Mark (1897) 22 VLR 636 at 640; Food Channel Network Pty Ltd v Television Food Network GP (2010) 185 FCR 9 (Keane CJ, Stone and Jagot JJ) at [49] – [55], Pham Global Pty Ltd v Insight Clinical Imaging Pty Ltd (2017) 251 FCR 379 (Greenwood, Jagot and Beach JJ) at [18] – [19]. ?
If the Registrar or a prescribed court, having regard to the circumstances of a particular case, thinks fit, the Registrar or the court may decide that a person has used a trade mark if it is established that the person has used the trade mark with additions or alterations that do not substantially affect the identity of the trade mark. ?
In fact, the issue was raised in obiter dicta by Yates J at [647] – [649]. ?
In the 7Now case, the Full Court reminded us that the classic statement of when a sign is used as a trade mark for the purposes of Australian law is the Oil Drop Man case:[1]
With the aid of the definition of ‘trade mark’ in s. 6 of the [1955 Trade Marks] Act, the adverbial expression [‘as a trade mark’] may be expanded so that the question becomes whether, in the setting in which the particular pictures referred to were presented, they would have appeared to the television viewer as possessing the character of devices, or brands, which the appellant was using or proposing to use in relation to petrol for the purpose of indicating, or so as to indicate, a connexion in the course of trade between the petrol and the appellant. Did they appear to be thrown on to the screen as being marks for distinguishing Shell petrol from other petrol in the course of trade?
In that case, the High Court held that Shell was not using a caricature image of an animated oil drop man as a trade mark in its television advertisements and so did not infringe Esso’s registered trade marks depicting a “humanized oil-drop”. A surprising aspect of that decision is that, in doing so, the High Court overruled Windeyer J who had found it did infringe.
Unlike the High Court (who did watch the ads), we can’t see how Shell’s Oil Drop Man appeared. So it has always been something of a challenge to understand why.
Now Dr Vicki Huang, who professes law at Deakin University, has published an article revealing to the world some images from the original appeal books: “Uncovering secrets of Australia’s landmark High Court cases” (2024) 54 Australian Bar Review 209. (Unfortunately, its behind a paywall.) If you are a trade mark lawyer trying to grapple with what the late Prof. Lahore always regarded as the most elusive concept in Australian IP law, you should read it!
Seven had registered 7NOW as a trade mark, TM No 1540574, in respect of a range of goods and services in classes 9, 35, 38 and 41. 7-Eleven filed non-use applications against the trade mark on the s 92(4)(b) grounds.
By the time the proceeding reached the Full Court, four categories of goods and services in classes 9 and 35 were still in dispute (7-Eleven not pursuing the other services):
Category 1: computer software (in class 9);
Category 2: the promotion and sale of goods and services for others including through the distribution of online promotional material and promotional contests (in class 35);
Category 3: retail and wholesale services including retail trading via television programmes and by telephone and electronic means including the Internet (in class 35); and
Category 4: the bringing together, for the benefit of others, of a variety of goods enabling customers to conveniently view and purchase those goods including by mail order, telecommunications, website or television shopping channels (in class 35).
The trial Judge found that Seven had not discharged its onus under s 100 to show it had used its trade mark in respect of any of the categories and refused to exercise the discretion under s 101 not to order removal.
The Full Court allowed Seven’s appeal in relation to the promotion of goods and services for others only in Category 2 but otherwised dismissed it. As the Full Court emphasised in its quote from the Oil Drop Man case, the different outcomes arose from whether or not Seven’s use of its sign was sufficiently connected with the particular category.
How Seven used its 7NOW mark
It seems Seven’s evidence of use in the relevant period was limited to a one-page website:
As you can see, there is a 7NOW banner at the top of the page. Then (during the relevant period) there were 48 “tiles” which, if clicked, transferred the user to a different websites (some of which were associated with Seven such as 7plus and 7travel while others were not) and, finally, there was a panel with four columns of links at the bottom of the page.
Category 2 services
In effect the website operated as advertising the services being offered from the various websites linked through the 48 “tiles”. As the Full Court explained at [105], the
grid of those icons – [brought] to the consumer’s attention the services and goods offered on the linked websites and [made] it easier for the consumer to acquire those services and goods….
That was the performance of promotional services and those services were provided by reference to the 7NOW badge which clearly functioned as a trade mark. Accordingly, at [106]:
In our view, the connection shown on the website between the 7NOW mark, which has the character of a brand, and the actual performance of those promotional services is sufficient to mean that it has been used as a trade mark in relation to those services. ….
But, as the 7NOW website did not actually offer any goods or services for sale as opposed to the linked websites, the website was not “a means of delivering the service of selling goods. And so, Seven had not proved use of the mark in relation to sales services.
The other categories
To establish use in relation to computer software, Seven relied on the bottom two links under the 7Plus logo in the first column in the page “footer”:
“Get the iOS App
“Get the Android App”
If you clicked on one or other of those links, you were transferred to the page in the App store or the Android Play store where you could download the 7Plus app. 7NOW did not appear anywhere on those pages.
The Full Court agreed with the primary Judge’s conclusion that 7NOW had not been used to distinguish the (7PLUS) software from software being offered by third parties. A number of considerations from the overall context of the use played into this conclusion. These included:
7NOW appeared only once on the website – at the top of the page and separated from the Store links by the 48 “tiles”;
the links to the stores were very much smaller and much less prominent that the tiles;
the links appeared under the, relatively speaking, prominent 7PLUS logo
there was no reference to the 7plus app at all;
getting to the software involved a 3-step process
the user had to navigate to the 7NOW page
the user then had to scroll down to links at the bottom of the page and perceived and read the words under the 7PLUS logo; and
then click on the link and be transferred to the page;
the 7NOW trade mark did not appear on the Store page
The same problem infected the claimed use for the other categories. Seven’s argument was that its 7NOW trade mark was being used for retail services because a user could buy goods of services from the websites linked by some of the tiles. So for example if one clicked on the tile for 7travel or House & Garden, this is what happened:
As with the computer software, however, the Full Court found the consumer would think the goods or services being offered from the linked website were being offered by reference to the trade mark(s) on that website. The connection between the 7NOW trade mark and the service of offering the goods or services in question for sale was not so much attenuated as severed.
If you do a Google search to buy some product, such as say a book, and the search results bring up, say, 10 or 20 links to sites from which you can buy the book, would you think Google was engaging in retailing the book?
Discretion
Finally, the Full Court rejected the appeal from the refusal to exercise the discretion against removal.
Seven placed heavy reliance on its family of 7-formative marks and its reputation in particularly the 7plus mark given there were some 12.5 million registered users of the 7plus app.
A number of factors played into the Full Court’s conclusion at [166] that the risk of confusion if another trader were to use 7NOW in relation to, say, retail services [via convenience stores] was so high that it would be reasonable to retain the registration for all categoires. Seven Network (Operations) Limited v 7-Eleven Inc [2024] FCAFC 65 (Burley, Jackson & Downes JJ)
Last week reviewed the notice and take down / moderation procedure the majority of the Full Federal Court adopted to limit the injunction against Redbubble’s trade mark infringement. In the course of allowing Redbubble’s appeal, the Full Federal Court also unanimously flagged significant questions about when the use of a trade mark on a website overseas may constitute trade mark infringement in Australia.
On these matters, Perram and Downes JJ delivered the main reasons and Nicholas, Burley and Rofe JJ agreed.
You will recall, Redbubble operates a website to which creators can upload their designs and customers can then buy merchandise to which the customers have chosen to have the designs applied. Once a customer has placed an order, Redbubble undertakes the fulfilment function including having the products manufactured and delivered to the customer branded with Redbubble’s trade marks.
Redbubble’s servers (at least in this case) are in the United States.
This part of the appeal concerned the second part of the trial – transactions 8 to 11. In the case of these “transactions”, Hells Angels’ trade marks officer in Australia merely viewed the trade mark infringing images on Redbubble’s website; he did not even make a trap purchase. There was no evidence that anyone else in Australia ever viewed the images or bought them.
Did this infringe?
The short answer is “yes”. But it is how the Full Federal Court got there that will require careful consideration in the future.
At first instance
The starting proposition is that for trade mark infringement in Australia there must be unauthorised use of the trade mark (or a substantially identical or deceptively similar sign) as a trade mark – i.e. as a badge of origin.[1]
Way back in 2005, Merkel J had concluded in Ward v Brodie the fact that a website was accessible from Australia was not sufficient to establish use; it was necessary to show that the website was directed at or targeted Australia.[2]
At first instance in this case, Greenwood J having found that the trap purchases (transactions ##1 to 7) infringed also found that transactions ##8 to 11 infringed even without a purchase. In doing so, his Honour applied the proposition that he had propounded in his 2019 ruling (which also involved trap purchases) at [469]:
The capacity to engage, in Australia, through the website, as Mr Hansen did, constitutes use in Australia by Redbubble. [3]
The appeal
Redbubble had not disputed that proposition at trial. It did seek to raise it as Ground 1 of its appeal. As it had not sought to argue the ground at first instance, however, this would have required leave and ultimately it did not press it.
So, it was unnecessary for the Full Federal Court to deal with the issue. At [48], however, their Honours placed a question over the correctness of Greenwood J’s proposition. Perram and Downes JJ said:
For the reasons which follow, to the extent that [469] of the 2019 judgment deals with the situation disclosed by Examples 8 to 11, we would reserve the correctness of that statement for a case where it is directly raised.
In the following paragraphs, their Honours identified at least three issues which would need to be addressed.
Why is a trap ‘viewing’ not an authorised use
The first issue was why a trap viewing was not a direct infringement.
Here, Perram and Downes JJ considered the trap purchases in Ward v Brodie had not been infringing because (now repealed) s 123(1) of the Trade Marks Act had provided it was not an infringement to use a trade mark in relation to goods to which the trade mark had been applied by or with the consent of the trade mark owner.
