So I came across the following Facebook ad and it instantly grabbed my attention — mostly because the picture was so… disturbing might be the word.  But eye catching and interesting also fit.  See for yourself:




And this ad is a great example of having the Headline, Picture, and Body Text work together non-redundently.  By that I mean all the pieces fit together, but your “need to know” is provoked by both the picture and headline, forcing you to read the body text to satisfy your curiosity.


So what you get is the promise of seeing the results of an eye tracking test comparing pictures (presumably baby pictures like the kind in the ad) featuring either cake or worms.  To me, an advertising junky with more than a little experience and interest in eye tracking, the promise of those test results was more than worth a click.  In other words, the ad worked on a number of very important levels:


  1. The picture grabbed my attention amidst the visually-crowded and diversion-filled field of Facebook

  2. The headline and text intrigued me enough to READ the ad / offer
  3. The offer was compelling enough to garner the click


So a lot could be learned just at that level, but then things got even more interesting…




As you can see, the Landing Page instantly made good on the ads promise by both offering up a WhichTestWon-style split test quiz featuring the same baby picture from the ad. In the website optimization world, this match-up between messaging & pictures in the ad and on the landing page would be described as good “scent.” Except that…


The sample ads being tested had multiple variables.  It wasn’t the same page with simply different pictures, because the “cake” page strategically used visually prominent and “hot” colors to draw the eye up towards the top of the page where the product/offer/messaging was placed.  So even if the worm picture is more eye-grabbing, the comparative visual prominence and eye-tracking performance of the two ads might not reflect that.  And indeed that was exactly the case as the next click revealed this:




This next page showcases the superior visual attention-getting properties of the high-color ad featuring the cake pictures — at least according to 3M’s proprietary Visual Attention Service heatmaps.


Unfortunately, though, the page doesn’t show us the heatmaps from BOTH ads so we can see for ourselves – we only get shown the winning ad and have to take 3M’s word for the results.  So that’s something of a fail.


And then there are the factors I mentioned before: that the use of visually prominent colors in the alternate ad “biased” the results away from what most people would have assumed was the more attention-grabbing ad.


And how do I know that the baby-with-worm picture is really the more attention-grabbing photo?


Because that’s the picture 3M chose to use for their own (real) Facebook ad!


So what are the takeaways from this?


2 Things:


  1. Visually Prominent Colors are no joke when it comes to catching people’s eye.  This is a tactic we’ve covered in this column before, usually taking the form of a bright red border around the ad’s picture.  But it can even be used as a red labelled call-out, as demonstrated by this ad:




  2. Pictures with faces, “Story Appeal,” and/or “Alarm Triggers” are equally attention-grabbing, even if they are not as visually prominent.  This is because they work on an emotional, rather than purely perceptive level. When you see the baby face, your eye is drawn to it, because we are hard-wired to pay attention to faces and babies. And when you see the baby holding a worm, alarm and curiosity further peak your emotional response.  In other words, the picture both draws your eye and your sustained attention, causing you to actually read the ad.  A red label might get me to “see” your ad, but it hardly guarantees that I’ll read it.  But a great picture like the baby with worms usually WILL get me to read the ad.  See the difference?


So why not use both techniques?  Good question: why wouldn’t you?  At the very least, it would be worth testing, don’t you think?