"Exotic" marks, being colour, smell and shape marks, are notoriously tricky to register. Cadbury's purple mark and Mattel's Scrabble tile mark have recently been under the scrutiny of the Court of Appeal. So what was the outcome? These types of marks can be registered, but the devil is most definitely in the detail.
Both Cadbury and Mattel came off badly. Their registrations were deemed not to be signs capable of graphical representation as they each covered too many variations and so were not precise.
Both parties attempted to rely on acquired distinctiveness to support their marks. Proving acquired distinctiveness can get you past the requirement to demonstrate that your sign, which is capable of graphical representation, is in fact capable of distinguishing the goods or services of one undertaking from those of another. However, such distinctiveness is irrelevant to whether the thing in your application is a sign and whether it is capable of graphical representation.
It is all too easy to scrutinise the applications in these two cases and find them wanting. But it would be useful to know whether the marks were in accordance with the Registry guidelines applicable at the time and, therefore, whether the current scrutiny is very much one of hindsight.
Article 2 of the Trade Marks Directive 2008/95/EC of 22 October 2008 (TMD) requires all marks to be (a) a sign; (b) capable of being represented graphically and (c) capable of distinguishing the goods or services of one undertaking from those of other undertakings.
The Court of Justice of the European Union (CJEU) has considered exotic marks in a series of well-known decisions. These include: Libertel Groep BV v. Benelux-Merkenbureau Case C-104/01  ECR 1-3793, Sieckmann v. Deutsche Patent -und Markenamnt Case C-273/00  ECR 1-11754, Heidelberger Bauchemie GmbH v. Bundespatentgericht Case C-49/02  ECR 1-6129 and Dyson v. Registrar of Trade Marks Case C-321/03,  ECR 1-687.
The outcome of the above decisions has been a detailed set of conditions but which may be (clumsily!) paraphrased as a need for precision, clarity, specificity, self-containment, objectivity and uniformity.
In short, your application shouldn't amount to a multitude of forms that arise out of one registration (the solution is to register additional marks). This need for precision raises obvious difficulties, particularly in relation to colour, which may be overcome in part by reference to a Pantone colour (or other recognised reference system), depositing a swatch of such colour and a prose description.
Cadbury's purple - Société des Produits Nestlé S.A. v. Cadbury UK Limited  EWCA Civ. 1174
Cadbury sought to register "The colour purple (Pantone 2685C), as shown on the form of application, applied to the whole visible surface, or being the predominant colour applied to the whole visible surface, of the packaging of the goods."
The description was accompanied by a rectangular swatch of colour and a specification (post amendment) of "chocolate in bar and tablet form; chocolate for eating; drinking chocolate; preparations for making drinking chocolate."
Mattel's Scrabble tile - (1) J.W. Spear & Son Limited, (2) Mattel Inc, (3) Mattel UK Limited v. Zynga Inc.  EWCA Civ. 1175
The Mattel mark consisted of the shape on the left and the description "The mark consists of a three-dimensional ivory-coloured tile on the top surface of which is shown a letter of the Roman alphabet and a numeral in the range 1 to 10.".
It was registered in respect of "Computer game adaptations of board games" in Class 9, "Board games" in Class 28 and "Organisation of competitions and exhibitions, all relating to board games"
What did the debate (and the judgments) hinge on?
In relation to Cadbury, the debate was - what does "or being the predominant colour" mean?
People can understand the Pantone reference and the colour swatch and presumably accept that it is to be applied to the packaging of the specified goods. But when is such a colour the "predominant" part of the whole visible surface?
Take the example of a typical tablet of chocolate. If the whole of the upper surface is purple but the side surfaces and the lower surface of the external wrapper is white, then the predominant area of the wrapper is white, not purple. However, the consumer in a shop will regard it as predominantly purple.
Likewise if such a wrapper consisted of yellow and purple dots but the purple ones were slightly bigger, then the percentage of the wrapper covered by purple would be greater than the yellow and so purple would be predominant. In this instance, it's very unlikely that a consumer would consider such a design to emanate from Cadbury.
The uncertainty engendered by the alternative "or" and the confusion regarding the meaning of "predominant" meant that Nestle's appeal was granted. The Cadbury application was simply not "a" sign, it was potentially many signs and consequently it was not capable of graphical representation.
The fact that the Cadbury purple is enormously well-known, to such an extent that the buying public can easily distinguish even other purple wrapped chocolate products (eg Milka) from Cadbury purple wrapped chocolates was accepted and acknowledged.
But this fact only goes to the assessment of whether the sign is capable of distinguishing the goods and services of one undertaking from those of another, not whether any particular thing is a sign which is capable of graphical representation.
In relation to Mattel the problems were manifold. How big is the tile? What is "ivory-coloured" exactly? What letter, how big should it be and where should it be on the top surface of the unknown size and colour tile?
Likewise, how big should the numeral be and where should it be? Are such letters and numerals to be in any typeface at all? Also where are the letters and numerals to be placed relative to each other?
It is very clear that the Mattel registration simply cannot be "a" sign capable of graphical representation. The mark's rendition on the register was simply too vague and permissive of too many variations to give any certainty whatsoever.
The Scrabble tile appeal arose out of a summary judgment at the High Court. Mattel sought to argue that their appeal should prevail and that they should be allowed to prove the distinctiveness of the tile mark at trial. But again they were muddling acquired distinctiveness with the requirement that all marks must fundamentally be a sign capable of graphical representation before you get to any consideration of their ability to distinguish.
It is not clear whether the marks were, in fact, filed in accordance with Registry guidelines at the time their applications were filed.
It may not have helped either of the current cases, especially the Mattel mark, which was vague in so many ways, but there are many colour marks on the register that pre-date the line of cases requiring clarity in the form of Pantone references or similar.
Should they all be revoked now if challenged? The caselaw suggests that the answer is yes, especially in cases where the colour is what makes the mark distinctive. But is this the approach we should be adopting?
This hindsight view of the trade marks register could be a boon to would-be infringers; no matter that they are using almost an identical sign, as the registered mark can be taken off the register for want of being capable of graphical representation.
There will be many who consider this "cleaning up of the register" to be a great idea. But it could be very unfair on trade mark owners. Perhaps a solution could be to set up a simple procedure to allow post-grant amendment of a trade mark registration, provided it only narrows the mark's scope (for example specifying a Pantone reference), in a similar way that patents can be amended.