With the repeal of that provision, however, that proposition could no longer be applied. Further, at [50] their Honours questioned whether the replacement provision, s 122A,[4] “could be pressed into service” instead. Although their Honours expressed no concluded view at this stage.
However, Perram and Downes JJ at [51] questioned Merkel J’s conclusion in Ward v Brodie that a trap purchase was not authorised use under s 8(1). As a consequence, their Honours considered it would also be arguable that at least some of the trap viewer’s actions in viewing the images (i.e. requesting Redbubble to serve the images to the trap viewer) might also be authorised use and so fall within the defence provided by s 122(1)(e).
Noting once again that these matters had not been argued and so did not need to be decided, at [52] Perram and Downes JJ considered whether authorised use could in fact be made out could be “highly dependent on the particular facts”.
The reason for this warning lay in the different nature of some of the trap “views”. Mr Hansen, the Hells Angels’ trade mark officer had navigated to the Redbubble website and specifically requested the image displayed in transaction #8. The Redbubble website, however, included a carousel feature. So that, when image #8 was displayed the website automatically presented to him other images (##9 and 11) which Redbubble recommended to him.
The geographical reach of s 120(1)
The second issue the Full Federal Court raised was the geographical reach of infringement under the Australian Act. That is, there must be use as a trade mark in Australia. The Act does not reach acts outside Australia (if they do not involve trade mark use in Australia).
Is viewing an overseas website from Australia enough
Thirdly, Perram and Downes JJ at [57] considered it is open to question whether a website overseas which is merely viewed by people from Australia (other than a trap “viewer”) without purchase would constitute use as a trade mark in Australia. Their Honours noted that the previous decisions (apart from Greenwood J’s decision under appeal) including Christian v Nestlé involved an actual trade in Australia.
Noting that Moorgate Tobacco v Philip Morris established a threshold requirement for trade mark use that there “be an actual trade or offer to trade in the goods in Australia”, their Honours explained at [62]:
The question of whether mere overseas projection without a local trade in the goods can amount to trade mark use in Australia is, in our view, a question of considerable difficulty. It is made potentially more complex in this case because although the website is hosted from servers in the United States there is no doubt that Redbubble conducts business in Australia. The difficulty is that that business does not appear to have involved, in the case of Examples 8 to 11, any more than projection into the Australian market without any consequent trade in goods bearing the marks. An important question is whether the necessary geographical nexus for use of a trade mark in Australia can be established by the mere fact that the trader is engaged in trade in Australia albeit not in relation to the infringing trade mark. The answers to these questions are not self-evident. (emphasis supplied)
At [63], Perram and Downes JJ concluded:
On the current state of the authorities, we regard the matter as undetermined. At no point in either of the trial judge’s decisions does his Honour traverse these issues. We would therefore not read [469] of the 2019 judgment as resolving them.
Perhaps the issue that concerned their Honours is that, in the bricks and mortar world, Yanx established that consumers in Australia who bought “Yanx” cigarettes in the USA and imported them into Australia for their own personal use did not use the Yanx trade mark as a trade mark. The transaction was completed in the USA and the goods, when imported for the consumers’ own personal use, was no longer in the course of trade.
On the other side of the ledger, however, Deane J said in Moorgate at 433 –434:
The cases establish that it is not necessary that there be an actual dealing in goods bearing the trade mark before there can be a local use of the mark as a trade mark. It may suffice that imported goods which have not actually reached Australia have been offered for sale in Australia under the mark (Re The Registered Trade Mark “Yanx”; Ex parte Amalgamated Tobacco Corporation Ltd., at pp 204–205) or that the mark has been used in an advertisement of the goods in the course of trade (The Shell Co. of Australia v. Esso Standard Oil (Australia) Ltd., at p 422). In such cases however, it is possible to identify an actual trade or offer to trade in the goods bearing the mark or an existing intention to offer or supply goods bearing the mark in trade. In the present case, there was not, at any relevant time, any actual trade or offer to trade in goods bearing the mark in Australia or any existing intention to offer or supply such goods in trade. There was no local use of the mark as a trade mark at all; there were merely preliminary discussions and negotiations about whether the mark would be so used.
One might think that a website which was directed at, or targeting, Australians was making an offer to trade here or had an existing intention to offer and supply here, even if there is no actual sale. At least arguably, that does not seem very different, if at all, to advertisements in magazines circulating in Australia with the aim of soliciting custom. Moreover, (and this may require evidence in a particular case), if one clicks on the “Buy Now” or “Purchase” button on most websites, the whole transaction is automated and does not involve a volitional decision by the website operator whether or not to complete the transaction.
So why did Redbubble infringe
At the risk of simplifying the arguments very significantly, Redbubble’s argument was a kind of de minimis argument that the infringing images were not available to consumers in the ordinary course of trade.
This argument had two main strands to it. One strand was that the Hells Angels had been able to identify the accused images only through a prolonged period totalling some 4.5 hours over approximately 7 hours – typing in “Hells Angels” and filtering for “Newest”. Redbubble contended this was not the behaviour exhibited by ordinary consumers who, for example, spent on average spent less than four minutes on the site. The other strand was the claimed short period of time the images were available on the website.
Perram and Downes JJ did not think the primary judge had erred in rejecting Redbubble’s argument. Their Honours further pointed out that, if Redbubble had wanted to prove that an image was unlikely to be found, it should have provided evidence of what a search would have revealed at the relevant time. The carousel function also contradicted the argument.
Putting aside the factual problems, their Honours considered there was a more general objection to Redbubble’s argument. This was not a case of a consumer using a general search engine like Google or Bing and having to filter results. Rather, it was a case involving a search of a specific website with specific functionality designed to facilitate locating desired iterms. Having noted the search function and the catalogue Redbubble provided were central components of its business model, their Honours at [81] rejected the argument:
one must distinguish between, on the one hand, the difficulties a consumer may encounter in finding what they are searching for on a website explicitly designed for the purpose of helping them do so and, on the other, the idea that such difficulties entail that the website is not engaged in the ordinary course of its trade. We do not think that the fact that it might be difficult to locate goods bearing infringing trade marks in a poorly laid out store can mean that the goods are not being offered for sale in the ordinary course of the trader’s business. We do not think any different principle applies to a website of the present kind.
Some other matters
Patches and badges of affiliation
At [226], Perram and Downes JJ appeared to suggest that the use of the Hell’s Angels trade mark as “patches” on jackets and the like to indicate exclusive membership of the club would not be use as a trade mark. Pointing out that there may be trade mark use where the sign serves dual purposes, Nicholas, Burely and Rofe JJ at [255] expressly reserved that proposition for future consideration.
Nominal damages?
It is also worth noting that the Full Federal Court rejected Greenwood J’s award of $8,250 as nominal damages (if indeed it was nominal) on the basis that such an amount could never be considered “nominal”, whether it was calculated as $750 per infringement or as $8,250 on a global basis.
After reviewing the amounts that had been awarded in other cases as nominal damages, the Full Federal Court at [127] awarded the sum of $20 per infringement making, in total, $100.
Given the error in calculating the damages, the award of additional damages ($70,000) was also set aside since the amount awarded as damages was relevant to that assessment even if the amount awarded as additional damages did not need to be proportionate.
Most of the factors listed in s 126(2) did not support an award of additional damages and, while there was a ‘mild’ need for general deterrence, the trivial quantum of infringements led to no additional damages being awarded.
See also Christian v Société Des Produits Nestlé SA (No 2) [2015] FCAFC 153; 327 ALR 630 at [78]. An approach recognising that, as a website on a server overseas is accessible by anyone in Australia with an internet connection, the trade mark owner’s rights would be set at nought if infringement could be avoided simply by setting up a website on the internet while at the same time the owner’s rights (and the Court’s powers of enforcement) are territorially limited: Lifestyle Equities CV v Amazon UK Services Ltd [2024] UKSC 8 at [3]. ?
I am not aware of any decided cases on the interpretation of this provision. In the meantime, my attempt to understand it can be found in Warwick A Rothnie, ‘Unparalled importing and trademarks in Australia,’ (2020) 21(3) Business Law International 229 (behind a paywall I’m afraid). ?
A unanimous High Court has upheld Self Care’s appeal and ruled that PROTOX and “instant Botox® alternative” do not infringe Allergan’s BOTOX registered trade mark. Nor was “instant Botox® alternative” false, misleading or deceptive conduct contrary to the ACL.
The High Court’s ruling that the reputation of the registered trade mark has no part to play in infringement under section 120(1) has finally settled that issue. More interestingly, in explaining why PROTOX did not infringe BOTOX their Honours also may also have changed how infringement is assessed. Thirdly, the High Court’s explanation why “instant BOTOX® alternative” did not infringe confirms that the plain English 1995 Act fundamentally changed the nature of trade mark use.
Some facts
Allergan owns various registered trade marks in Australia for BOTOX including in class 5 for “pharmaceutical preparations for the treatment of … wrinkles” and in class 3 for “anti?ageing creams” and “anti?wrinkle cream”.[1]
Allergan’s BOTOX product is an injectable pharmaceutical product containing botulinum toxin, type A which is administered by healthcare professionals and which can last for several months. That is, a class 5 product type. It does not sell an anti-ageing or anti-wrinkle cream. Its class 3 registration, however, is a defensive registration under section 185. As the High Court pointed out at [17], it was the reputation Allergan had derived from its extensive use of BOTOX for the goods in class 5 that was the basis for the defensive registration in class 3.[2]
Self Care markets anti-wrinkle creams under the trade mark FREEZEFRAME. Its FREEZEFRAME products come in at least 2 lines – PROTOX and INHIBOX. These creams could be self-administered and could reduce the appearance of ageing for up to a few hours. The image below shows the PROTOX packaging the subject of the litigation:
The INHIBOX labels are similar, but bearing INHIBOX AND the slogan “instant BOTOX® alternative”.
Some differences between trade mark infringement and passing off / ACL
To consider what the High Court did when finding PROTOX did not infringe BOTOX, I want to recall four or five main differences between actions for “traditional” trade mark infringement and passing off or misleading or deceptive conduct contrary to the ACL.
