The first thing I learnt from Gordon Brown (one of the founders of research company Millward Brown, not the PM), was that people don’t pay much attention to advertising. They only notice and remember certain elements/moments/aspects of ads i.e. the (few) bits they find interesting, entertaining or emotionally compelling. They ignore or quickly forget everything else - sometimes the whole ad.
 
This is why the best way to find out whether your ad is effective or not is to establish what, if anything, people find memorable about it. Then you need to determine whether the things people do remember are connected to the brand and intended message. This useful framework has been referred to as the 'Creative Magnifier/V-diagram’ (Millward Brown) and relates directly to the 'ABC’ of ad effectiveness (Unilever’s acronym following 'SMILE’).
 
Here’s an outline for what I consider to be the shortest, most valid design for an advertising per-test:
1.    SHOW AD
2.    Ask 'Is there anything about the ad that sticks in your mind?’ (Yes/No)
3.    IF YES: 'What sticks in your mind?’
4.    How strongly is XXXX (BRAND) connected with what sticks is your mind? (Strongly, Slightly, Not at all)
5.    Thinking about what sticks in your mind, what impressions does it give you about XXXX (BRAND)?
6.    POSSIBLY ALSO ASK: Which of these applies to you? (The ad makes me... Much more likely to buy the brand, Slightly more likely, No more or less likely, Slightly less likely, Much less likely)
 
Responses to these five (or six) questions, analysed appropriately, would tell you all you need to know about an ad’s potential effectiveness. I’d love to develop a streamlined research product based on this thinking if anyone is interested?