A group I had never heard of before, the Future of Privacy Forum, recently designed an icon for use on targeted online advertising. This symbol, and its design process is intriguing because it is far better than how most symbols are designed, although it was designed by OgilvyOne Worldwide, an advertising agency.
Now, advertising agencies are no slouches when it comes to in-depth research, but OgilvyOne is still an advertising agency, not an expert in symbol design, and there is nothing in any of the articles, or even their research report, to indicate they have any domain knowledge, or even made a cursory glance at any of the dozens of international standards and other guidelines for the design and testing of symbols.
Despite that, I think they did a mostly reasonable job in their research, though if they had had a passing familiarity with the proper testing procedures and existing symbol conventions, they may have produced a much better symbol. But 6 months after its introduction, this symbol still has not be used in any advertising I’ve seen, and perhaps part of the reason is that the design and interaction design of the symbol isn’t compelling enough for advertisers to buy in. And as the group did not define any interaction design, that is a large part of why it may not be compelling enough. Designing a symbol isn’t enough; people have to know that it will behave consistently to be fully comfortable with it, and making advertising subjects comfortable is entirely what this symbol was designed to do.
The set of three symbols that were tested by an OgilvyOne sister company (see Seeking a Symbol for ‘This Ad Knows About You’) are not all that varied, and show some classic newbie symbol design problems:
- The set selected for testing isn’t large enough, or dissimilar enough.
- There was no variation in colors, and no thought to selecting a color compatible with existing symbol and sign standards was given.
- One of the symbols look an awful lot like the head of a Phillips screw, for no good reason.
- One of the symbols was intentionally designed to appear similar to the international standard power symbol, but targeted advertising has nothing to do with cycling power, so this is an entirely inappropriate design intent.
As for the testing, OgilvyOne broke one of the cardinal rules in symbol testing by offering multiple-choice answers! The proper procedure is to present the symbol alone, or, in this case, in the proper context of an ad, and ask the test subjects to fill in the blank of what they think the symbol means, without any text description or other clues. This is how everyone except for the test subjects will view the symbol, so this is a necessary test method. And the results generally need to be 85% or better recognition without such prompting.
The Future of Privacy Forum’s final Behavioral Notices Study report does an excellent job of describing the response that typical Internet readers have to the symbols, and how the design of the symbol and the wording used with it changes how many of them feel comfortable with the targeted advertising process.