For “traditional” trade mark infringement (that is, infringement under section 120(1)), the trade mark owner just has to prove that the trade mark was registered – there is no need to prove reputation; just the fact of registration;
For “traditional” trade mark infringement at least, it was necessary to show that the accused conduct was conduct in relation to the goods or services for which the trade mark was registered whereas passing off and the ACL were not so limited;[3]
Trade mark infringement can occur where a reasonable member of the public is caused to wonder whether or not there is some connection between the accused conduct while passing off and the ACL require a likelihood of deception or being misled;[4]
At least for trade mark infringement, the accused use must be use as a trade mark; that is, as a “badge of origin” to identify trade source; and
“Traditional” trade mark infringement required a comparison of the mark as registered to the particular sign alleged to infringe alone. The Court has ignored the use of other marks or indicia that may distinguish the relevant goods. In contrast, the comparison for false or misleading conduct or in passing off involves the accused use in context of all the circumstances.
This last point is well illustrated by the June Perfect case.[5] There, Saville Perfumery had “June” registered in fancy script for toiletry articles including shampoo and lipsticks. June Perfect brought out its own range of lipsticks and shampoo under the name “June”. The packaging made it clear that the goods were the products of June Perfect.
The House of Lords held there was a clear case of trade mark infringement as the comparison was between the mark as registered and the sign used by June Perfect. On the question of passing off, however, the House of Lords accepted that June Perfect might be able to use its name in such a way that the trade source of the goods was clearly distinguished from Saville Perfumery. While there was an injunction to restrain June Perfect from infringing the trade mark, the passing off injunction restrained only the use of “June” without clearly distinguishing the trade source of the articles from Saville Perfumery.[6]
There has been some relaxation over time to propositions 1 and 2.
First, section 120(2) extends the trade mark owner’s rights to cover not just the goods or services specified in the registration but also to things of the same description or closely related. Unlike the case with infringement under s 120(1), however, it is a defence to this extended form of infringement if the alleged infringer can show that the way they use their sign is not likely to deceive or cause confusion. Thus, the proviso to s 120(2) states:
However, the person is not taken to have infringed the trade mark if the person establishes that using the sign as the person did is not likely to deceive or cause confusion.
Thus, Burley J quoted with approval Yates J’s dictum:[7]
So too it is recognised that, for the purposes of considering infringement under s 120(1), it is beside the point that the alleged infringer has added other material to the impugned trade mark, even if those steps were taken to avoid the likelihood of deception: Saville Perfumery Ltd v June Perfect Ltd (1941) 58 RPC 147 at 161 (Sir Greene MR) and at 174 (Viscount Maugham); Lever Brothers, Port Sunlight Limited v Sunniwite Products Ltd (1949) 66 RPC 84 at 89; Mark Foy’s Ltd v Davies Coop and Co Ltd (1956) 95 CLR 190 at 205; Polaroid Corporation v Sole N Pty Ltd [1981] 1 NSWLR 491 at 495; New South Wales Dairy Corporation v Murray Goulburn Co-Operative Company Limited (1989) 86 ALR 549 at 589; Polo Textile Industries Pty Ltd v Domestic Textile Corporation Pty Ltd (1993) 42 FCR 227 at 231–232. Considerations of this kind, if raised by an alleged infringer, are relevant when considering infringement under s 120(2) and may be relevant when considering infringement under s 120(3). However, the general position under s 120(1) is that infringement cannot be avoided by, for example, the use of additional matter if the mark itself is taken and used. Once again, if the test is not applied in this fashion a trade mark owner may be deprived of the monopoly conferred by registration. (emphasis supplied by Burley J)
As the High Court recognised in the Self Care case, the 1995 Act introduced a further broadening of what could be infringement in s 120(3). If a trade mark owner could show that its trade mark was well-known in Australia, it could claim infringement by use of a sign on wholly unrelated goods or services where the use would be likely to indicate a connection to the trade mark owner and the trade mark owner’s interests were likely to be prejudicially affected.[8]
With that background, we can turn to the High Court’s reasons.
Self Care and some principles
The appeal is concerned only with infringement under s 120(1). The extended versions of infringement for similar or closely related products (s 120(2)) and “famous” or “well-known” trade marks (s 120(3)) were not in issue in this case.
The High Court at [22] pointed out that infringement under s 120(1) requires 2 distinct questions to be addressed:
Did the alleged infringer use the sign “as a trade mark” – that is, as a “badge of origin” to indicate trade source?
If so, was the sign deceptively similar to the registered trade mark?[9]
These are, as the High Court emphasised, two different issues and the High Court approached them separately.
Use as a trade mark
The High Court confirmed that whether a sign is being used as a trade mark is to be determined objectively, without reference to the subjective intentions of the user. At [24], their Honours explained:
Whether a sign has been “use[d] as a trade mark” is assessed objectively without reference to the subjective trading intentions of the user[50]. As the meaning of a sign, such as a word, varies with the context in which the sign is used, the objective purpose and nature of use are assessed by reference to context. That context includes the relevant trade[51], the way in which the words have been displayed, and how the words would present themselves to persons who read them and form a view about what they connote[52]. A well known example where the use was not “as a trade mark” was in Irving’s Yeast-Vite Ltd v Horsenail[53], where the phrase “Yeast tablets a substitute for ‘Yeast-Vite’” was held to be merely descriptive and not a use of “Yeast-Vite” as a trade mark. Therefore, it did not contravene the YEAST-VITE mark. [citations omitted]
At [25], their Honours affirmed the longstanding principle that the existence of a descriptive element or purpose was not determinative if there were several purposes for the use of the sign. So long as one purpose is to distinguish the trade source, that will be sufficient.
Further, their Honours acknowledged that the presence of ‘a clear dominant “brand”’ can be relevant to assessing the balance of the label or packaging, but that did not mean that another sign on the labelling was not also functioning as a trade mark.
For the last proposition, the High Court cited Allsop J’s decision in the Budweiser case at [191]. In that case, Anheuser-Busch, the owner of trade mark registrations for BUDWEISER successfully sued the Czech company for infringement by the latter’s use of BUDWEISER on labels such as:
Front label (black & white)Front label (black & white)
At [191], Allsop J explained:
It is not to the point, with respect, to say that because another part of the label (the white section with ‘Bud?jovický Budvar’) is the obvious and important ‘brand’, that another part of the label cannot act to distinguish the goods. The ‘branding function’, if that expression is merely used as a synonym for the contents of ss 7 and 17 of the TM Act, can be carried out in different places on packaging, with different degrees of strength and subtlety. Of course, the existence on a label of a clear dominant ‘brand’ is of relevance to the assessment of what would be taken to be the effect of the balance of the label.
Turning to the PROTOX label, there cannot really be any dispute that PROTOX is used as a trade mark. The question then is whether it is deceptively similar to BOTOX.
The test for deceptive similarity
The High Court discussed the principles for determining whether a trade mark is deceptively similar to another at [26] – [51].
Noting that section 10 defines a deceptively similar mark to be one that so nearly resembles the registered trade mark that it is likely to deceive or cause confusion, at [26] the High Court stated the resemblance of the two marks must be the cause of the likely deception or confusion. And this involved an assessment of the two marks as a whole taking into account both their look and sound.
At [27], their Honours endorsed the much quoted explanation of the principles given by Dixon and McTiernan JJ in Australian Woollen Mills at 58 CLR 658:
“But, in the end, it becomes a question of fact for the court to decide whether in fact there is such a reasonable probability of deception or confusion that the use of the new mark and title should be restrained.
In deciding this question, the marks ought not, of course, to be compared side by side. An attempt should be made to estimate the effect or impression produced on the mind of potential customers by the mark or device for which the protection of an injunction is sought. The impression or recollection which is carried away and retained is necessarily the basis of any mistaken belief that the challenged mark or device is the same. The effect of spoken description must be considered. If a mark is in fact or from its nature likely to be the source of some name or verbal description by which buyers will express their desire to have the goods, then similarities both of sound and of meaning may play an important part.
At [28], their Honours emphasised the artificial nature of the inquiry. Stating at [29]:
…. The notional buyer is assumed to have seen the registered mark used in relation to the full range of goods to which the registration extends. The correct approach is to compare the impression (allowing for imperfect recollection) that the notional buyer would have of the registered mark (as notionally used on all of the goods covered by the registration), with the impression that the notional buyer would have of the alleged infringer’s mark (as actually used). …. (original emphasis) (citations omitted)[10]
Returning to this issue, at [33] their Honours emphasised that “the court is not looking to the totality of the conduct of the defendant in the same way as in a passing off suit”.[11] The High Court continued:
…. In addition to the degree of similarity between the marks, the assessment takes account of the effect of that similarity considered in relation to the alleged infringer’s actual use of the mark, as well as the circumstances of the goods, the character of the likely customers, and the market covered by the monopoly attached to the registered trade mark. (citations omitted)
Cases approved by the High Court in Self Care have acknowledged that questions of some subtlety can arise assessing the context of a use to determine if the sign is being used as a trade mark and assessing whether the infringing sign is deceptively similar.[12]
All of the cases endorsed by the High Court in these propositions, however, make the same point: the comparison is between the registered trade mark and the mark being used by the alleged infringer without regard to the totality of the conduct by the infringer such as the presence of other trade marks or disclaimers.
One example of the role of impression in this mark to mark comparison, expressly cited by the High Court at [29], is the Chifley Tower case.[13] There, MID Sydney’s registration of CHIFLEY TOWER for building management services was not infringed by Touraust’s proposed use of CHIFLEY for the names of the hotels it managed – such as “Chifley on the Wharf” or “The Chifley”.
One reason was that the services were not the same or of the same description.
Importantly for present purposes, the Full Court also found the marks were not deceptively similar because the public was familiar with many different uses of “Chifley” – apart from MID Sydney’s. This included the name of the Prime Minister, a restaurant and numerous geographical places. With that general background knowledge, therefore, the distinctive power of MID Sydney’s trade mark lay in the combined term, not in the common element CHIFLEY alone.
