Wednesday 15 September 2021

Afro Leo


In the recent case of iCollege (Pty) Ltd (iCollege) v Excellence Skills Development and Mentoring (Expertease Skills) and ZA Central Registry NPC (Domain Name Registry), iCollege (Pty) Ltd sued Expertease Skills for trade mark infringement and passing off for their use of the mark ICOLLEGE, and requested the transfer of the domain name from the Domain Name Registry. The citation of the Domain Name Registry is a mechanism for enforcement of the judgment, and they did not take part in proceedings.

It is apparent from the decision that the High Court, in dismissing the claim, has still not got to grips with the proper application of the trade mark infringement tests, and this post will focus on these errors and omissions and once again appeal to the court and legal practitioners to adopt and follow the step-by-step approach used in Europe (see earlier posts here, here and here) as it will help reduce errors in application. The majority of the decision surrounds the trade mark infringement claim under S34(1)(a) and so this post focuses on that.

The claim for passing off and Section 34(1)(b) and (c) are not dealt with in any detail and are abruptly dismissed as the Section 34(1)(a) claim was not successful, albeit that there are also obvious errors in this part of the judgement (for example at para 69 it states – “The test for passing off with regards to section 34(1)(c) is set out by the court in Laugh It Off Promotions”).

Facts and findings

iCollege provides accredited and skills-based training with the Sector Education and Training Authority (SETA) at its 17 campuses nationwide and at an e-learning centre. It is the registered proprietor of two device trade marks for iCollege covering education and training services applied for in 2013 and 2018. The 2018 mark is the colour mark as indicated in the table below.

At para 4 the judge makes a finding that iCollege has made “continuous and extensive use” of the trade mark since 2013.

Expertease Skills is an internet-based business that provides E-Learning courses to consumers located worldwide for international, technology vendor specific information technology certifications. It uses the marks illustrated in the table below.


In dismissing the iCollege’s the High Court in Pretoria made the following findings:

Comparison of the services (paras16-26)

“From the papers it would thus seem that the main similarity between the business of the applicant and the respondent is merely the fact that both render services in relation to “education and training services” (I have already pointed out that the applicant’s trade mark is registered in relation to, inter alia, “education and training services.”) However, in my view, it does not follow that the mere fact that both parties render “educational and training services” would mean that they in fact render the same or similar educational and training services, nor does it follow that the target market for these services (the potential notional customer) is the same (para 26). This sentiment is repeated at para 39.

Whilst the judge is correct that the two parties do not provide the same services, she confuses the test of passing off for that of trade mark infringement. For trade mark infringement purposes the services are identical in that they are both education and training services or, put differently, the services covered by iCollege’s trade mark are wide enough to cover those in respect of which Xpertease Skills uses their trade marks. This is a critical error.

Comparison of marks (paras 48-55)

Trade mark infringement was pleaded under all provisions of Section 34 and the test is the same for each of the infringements contemplated by that section. What is required of the judge is to make finding as to whether the marks are identical, similar or different. If the marks are different, the test stops there – there can be no infringement.

In this case, as Xpertease Skills uses a different mark to that which is registered, the marks are obviously, not identical. The question then is if they are not different, they must be similar and what that level of similarity is. Importantly, for this part of the test, the judge should not determine whether there is a likelihood of confusion. That would be premature as it involves the interplay of the other parts of the infringement tests e.g. similarity/identity of services, notional consumer, strength of the earlier mark etc.

Para 48-55 unfortunately misapplies the test for similarity of the marks and conflates it with other aspects of the tests for infringement. This makes it difficult to determine what the court’s conclusion is on the similarity of the marks. The test is also restricted to a comparison on the logo marks in use and appears to ignore the fact that Xpertease Skills also use the word mark on its own. This would reduce visual difference in the marks.

At para 50 the judge finds that “but for the word iCollege / ICOLLEGE, the two marks are not, in my view, similar. The ultimate test however is, as stated by the court in Cowbell, whether, on comparison of the two marks, there is a likelihood of confusion”. Perhaps by this the judge is saying that the marks are only similar in that they include the word iCollege, and that the rest of her analysis focusses on that. However, even this liberal reading of the reasoning it does not make sense because it fails to then draw on the conclusion on the similarity/identity of the goods in assessing the likelihood of confusion.