While this should not be surprising to trade mark lawyers, therefore, where it becomes interesting lies in what the High Court did when finding PROTOX did not infringe BOTOX.
Before turning to that issue, however, the High Court squarely addressed the role of reputation in infringement proceedings under section 120(1).
The role of reputation
Noting that the role of reputation has been contentious for a number of years, the High Court ruled at [50] that reputation is not relevant to infringement under section 120(1).
A number of considerations led the High Court to this conclusion. The first point at [37] was that it is registration which confers the rights in the trade mark on the owner and defines the scope of the registration. If considerations other than the registration could be taken into account “the level of protection afforded to that right would vary and be inherently uncertain.”
Another point was that the legislation specified various matters to be entered on the Register and available for public inspection. Reputation was not one of those matters and at [40] taking into account the reputation which had accrued to a trade mark would be contrary to the objective of the registered trade mark system:
which is to provide “a bright line that delineates the property rights” of a registered owner, for the benefit of the owner and the public, and runs the risk of collapsing the long standing distinction between infringement and passing off. (citations omitted)
Further, the Trade Marks Act expressly identified a role for reputation in four places:
section 60 providing a ground of opposition on the basis of the reputation in the opponent’s trade mark;
the provision for registration as a ‘defensive’ trade mark provided by section 185;
the extended form of infringement provided by section 120(3); and
the provision by section 24 for “genericide” when a trade mark has become known as the generic description of the goods or services.
Why PROTOX did not infringe BOTOX
At [63], the High Court summarised the trial judge’s finding that PROTOX was not deceptively similar to BOTOX. His Honour accepted that the two marks looked and sounded very similar but less so in idea or meaning. Further, the trial judge had held that the reputation of BOTOX was so strong that it was not likely to be recalled imperfectly. Even if there was imperfect recollection, no-one was likely to be deceived. His Honour was reinforced in this conclusion by the close proximity of PROTOX to FREEZEFRAME and the lack of evidence of actual confusion.
At [64], the High Court noted the Full Court held the trial judge had erred by failing to consider whether the use of PROTOX might cause people to wonder if there was some connection to the owner of the BOTOX mark. In finding deceptive similarity, however, the Full Court had made two errors.
First, it had relied on Allergan’s reputation in BOTOX for pharmaceutical preparations to conclude that the public might wonder whether PROTOX was some form of brand extension. Secondly, in doing so, their Honour’s had relied on the way Allergan actually used BOTOX rather than taking into account its notional use for anti-wrinkle creams in class 3.
Considering the effect of the use of PROTOX on potential customers of anti-wrinkle creams in class 3, the High Court accepted at [69] that “pro” and “bo” looked and sounded similar and the common element “otox” was both distinctive and identical. But consumers would not have confused PROTOX or BOTOX:
…. The words are sufficiently different that the notional buyer, allowing for an imperfect recollection of BOTOX, would not confuse the marks or the products they denote. The visual and aural similarities were just one part of the inquiry. (emphasis supplied)
Despite the surprise many trade mark practitioners have felt about the trial judge’s similar conclusion, up to this point the High Court’s reasoning can be seen as consistent with the extensive array of case law endorsed by the High Court which distinguishes trade mark infringement from passing off. After all, as the High Court emphasised from Australian Woollen Mills, the ultimate conclusion onabout deceptive similarity is a question of fact.
However, the last sentence from [69] quoted above picks up what their Honours had said in [68]. In considering the visual and aural impact of PROTOX, it was permissible to have regard to both the packaging and the website from which PROTOX was promoted:
it was necessary to consider the marks visually and aurally and in the context of the relevant surrounding circumstances. Considering both the packaging and the website for Protox accords with assessing the “actual use” of the PROTOX mark as required by the test for deceptive similarity. ….
The High Court then explained at [70] that the packaging and the website together dispelled the risk of implied confusion:
…. The notional buyer sees the PROTOX mark used on a similar product – a serum which is advertised on its packaging and website to “prolong the look of Botox®”. While the reputation of BOTOX cannot be considered, the relevant context includes the circumstances of the actual use of PROTOX by Self Care. “[P]rolong the look of Botox®” may suggest that Protox is a complementary product. However, as was observed by the primary judge, “it will be the common experience of consumers that one trader’s product can be used to enhance another trader’s product without there being any suggestion of affiliation”[136]. In this case, the back of the packaging stated in small font that “Botox is a registered trademark of Allergan Inc” and, although the assumption is that Botox is an anti?wrinkle cream, the website stated that “PROTOX has no association with any anti-wrinkle injection brand”. (emphasis supplied)
It is very difficult, with respect, to see how these conclusions sit with the High Court’s earlier endorsement of the authorities that additional matter such as the presence of disclaimers does not avoid infringement.
Perhaps, given the copious citation of case law endorsing the “traditional” position that it is a mark to mark comparison only, the role of the packaging and the website will ultimately be characterised as reinforcing the finding of deceptive similarity rather than determining it. Indeed, at [71], their Honours concluded there was no real, tangible danger of deception or confusion:
…. As explained, the marks are sufficiently distinctive such that there is no real danger that the notional buyer would confuse the marks or products. The similarities between the marks, considered in the circumstances, are not such that the notional buyer nevertheless is likely to wonder whether the products come from the same trade source. That conclusion is reinforced by the fact that the PROTOX mark was “almost always used in proximity to the FREEZEFRAME mark” and that there was “no evidence of actual confusion”.
instant Botox® alternative
As noted at the outset, the High Court also found that Self Care’s use of “instant Botox® alternative” did not infringe Allegan’s trade mark. Nor was it misleading or deceptive conduct in contravention of the ACL. Given the length of this post, however, consideration of those issues will have to await another day.
Edit: on 3 April to clarify that it is the ultimate conclusion about deceptive similarity that is the question of fact. Thanks, Craig Smith SC.
That is, Allergan has used BOTOX so extensively, its use by someone else in relation to class 3 goods such as anti-ageing creams will falsely indicate a connection with Allergan. Where the reputation in the trade mark is so extensive to achieve a defensive registration, it does not matter whether the trade mark owner actually uses the trade mark for the goods or services covered by the defensive registration. ?
At [19], the “overwhelming” and “ubiquitous reputation of BOTOX”. ?
For an extreme case where the services were so far removed from the goods associated with the reputation – Tabasco sauce – that deception or misrepresentation were so unlikely that there was no contravention of the ACL or passing off, see McIlhenny Co v Blue Yonder Holdings Pty Ltd formerly trading as Tabasco Design [1997] FCA 962; 39 IPR 187. ?
Compare Trade Marks Act 1995 (Cth) s 10 to the High Court’s approval in Campomar Sociedad, Limitada v Nike International Limited [2000] HCA 12 at [106] quoting Taco Co of Australia Inc v Taco Bell Pty Ltd (1982) 42 ALR 177 at 201 (Deane and Fitzgerald JJ). ?
Saville Perfumery Ld. v. June Perfect Ld. (1941) 58 RPC 147. ?
As Lord Tomlin explained at 176, “It seems to me, and the form of the second injunction supports the view, that these Appellants may be able by proper precautions to sell the three articles in connection with their name of June Perfect Ld., while clearly distinguishing those goods from the Respondents’ goods. If that can be done there is no probability that the ultimate purchaser will be deceived.”. See also e.g.Puma Se v Caterpillar Inc [2022] FCAFC 153; 168 IPR 404 (Nicholas, Rofe and McElwaine JJ) at [43] (Nicholas, Rofe and McElwaine JJ); In-N-Out Burgers, Inc v Hashtag Burgers Pty Ltd [2020] FCA 193; 377 ALR 116; 150 IPR 73 at [80] and [160] (Katzmann J) (affirmed on appeal) and many others. ?
If you know of a court case where s 120(3) has been successfully asserted, please let me know. ?
Curiously, s 120 does not in terms require the trade mark owner to prove that alleged infringer did not have the owner’s consent to use the trade mark. An alleged infringer who claims to be licensed or set up consent must do so by way of [section 123][s123] in the case of services or, in the case of goods, the wonders of [section 122A][s122a]. (I tried to untangle the latter provision in Warwick A Rothnie, ‘Unparalleled importing and trademarks in Australia,’ (2020) 21(3) Business Law International 229.) ?
See e.g.Optical 88 at [95] and Budweiser at [226]. Generally, one might have thought the emphasis in actual use in an infringement context lay in contrast to the situation at the examination and opposition stages where it is necessary to consider all fair and reasonable notional use that may be made by the applicant within the scope of the applied for registration. ?
The Full Court has dismissed Urban Alley’s appeal from O’Bryan J’s rulings that URBAN ALE was invalidly registered as a trade mark and La Sirène’s use of URBAN PALE did not infringe URBAN ALE. The decision may provide some helpful clarification of the test of substantial identity and, perhaps, urges caution against his Honour’s conclusion that URBAN PAEL was not used as a trade mark.
Urban Alley had registered URBAN ALE for beer. La Sirène started selling a Farmhouse Style Urban Pale [beer] by La Sirène under this trade mark:
Not capable of distinguishing
The Full Court upheld O’Bryan J’s conclusion that URBAN ALE lacked any capacity to distinguish as essentially descriptive or laudatory. As Urban Alley had not used the term before it registered its trade mark, it was not registrable under s 41.
This was really just the result flowing from the facts arising on the evidence. So a successful appeal was always a steep hurdle.
Practice tip: if you are going to adopt something as a trade mark, it will be unhelpful to refer to it in marketing materials in terms like:
The signature Urban Ale sits somewhere between a classic Australian golden ale and a Belgian blonde, with pleasant tropical notes but a crisp, clean finish. This is a premium beer for the people and is described as a ‘celebration of our great city, a tribute to the laneway culture and a blend of the old and the new’
and
Name: Once Bitter Style: Urban Ale (Somewhere between an Aussie Golden Ale and Belgian Blonde) ABV: 4-5%
One could be mistaken for thinking the trade mark was “Once Bitter”!
Deceptively similar to prior conflicting registration
Likewise, Urban Alley was unable to overturn O’Bryan J’s conclusion that URBAN ALE was deceptively similar to a prior registration for URBAN BREWING COMPANY and so invalid under s 44.
On this part of the case, Urban Alley argued O’Bryan J’s reasons for concluding the two marks were not substantially identical were inconsistent with his Honour’s conclusion that they were deceptively similar and so the latter conclusion was wrong.
In relation to substantial identity, O’Bryan J held that the inclusion of BREWING COMPANY in the prior mark conveyed a different meaning to ALE.
The Full Court rejected Urban Alley’s attack at [98] – [99]:
A side-by-side comparison of two marks is a studied comparison. It highlights the differences between the marks just as much as it shows their sameness, in order to reach a conclusion as to whether the two marks are, in fact, substantially identical. The primary judge’s observation must be understood as having been made in that light.
The test of deceptive similarity is fundamentally different. It is not a studied comparison. Rather, it is a comparison between one mark and the impression of another mark carried away and hypothetically recalled, paying due regard to the fact that recollection is not always perfect.
Thus, when considered from the perspective of deceptive similarity and imperfect recollection the differences which were apparent from a side by side comparison lost much of their significance. In that assessment, Urban Alley’s challenge overlooked the significance of URBAN being the first word of both marks and the close association in meaning of “brewing company” and “ale”. At [106], the Full Court explained:
The appellant also submitted that there is “no relevant trade mark resemblance” between the words “ale” and “brewing company”. This submission requires careful consideration. As the appellant’s submission recorded immediately above recognises, each compared mark must be considered as a whole. It is impermissible to dissect each mark to emphasise its disparate elements and then compare the disparate elements of each in order to reach a conclusion on deceptive resemblance. To start with, this would leave out entirely the impact of the common element “urban”. It would also ignore the synergy between the word “urban” and the other word(s) in each mark. This synergy contributes to the impression gained of each mark, which is carried forward into the relevant comparison between the two. This last-mentioned consideration brings into play the primary judge’s finding that there was a clear association in meaning between “brewing company” and “ale”. Given that clear association, coupled with use of the common element “urban”, it is understandable that the primary judge reached the conclusion he did on the question of deceptive similarity.
It might be thought that the strong emphasis on the narrow scope of the substantial identity test, requiring a studied side by side comparison, is a very welcome brake on the ruling in Pham Global.
No infringement
Having ruled that Urban Alley’s trade mark was invalidly registered on two alternative grounds, their Honours pointed out that Urban Alley’s appeal against the finding that La Sirène did not infringe must fail. So, it was strictly unnecessary to consider whether La Sirène’s use of “Urban Pale” would have been an infringement.
Speaking obiter dicta, the Full Court emphasised that O’Bryan J’s finding that La Sirène did not use URBAN PALE as a trade mark, despite its prominence, turned very heavily on the “overwhelmingly descriptive” nature of the expression. At [119], their Honours said:
Thus, it is entirely possible—indeed likely—that, absent the finding of the Word Mark’s lack of inherent adaptation to distinguish because of the ordinary signification of the word “urban”, the primary judge would have come to a different conclusion on trade mark use in relation to the respondent’s use of the name “Urban Pale” on the depicted label. This is particularly so when regard is had to the prominence and location of the name “Urban Pale”. Such use would normally be regarded as persuasively suggesting trade mark use, a consideration which his Honour seems to have recognised in the next paragraph of his reasons, where he said:
205 It is apparent that the labelling of the La Sirène Urban Pale product features the words “Urban Pale” in large lettering and an emboldened font. It is the most prominent name on the labelling. However … I do not consider that that prominence converts the essentially descriptive name into a trade mark indicating the source of origin of the product.
Even so, it will be necessary to treat the finding that Urban Pale was not used as a trade mark very carefully and confined to its particular facts. On this part of the case, the Full Court concluded at [120]:
Be that as it may be, our resolution of Grounds 1, 2 and 3 of the appeal adversely to the appellant necessarily means that Ground 5 of the appeal should be dismissed, as we have said.
Redbubble’s online market place has survived the Hells Angels’ copyright infringement claims, but did infringe their registered trade marks. The reasoning, however, leaves questions hanging over Redbubble’s business model.
Redbubble provides an online market place. Artists can upload their artwork and potential buyers can browse the site to purchase the artwork or merchandise such as t-shirts and coffee cups emblazoned with the artwork. If a purchase is made for, say, a t-shirt with a particular artwork printed on it, Redbubble’s system arranges for the order to be placed with a fulfiller and ultimately shipped in packaging which bears a Redbubble trade mark.
The claims in this case related to uploaded images of a Hell’s Angels membership card featuring a helmeted death’s head in profile:
and registered trade marks featuring versions of the death’s head: Trade Marks Nos 526530,723291, 723463, 1257992 and 1257993.
At 552 paragraphs long, this post is going to be a high level overview only.
Copyright
A key feature in the case is that Hells Angels Motorcycle Corporation (Australia) Pty Ltd is not the owner of the copyright or the registered trade marks. It contended it was the exclusive licensee in Australia of those rights; the exclusive licences having been granted by Hells Angels Motorcycle Corporation, a US corporation.
The Hells Angels lost the claim of copyright infringement. They did so, however, because they could not prove Hells Angels USA was the owner of the copyright. As a result, Hells Angels Motorcycle Corporation (Australia) Pty Ltd could not be the exclusive licensee.
Reaching this conclusion required Greenwood J to explore, amongst other things, the notion of publication and whether the supply of membership cards was supply of copies of the work to the public. And the non-applicability of the US “work for hire” doctrine in ownership disputes under Australian law.
Redbubble is still in trouble.
First, if the applicants had been able to prove title to the copyright, Redbubble would have infringed.
Contrary to Hells Angels’ arguments, Redbubble was not liable for infringement by uploading the images. That was done by the artists in question. In the examples in question, the acts involved uploading images to websites outside Australia. For example, Example 1 was uploaded by an individual in Virginia in the USA. So the uploaders themselves were not liable as their actions did not involve any act done in Australia. At [428] – [429], Greenwood J ruled that, even though the images were made available online to the public in Australia, the artists (uploaders) did not infringe because they did no act in Australia.
…. the act of the artist in uploading the image to the website and thus making the work available online to the public must be an act “done” (that is, an exercise of the exclusive right), “in Australia” and therefore, none of the artists in the examples in suit can be regarded as a “primary infringer” in the territorial sense contemplated by s 36(1) because the relevant act was not done “in Australia”.
His Honour found, however, that Redbubble would be liable for communicating the images to the public in Australia as it was the person who was responsible for determining the content of the communication for the purposes of s 22(6) when a potential customer in Australia viewed the image on the website. Redbubble’s business model was crucial here. At [435], his Honour explained:
The business model as described by Mr Hosking and its working operation as described by Mr Kovalev makes it plain that Redbubble is not in the nature of an ISP linking a user to remote websites. It is not an intermediary providing a transmission service between particular participants. It owns, operates, manages and controls the website and conducts a transactional enterprise in which it facilitates the uploading of images, the interrogation of those images in Australia, relevantly, by users, with a view to enabling sales to consumers of articles bearing the relevant images. It has a detailed business model in which it derives revenue from each transaction and controls every step of the transactional engagement between an artist and a buyer. It confirms the sale. It facilitates payment. It organises a fulfiller to apply the work to the relevant goods. It facilitates delivery of the goods to the buyer. It generates email responses which not only confirm the order but track every step of the transaction. It affixes its own trade marks to the goods. It says that it does not directly do that but there is no doubt that an essential part of its business model is ensuring that fulfillers affix the Redbubble trade marks to the goods. The labels bearing the trade marks are on the goods as delivered to each buyer. Although I will address the trade mark case shortly, the reference to Redbubble’s trade marks, in this context, is simply to note another feature of the extent of Redbubble’s engagement in and association with each transaction. It is Redbubble’s business. But for the Redbubble website, the transactions would not occur. The artworks would not be available online to consumers in Australia to consider and appraise with a view to purchasing a product bearing the artwork. The entire focus of the business model is to enable works to be made available online so that consumers can pick and choose amongst the works so as to have them applied to goods. It would be difficult to imagine a more directly engaged participant than one deploying the business model adopted by Redbubble. Although Redbubble describes itself as the “agent” of the artist (presumably as principal), the relationship is not, in truth, a relationship of agent and principal. Redbubble acts as an “independent contractor” to “facilitate the transaction” as the Redbubble User Agreement and Appendix A to the Services Agreement makes plain: [245] and [246] of these reasons. The artist, in truth, is not the “seller” in the classic sense in which that term might be understood because Redbubble is the supplier as the facilitator of all of the essential elements of the transaction with the consumer in an analogous way to that discussed in: International Harvester Company of Australia Pty Ltd v Carrigan’s Hazeldene Pastoral Company [1958] HCA 16; (1958) 100 CLR 644 at 653; Heidelberg Graphics Equipment Ltd v Andrew Knox & Associates Pty Ltd (1994) ATPR 41326 at 42, 31011, notwithstanding that the nature of the technology is different to the forms of distribution arrangement in those cases.
His Honour would, if necessary, have also found Redbubble liable for authorising the conduct if it had been infringing.
Trade Marks
Secondly, Greenwood J found Redbubble liable for infringement of the Hells Angels’ registered trade marks on works such as t-shirt designs featuring the death’s head logo.
The crux of this finding came back to Redbubble’s business model. Greenwood J accepted that the artist who uploaded the image was using the trade mark as a trade mark. Unlike the copyright test, there was no requirement that the artist be in Australia. However, so was Redbubble.