There are also other errors such as the determination of “material link” para 54. The term material link as used in the context of the Verimark case cited in the judgement, and in Europe is there to determine whether the use of the trade mark by the alleged infringer is trade mark use (as opposed to pure incidental or descriptive use) and in the dilution test. In this case it is trite that both parties are using iCollege as a trade mark and so the concept of ‘material link” under S34(1)(a) is met. It is use of a mark by the alleged infringer in the course of trade in the provision of educational services.

Notional Consumer (paras 56 – 65)

The penultimate finding under trade mark infringement is at para 63 and illustrates that if there is a problem in the application of part of the test, it perpetuates itself:

“ Apart from the fact that the applicant placed no evidence before the court to substantiate a likelihood of confusion within the context of this matter, I am not persuaded that any customer would be confused into believing that the educational services presented by the respondent are somehow linked to the applicant. I have already pointed out that there are material differences between the content of the courses presented by the parties as well as their respective teaching methods. The common element of the marks (“college” or the composite mark “icollege”) is, at best, weak in that it is, as already pointed out, descriptive of the educational vehicle both parties use to offer their educational services. The mark does not describe the applicant’s or respondent’s educational services. In fact, the trade mark does not tell the potential customer anything about the product. As such, the trademarks do not serve as a badge of origin of the educational services it presents. Adding the prefix “i” (composite mark) also does nothing more than to refer to the internet.”

The evidence to substantiate a likelihood of confusion for trade mark infringement purposes is through the application of the applicable test. Whilst it may be said that more evidence could have better served iCollege’s case, there cannot be said to be “no evidence”  (for example, iCollege’s registration certificates are evidence of their rights). Furthermore the conclusions on likelihood of confusion are more appropriate to a passing off test.

The correct conclusions based on what is in the judgement in this case, appear to be better summarised as follows:

a.) the marks are similar if not highly similar in that they consist of the identical components “iCollege” which are a material part of the marks in the sense that they are the only verbal component, and visually prominent. They are also conceptually identical insofar as a meaning can be ascribed to them. The judge does conclude that the marks are at least similar.

b.) the services are identical in that those offered by Xpertease Skills are covered by the trade mark registrations. The judge acknowledges this but appears the be confused by the test of passing off which compares services in use.

c.) the marks have been “continuously and extensively” used (concluded in para 4) but are not well known. The marks consist of the descriptive words “college” and “I” for “internet” which are not strong components of the mark.  Portions of the registered mark(s) (eg COLLEGE and I meaning INTERNET) is/are therefore relatively weak mark inherently. However, as iCollege’s marks has/have been used over almost ten years it has elevated their distinctiveness. In layman’s terms, not as strong as the iPhone mark but still a protectable mark and, of course, it is registered. On balance one may say that that iCollege’s mark is deserving of ordinary protection.

d.) the consumer (of educational and training services) is an ordinary consumer including those seeking educational services online. Those seeking educational services online may have a heightened level of awareness referred to in YuppieChef and may be more discerning.

e.) there is trade mark use of iCollege by the alleged infringer (as opposed to non trade mark use or incidental use). It is unauthorized.

So, based on the judgement, we have unauthorised trade mark use of a similar (if not highly similar) mark applied to identical services covered by the registered trade mark offered to an ordinary consumer seeking educational services online, who can be discerning.

In order come to the finding that the court did, it would need be convinced that there are material differences in the marks or that the consumer is so discerning that they would not likely be confused. I am not convinced that there are material differences in the marks or that the consumer is that discerning, but then who am I. My point is that had the court  applied the test in this way (or perhaps had the papers have been pleaded in this way), the court would not make the mistakes highlighted above, be less open to an appeal and, if there was an appeal, the basis for the appeal would be clearer. It would make for greater clarity, expediency and clarity all round.

Afro Leo

Afro Leo

Subscribe via email (you'll be added to our Google Group)