At [460] – [461], his Honour explained:
As to Redbubble, that company is “in the business” of facilitating the supply of products bearing the uploaded image of Ms Troen (in this example) or, put another way, Redbubble is in the business of facilitating the supply of clothing bearing, put simply, the registered trade marks of HAMC US (in this example). Redbubble is not the “seller” of artwork. However, it is the supplier, in the sense that it is responsible for all of the transactional supplyside elements of a transaction for the supply of goods bearing the applied works. (emphasis supplied)
Redbubble has created a business model designed to enable users, in Australia (and, for that matter users in all jurisdictions in which the website is accessible), to find images through the website comprised of, in this example, Ms Troen’s image made up of the identified trade marks of HAMC US. Redbubble enables images containing the relevant trade marks to be presented to buyers of particular goods (nominated by the artists from the website categories of those goods to which the work can be applied) expressly for the purpose of facilitating the supply of goods (clothing, in this example) to which the marks are applied. It does so by and through the functions and protocols of the website engaged by Mr Hansen (and other potential viewers of the image), in Australia.
His Honour elaborated on why Redbubble’s conduct attracted liability at [462] – [469]. While this and two other examples infringed, his Honour found that, on the particular facts, Example 2 was not infringing use.
Greenwood J’s reasons also include an extended consideration of whether Hells Angels Australia was an authorised user; ultimately concluding it was.
Greenwood J, however, rejected Hells Angels’ claims that use of “Hells Angels” as search terms, or key words, within the Redbubble site was infringing. At [542] explaining:
542. …. However, I am not satisfied that this use, in itself, is use of the word marks as a trade mark, at this point in the functionality of the website. I take that view because I am not satisfied that using the term as a search term to find a relevant image is use of the term as a “badge of origin” of Redbubble. It is, undoubtedly, a use which is designed, quite deliberately, to lead a consumer by the “search nose” to images, marks, devices, livery and badging somehow or other connected with the Hells Angels Motorcycle Club.
….
544. … use of the word marks … as a search term is a search step along the way to use of the image and thus the registered trade marks, as trade marks but use of the word marks at the point of searching is not, in itself, in my view, use as a trade mark. (original emphasis)
It appears that, at the stage of entering the search term, it is not being used to identify things offered under the aegis of the Hells Angels, but just to locate things about the Hells Angels in some way.
This is the second ruling at first instance where Redbubble has been found to infringe.
While Redbubble’s business model does leave it exposed along the lines indicated above. It is worth noting that Greenwood J awarded only nominal damages of $5,000 in respect of two of the three infringements. His Honour did not allow even nominal damages in respect of the third infringement, Example 4, as it was online for a short period, viewed only 11 times and no sales resulted.
Greenwood J expressly rejected any claim for exemplary damages. His Honour does not go into reasons. Perhaps, Redbubble’s business model did protect it. The evidence was clear, for example, that Redbubble had a policy relating to infringement claims and implemented it promptly.
Frucor’s attempt to register the colour green as the predominant colour applied to its “V” energy drinks has failed. The colour being applied for was not properly defined and, consequently, had not been used as a trade mark. Yet again, consumer survey results did not help.
Frucor’s application included a green swatch, copied from its New Zealand trade mark application and, in accordance with reg. 4.3(7) a written description:
The mark consists of the colour green (Pantone 376c), as shown in the representation attached to the application, applied as the predominant colour to the goods, their packaging or labels.
The fundamental problem was that the green colour of the swatch was not Pantone 376c.
What’s the trade mark
Frucor argued that the written description should be given priority over the coloured swatch as a matter of construction of the trade mark application.
Its argument was that any trader looking at the Register would realise the colour swatch was inherently unreliable due to the potential for corruption through scanning, printing etc. Consequently, anyone looking at the Register would recognise that Pantone 376c was the subject colour.
Yates J considered this analysis was informed by Frucor’s subjective intention. The matter needed to be determined objectively. Reg. 4.3 required the application to set out a representation of the mark and, in that context, r. 4.3(7) required a description of the mark “as represented”. It was the representation which was primary. At [122], his Honour concluded:
…. A person inspecting the Register is entitled to act on the assumption that the trade mark applicant’s own depiction of colour in the representation accompanying the application is accurate.
Once the person inspecting the Register appreciated there was an inconsistency between the graphic representation and the description, there was no way to resolve the disconformity.
“V” green had not acquired distinctiveness
As the application was for a single colour mark, this meant that Frucor could not possibly demonstrate its trade mark had acquired distinctiveness under s 41(6)(a), the old (or pre-Raising the Bar version).[1] At [145], his Honour explained:
As the Full Court explained in Woolworths/BP at [79], it is important to appreciate that it is the use of the mark applied for, as a trade mark, that determines what can be registered. In that connection, the Full Court emphasised that the mark that is the subject of the application for registration must conform to the mark that was, for the purposes of s 41(6), used before the filing date. Because, in the present case, the mark is defined ambiguously—its features are uncertain and cannot be determined objectively—it is not possible for Frucor to establish the factual condition of s 41(6)(a) by reference to its own particular use of “V” Green. It follows that, for that reason alone, registration should be refused in the present case.
Yates J would have rejected the application even if one assumed there was no ambiguity and the colour claimed was Pantone 376c.
Yates J accepted that Frucor’s use of “V” green was substantial, consistent and conspicuous. At [163], his Honour had “no doubt that … those familiar with Frucor’s “V” energy drinks would have associated “V” Green as the colour of Frucor’s core energy drink product.” (emphasis supplied)
Yates J held, however, that this extensive use and recognition was not use as a trade mark.[2] First, at [164] the “consistent presence” and “dominating display” of the “V” logo was what consumers would recognise as performing the function of the badge of origin for the goods.
Accepting that there could be more than one trade mark for a product, the second consideration was the role colour played in the soft drink market including energy drinks, soft drinks, sports drinks, fruit juices and bottled water.
Packaging and labels in this field are often brightly coloured. Soft drink producers used colour to denote a range of things: sometimes product flavour, more generally some “varietal characteristic”. Frucor itself had different varieties of “V” which were presented in different liveries (scroll down): red for berry flavoured, silver for sugarfree, black and also a yellow in addition to “V” green which was reserved for the “hero” of the range. So at [166]:
…. Although Frucor’s use of “V” Green was pervasive and no doubt fundamental to its whole marketing strategy, it was, nonetheless, reminiscent of its core product. In this way, Frucor’s use of “V” Green was essentially descriptive, not distinctive in the trade mark sense. It denoted the core product in the “V” energy drink range. I am not persuaded that, somehow, consumers would understand that colour in relation to the core product was being used differently to colour in relation to other varieties within the “V” energy drink range, or any differently from how colour was and is used descriptively across the range of non-alcoholic beverages sold through trade channels such as supermarkets and convenience stores.
So at [167], even though more than one trade mark could be used on any given product, the “V” green colour did not function as a trade mark.
What the market survey didn’t prove
Frucor’s evidence from a market research consultant based on two surveys did not help. His evidence showed, for example, that some 77% of the survey respondents identified Pantone 376c with a brand of energy drink and and 85% of those identified the drink as Frucor’s “V” energy drink. As a result, the expert concluded that Pantone 376c was part of Frucor’s “brand identity”, having properties that went well beyond decorative or functional attributes so that it distinguished Frucor’s products from other traders’.
The main problem was that evidence the public associated, or even identified, a particular colour with a “brand” by itself is not enough to establish that the colour is functioning as a trade mark. At [171], his Honour explained:
… evidence of an association (or, I would add, identification) of a sign, including a colour, with a particular product does not mean, without more, that the sign is functioning or has functioned as a trade mark in relation to that product. One needs to have an understanding of how the sign was used, in the proper context and setting, before that conclusion can be drawn. Moreover, the conclusion is not purely a factual one.
The last sentence in that extract points out that whether something is used as a trade mark is a legal conclusion.
If you are going to advance someone to give evidence that a colour has been used as a trade mark – as Frucor’s market survey expert purported to do, you will also need to demonstrate that the witness properly understands the legal concept. More practically, if you are trying to convert your colour into a trade mark, you need to educate your public that it is being used as your trade mark. At it’s most rudimentary level (and bearing in mind the range of other issues), can you tell your customers to “look for” you product by reference to its colour?[3]
Market survey blues
Approaching the matter in this way enabled Yates J to avoid having to deal with the numerous criticisms about the survey methodology raised by Coca-Cola. A couple of points, however, do emerge.
At [168], Yates J was sceptical that the sample used in the surveys was representative. Frucor’s expert had not bailed up survey respondents randomly in the street or at the shopping centre. Rather, as is quite common in marketing, they were drawn from a panel of people who had signed up to participate in surveys.
That criticism is all the more compelling given the second point. Yates J was not at all comfortable accepting Coca-Cola’s criticisms of Frucor’s survey methodology in the absence of support from Coca-Cola’s own expert evidence.
Presumably, the same reasoning would apply under s41(3) in the “new” version. ?
At [148], Yates J recalled that this required “an understanding, from an objective viewpoint, of the purpose and nature of the use, considered in the context of the relevant trade” citing Woolworths/BP at [77]. ?
See for example the evidence in Philmac about its khaki and then terracotta plastic pipe fittings leading to the conclusion at [40]. For more recent US practice see here. ?
Like MICHIGAN for farm equipment and OXFORD for books, Burley J has ordered that Bohemia Crystal’s trade marks, BOHEMIA and BOHEMIA CRYSTAL be revoked because they are not distinctive of “glassware”.
Bohemia Crystal (BCP) had been formed in 1975 to distribute in Australia Skloexport’s products. Skloexport was the State-owned entity responsible for the export of all crystal and glassware products made in Czechoslovakia. In 1999, Skloexport went into liquidation and BCP took an assignment of its Australian trade marks. The main trade mark BCP used in this period, which had been registered since 1962 was the stylised BOHEMIA Glass mark, TM No. 319701:
Versions of this mark were used with or without the words “Made in Czech Republic” or substituting the word “Glass” for “Crystal”.
On 5 October 2001, BCP applied for and successfully registered BOHEMIA CRYSTAL for glassware and on 2 May 2003 BCP applied for and successfully registered BOHEMIA for glassware.
Host is an importer and supplier of catering goods and equipment in Australia. The business was started in 1999. One range within its 2,500 product lines is its range of glassware sourced from another Czech supplier, Forincorp, marketed under BANQUET BY BOHEMIA [1] or:
Host started importing this line in 2015. BCP made the fateful decision to start proceedings for infringement of its registered trade marks and contravention of the Australian Consumer Law for false and misleading conduct.
Burley J held that BCP’s trade marks lacked any capacity to distinguish and had not been used in such a way as to have acquired secondary meaning for the purposes of s 41(6).[2] Burley J also dismissed BCP’s allegations of misleading or deceptive conduct.
It is not going to be possible in a blog post to do justice to Burley J’s 376 paragraphs. Instead five points particularly caught my eye.
First, Burley J (who was a very experienced intellectual property barrister before his appointment) pointed out that the High Court in Cantarella referenced both ordinary consumers and traders as the criterion for whether or not a sign was inherently adapted to distinguish.
BCP argued that the test focused on what ordinary consumers would think the sign meant. It had found an expert who opined that ordinary members of the public would think “bohemia” was a reference to persons with an artistic or unconventional lifestyle.
After analysing Cantarella from [86] on, his Honour concluded at [93]:[3]
in a case such as the present, it is necessary to consider the ordinary signification of the words “Bohemia” and “Bohemia Crystal” in the context of the “target audience”, being traders and consumers of the relevant goods, to determine whether at the relevant dates other traders might legitimately desire to use these words or something similar in connection with their goods, for the ordinary signification which they possess. …. (emphasis supplied)
Here, despite the evidence of BCP’s principle witness, the evidence was largely one way. It was beyond dispute that for many centuries the geographic region known as “Bohemia” which is now in the Czech Republic had a strong reputation for producing high quality crystal and glassware. There was evidence that between at least four and ten different manufacturers used the term “Bohemia” at an important annual glassware trade show to signify the geographic origins of their products. There was also evidence from a number of dealers that the term signified to the public glassware originating from the Bohemia region. BCP’s own registered user agreements for the use of Skloexport’s trade marks had also required it to promote its products as from the geographic region, Bohemia.
Burley J concluded that other traders who had glassware manufactured in the region formerly known as Bohemia legitimately and honestly wanted to use that word to describe the geographic origins of their products. The fact that Bohemia was no longer a separate country (and had not been since the World War I) and not even the contemporary name of the region was not significant.[4]
Second, Burley J found that the evidence of use of the BOHEMIA and BOHEMIA CRYSTAL did not establish that those terms had been used by BCP as trade marks or in such a way as to have acquired secondary meaning. There were three aspects to this conclusion.
Following BP v Woolworths, promotion and use is not enough. It had to be shown that the signs as registered had been used in some way to identify the signs as being trade marks.
Next, for the most part the relevant evidence showed that what BCP had been using as a trade mark was Skloexport’s composite mark, not the terms as registered. This was not use of either trade mark as registered. Moreover it was the combination of the elements in the signs as a whole which comprised the distinctiveness. These signs should not be dissected into their component parts:
228 I have some difficulty with the proposition that the words “Bohemia” or “Bohemia Crystal” should in this context be regarded as having separate trade mark signification beyond the combination in which they appear in the composite marks described above. In my view, it is the combination of elements that is distinctive. The trade mark should be viewed as a whole and not dissected into parts. Although this is likely to be a matter of fact for each case, it is notable that several cases have cautioned against the proposition that separate elements should be so distilled; see Diamond T Motor Car Company [1921] 2 Ch 583 at 588, Fry Consulting Pty Ltd v Sports Warehouse Inc (No 2) [2012] FCA 81; (2012) 201 FCR 565 at [61], [63].
229 To my eye, the whole of the 701 mark is to be regarded as creating a complicated image that taken collectively represents a sign, or badge of origin. I do not think that the elements within it may be dissected or that they would be dissected by an ordinary consumer of goods within the relevant classes. In any event, I consider that the words “Bohemia Crystal” and “Made in Czech Republic” within the 701 mark tend to reinforce the descriptive, geographical signification of those words. ….
The third factor is the way that evidence was advanced did not help BCP’s case. A lot of the evidence was vague, or general, rather than specific to what needed to be proved here: use of the signs as trade marks before the filing date. In this respect, his Honour’s discussion will repay careful study as it is not uncommon to see evidence prepared for the Office suffering from similar problems.
Third, BCP did not demonstrate any sufficient reason why its trade marks should not be removed from the Register. Burley J accepted that, Host having established the marks were invalidly registered, BCP bore the onus of satisfying the Court that there was sufficient reason not to order cancellation.
Here, the evidence did not establish that BCP had acquired distinctiveness in its signs. Importantly, allowing BCP to keep its registrations would give it an unfair advantage. At [248], his Honour explained:
…. The presence of the existing ground of revocation via the operation of subsection 88(2)(a) and s 41 indicates an intention on the part of the legislature to ensure that historical registrations should not remain on the Register where they should not have been granted in the first place. In the present case, to permit such a course would advantage the unmeritorious registrant who has incorrectly had the benefit of the monopoly since the relevant dates. BCP is able to apply to register the Bohemia marks now, should it choose to do so.
Of course, if it were to do so, it would run the risk of other traders wishing to use the terms opposing (if the Registrar got suckered into accepting the applications in the first place).
Fourth, if his Honour had not found BCPs trade marks invalidly registered, Host would have infringed. Its attempt to rely on s 122(1)(b) would have failed. This part of the case essentially turned on Host’s use being BANQUET by Bohemia (emphasis supplied) rather than BANQUET from Bohemia.
Burley J accepted that s 122 could be invoked to protect trade mark use, not just descriptive use. However, Host’s form of use showed that Host was trying to assert origin in some particular trade source rather than some geographical origin. At [301]:
…. Ms Flint and Mr Sullivan adopted this language, notwithstanding the obvious difficulty with the perception of “by” and with no knowledge of either BCP or the Bohemia marks. However, I find that they did not do so for the purpose of using “Bohemia” to designate the geographical origin of the goods, but to designate the trade origin of the goods lying in a particular entity (which was ultimately Forincorp). Accordingly, the use does not fall within the defence ….
Fifth, BCP’s allegations of misleading or deceptive conduct also failed. A number of factors contributed to this including the particular trade marks BCP had actually used and the good old-fashioned Hornsby Building Information Centre proposition.[5] In contrast to the trade mark case, in addition, it was highly significant that Host’s market and BCP’s market were quite different. BCP’s market was member of the general public looking for premium quality products. Host’s customers, however, were cafes, restaurants, pubs, clubs, community groups and the like who were cost conscious but attended to their purchases with considerable care. So, for example, at [370]:
the typical reasonable consumer is most likely to perceive the October 2015 catalogue use to represent that the manufacturer or producer of the glassware is an entity known as “Banquet by Bohemia” or “Bohemia”, there is no more than a remote prospect that reasonable customers are likely to consider that the goods offered in the catalogue are offered with the sponsorship or approval of BCP or are offered by Host with the approval of BCP or that the Banquet products emanate from BCP. First, I do not consider that the typical Host customer who encounters this publication would be likely to be aware of BCP. Secondly, I consider that any Host customers who are aware of BCP would understand it to be a retailer of a range of glassware products sold under a range of different brands. Thirdly, to the extent that such customers perceive that BCP has a trade connection with products that it sells, those customers are likely to do so by reference to the common use of the 701 mark or the modified 701 mark. Without the presence of that mark, in my view they are unlikely to consider that the word “Bohemia” as it appears in the impugned uses connotes a connection or association with BCP. Needless to say, no such mark appears in the October 2015 catalogue. Fourthly, such customers would also be influenced by the geographical nature of the term and the material differences between the Host and BCP products such as price, quantity and quality. ….
Accor has trade mark registrations for “Cairns Harbour Lights” and “Harbour Lights”, which it used to promote accommodation at the Harbour Lights complex in, you guessed it, Cairns. It sued Liv Pty Ltd which was renting out apartments owned in the complex by others.
Amongst other things, the trial judge had held that “Cairns Harbour Lights” should be expunged from the Register and the registrations for “Harbour Lights” should be amended to remove some of the services including “accommodation rental services” and “rental of accommodation”. Liv, however was found to infringe through the use of:
(a) “Harbour Lights Cairns”; and
(b) “cairnsharbourlights.com.au”; and
(c) “harbourlightscairns.com.au”; and
(d) “harbourlightscairns.com”.
The Full Court has now allowed Accor’s appeal, revoking the trial judge’s orders to expunge “Cairns Harbour Lights” and remove some of the services for which “Harbour Lights” is registered.
At 360+ paragraphs, more detailed consideration will have to wait.
One interesting aspect is that the Full Court confirmed that Liv’s use of “keywords” (really metatags) in the source code of its website was trade mark use and so infringing:
323 The title used in the source data is “Cairns Luxury Accommodation – Waterfront Apartments – Harbour Lights – Cairns Queensland”. The primary judge finds that the use of the words “Harbour Lights” in that title appears to be merely a description of the waterfront apartments referred to in the title: PJ at [434]. As to the use of the keyword “Harbour Lights” (as described by the primary judge at [430] and quoted above), the primary judge regarded that use as also a reference to the apartments as those words appeared in the context of surrounding words such as “Cairns apartments”, “waterfront, luxury apartment” and “harbourside”. Thus, the words were not used as a badge of origin: PJ at [434].
324 The other words used in the source data as recited at [430] by the primary judge are these:
“content: = Harbour Lights Apartments in Cairns offer luxury private waterfront apartment accommodation for holiday letting and short-term rental”.
325 As to those words, the primary judge finds that the use of the words “Harbour Lights Apartments” in that phrase was, effectively, use as a business name for a business which offers “accommodation for letting and short?term rental” thus operating as a badge of origin to distinguish Liv’s services from others: PJ at [435]. Such use is use of a mark substantially identical with and deceptively similar to each of the registered trade marks in suit. It is use in relation to each of the Class 36 and Class 43 services other than “commercial real estate agency services”, “agency services for the leasing of real estate properties” and “hotel services”: PJ at [436]. (emphasis supplied)
The Full Federal Court has held that the licensor must actually exercise control over the licensee for a trade mark licence to be a valid licence.
The decision is part of a long running global battle between WILD TURKEY and WILD GEESE. The WILD TURKEY interests own and use WILD TURKEY around the world for bourbon whiskey; the WILD GEESE interests use, or want to use, WILD GEESE around the world for Irish whiskey. Instead of the usual battle about who was first to file and whether or not WILD TURKEY was confusingly similar to WILD GEESE or vice versa, there was an unusual twist in this fight: WILD TURKEY tried an end run, tacking on to a registration for WILD GEESE WINES.
Some background
A Mr O’Sullivan QC (and his partners) had established a winery in South Australia under the name WILD GEESE WINES (WGW) in 2000. In due course, WGW set out to register their trade mark. However, the WILD GEESE interests had already registered their trade mark in Australia for whiskey. It was cited against the WGW application and in 2005, WGW brought an application against the WILD GEESE interests’ registration to remove it for non-use. The WILD TURKEY interests had also brought a non-use action against the WILD GEESE interests’ registration.
Mr O’Sullivan (and partners) quickly came to the realisation that they did not to become embroiled in the intergalactic war being waged between WILD TURKEY and WILD GEESE whiskey. Instead, in 2007 WGW assigned its trade mark application and the benefit of its non-use application to the WILD TURKEY interests in return for an exclusive licence to use the trade mark in Australia for wine.
The non-use applications against the WILD GEESE interests’ trade mark was successful and the (now) WILD TURKEY interests registered the WILD GEESE WINES trade mark for wine and spirits that WGW had assigned to them.
In a case of sauce for the goose potentially being sauce also for the turkey, the WILD GEESE interests then brought an application to remove the WGW trade mark for non-use. The WILD TURKEY interests sought to defend that claim on the basis that the use of the trade mark by WGW was authorised use under the Act and so constituted use in the relevant period by the WILD TURKEY interests[1] as registered owner sufficient to defeat the non-use application.
As the removal application by the WILD GEESE interests was filed on 27 September 2010, the three year period in which the WILD TURKEY interests had to show use as a trade mark in good faith ran from 27 August 2007 to 27 August 2010.
Unfortunately for the WILD TURKEY interests, there were a few wrinkles.
WGW produced a merlot under its trade mark in 2004. Due to adverse climate conditions, it did not produce another vintage until 2011. However, wine from the 2004 vintage was for sale (and was sold) in relatively small batches during the non-use period.
Mr O’Sullivan (and his partners) realised that a valid trade mark licence required that the licensee’s use be under the licensor’s control. To that end, Mr O’Sullivan proposed quality control ‘conditions’ for inclusion in the licence:
WGW’s wines had to be of sufficient quality to qualify for an export licence from the Australian Wine and Brandy Corporation;
WGW had to supply samples of their wine to the WILD TURKEY interests if requested to do so.
Notwithstanding this, the licence arrangements did not have any practical effect on WGW’s operations and the WILD TURKEY interests never requested samples until after the WILD GEESE interests brought their non-use application.
The Registrar upheld the removal application. On appeal, Perram J considered that the Full Court’s decision in Yau Entertainment bound him to find that the possibility of control being exercised was sufficient for a valid licence and so, very reluctantly, allowed WILD TURKEY’s appeal.
The Full Court’s decision
All five judges considered that Yau Entertainment did not rule that the potential for the exercise of control by the licensor was sufficient for authorised use under the Act.
Control
Besanko J gave the leading judgment with which Allsop CJ and Nicholas J agreed.
After a detailed review of the legislative history and the case law, Besanko J concluded at [95] – [98] that “control” for the purposes of s 8 meant actual control. At [95]:
The meaning of “under the control of” in s 8 is informed by the principle stated by Aickin J in Pioneer, that is to say, that the trade mark must indicate a connection in the course of trade with the registered owner. The connection may be slight, such as selection or quality control or control of the user in the sense in which a parent company controls a subsidiary. It is the connection which may be slight. Aickin J was not saying the selection or quality control or financial control which may be slight. I think the principle stated by Aickin J informs the meaning of “under the control of” ….
His Honour acknowledged at [98] that whether there was actual control was a question of fact and degree, but “there must be control as a matter of substance.”
His Honour recognised that this conclusion was different to the conclusion reached under the Trade Marks Act 1994 by the House of Lords in Scandecor. That however was because UK law had taken a different course under the influence of EU law. Similarly, the CJEU’s decision in Ideal-Standard[2] was directed to a very different issue: exhaustion of rights.
WILD TURKEY did not actually exercise control
Besanko J went on to find that the WILD TURKEY interests did not actually exercise control over WGW’s use of the trade mark. Bearing in mind that it was a question of fact and degree, his Honour considered the most significant factor was that the licence arrangement had no practical effect on how WGW conducted itself.
At [107]:
The quality control provision in the Licence Agreement is that the wine be of a sufficient standard to obtain the approval for export of the AWBC. There was no evidence of the precise content of that standard. It was not an exacting standard as the approval rate shows (at [51] above).[3] The primary judge considered that the standard involved no more than a rejection of what he called truly undrinkable wine (at [55]). It is plain that the standard had no effect on Mr O’Sullivan’s wine making practices. He was interested in making good to high quality wine. At no time during the relevant period did [WILD TURKEY] contact Mr O’Sullivan about the wine he was making or selling or both. There was never any request by [WILD TURKEY] for samples under cl 3.1 or for the product to be supplied to the Australian Wine Research Institute under cl 3.2. [WILD TURKEY] never asked Mr O’Sullivan for any information about the use of the trade marks or Mr O’Sullivan’s wine making operations generally. There was no monitoring by [WILD TURKEY] and nothing to suggest that [WILD TURKEY] took steps to ascertain whether the terms in cl 3 were being complied with. I do not think s 8(3) was satisfied by the existence of cl 3 in the Licence Agreement.
The conditions in the licence that WGW could use the trade mark only for wine it manufactured and only on wines sold in Australia were restrictions, but they were not restrictions that went to the quality of what was produced necessary to maintain the connection in the course of trade with the (putative) licensor. At [108], his Honour explained:
…. These are restrictions but not ones like controls on quality or manufacturing process which might suggest a connection between the registered owner and the use of the trade marks in connection with the provision or dealing with goods in the course of trade. There is no evidence that [WILD TURKEY] monitored or informed itself as to whether WGW was only selling Australian wine in Australia. These requirements do not give rise to control. WGW was not permitted to amend or abbreviate the trade marks or use them in a scandalous fashion. These provisions seem to me to be standard provisions to be expected in a licence agreement for trade marks. There is no evidence of monitoring by [WILD TURKEY] of these provisions and they do not amount to control within s 8. Finally, the provision about standard liability insurance and [WILD TURKEY]’ ability to terminate the Licence Agreement for a material breach is not sufficient to constitute control under s 8 of the Act.
Thus, the use by WGW was not authorised use and the registrations for WILD GEESE for wines should be removed for non-use.
Some other points
Nicholas J agreed with Besanko J’s reasons. Nicholas J also pointed out that the use which would defeat a non-use application under s 92 had to be use as a trade mark in good faith. His Honour considered that the failure by the WILD TURKEY interests to exercise actual control over WGW would be a factor disqualifying that use from being use in good faith. As this line of attack was not actually argued by the WILD GEESE interests, his Honour did not decide the case on this basis. nonetheless at [132], his Honour said:
However, in considering whether or not the registered owner has exercised sufficient control over another person’s use of a mark so as to defeat an attack on the grounds of non-use, it is important to recognise that the boundary between “use” and “use in good faith” by the registered owner cannot be defined by a bright line. This is because the question whether there has been any use by the registered owner may itself depend on whether the control it is said to have exercised was real or genuine control as opposed to something that was merely token or colourable.
Allsop CJ agreed with both Besanko J and Nicholas J.
Katzmann J also found that authorised use required the licensor actually to exercise control over the licensee. That had plainly not happened in this case. Her Honour did accept that the WILD TURKEY interests’ request for samples in 2011 (after the non-use period and after the WILD GEESE interests had filed their non-use application) could lead to ‘a “‘retrospectant’ circumstantial inference”’[4] that control was actually exercised. But the inference that control had not been exercised was also open and, as the WILD TURKEY interests had not shown the inference they contended for was more probable than not, they would still lose. Her Honour pointed out that the wine show medals that the WILD TURKEY interests relied on to support the good quality of the wines did not survive scrutiny. The judges’ comments at the wine shows included:
Very disappointing class with no highlights. From this class it would appear to be unsuited to the region. No wines showed any varietal character or even just brightness of fruit and character.
Perhaps more importantly, there was no evidence that the WILD TURKEY interests had any idea that WGW’s wines had won any medals or whether the wine was of good, bad or indifferent quality.
Greenwood J also concurred in the result, but was not prepared to condemn the licensing arrangements between the WILD TURKEY interests and WGW in the strong terms used by the trial judge.
Wrap up
So, if you are acting for a trade mark licensor, make sure that it actually exercises control over its licensee(s). And, at least when the control relied on is quality control, make sure the control goes to the quality of the goods or services provided under the licence. The use won’t be authorised use otherwise. In that case, the licensor won’t be able to rely on it to defeat a non-use application as in this case. Even if that is not a risk, there will also be the danger that use which is not authorised use may render the trade mark deceptive and liable to cancellation.
If you have a comment or a question, please feel free to post it in the comments section. Or, if you would prefer, email me.
IHT Internationale Heitztechnik GmbH & Anor v Ideal-Standard GmbH & Anor [1994] 1 ECR 2789. ?
In the year ending 30 June 2010, only 40 wines out of 18,019 wines tested ultimately failed to receive export approval, and the figure in the following year was 43 wines out of 14,569 wines tested. ?