I've been playing around a lot with Traffic Junky. Getting familiar with the source, etc.
To date banners (animated and static) have had pretty good CTR's. But I noticed some animated placements (in good positions) seemed rather low.
Plus, the animated banners I was running were pretty decent.
Spent some time digging in and watching the placement (refreshing, etc.) - hard to do serious research in this niche by the way.
Anyways, the animated banners all looked very similar, pure T&A in a loop...hypnotic almost.
So I decided to try the good 'ol pattern interrupt.
I took an existing animated banner I made (getting okay CTR, but not great) and added 1 additional animation frame in PS for 2 seconds - a scary looking smiley face from hell.
It's the same "fuck anything walking" banner but with that little twist. It takes about 6 seconds to cycle, then boom - a messed up image appears that totally contrasts, then the animation repeats.
I've attached the split test in a screenshot - highlighted is the ad with the weird image. Almost double the CTR as the same ad with no weird image.
Still very early on with this, but I thought that was quite interesting.
2 conversions so far.
Anyone else try stuff like that? Making me very curious now...
That's pretty fucking creative. I don't think the doubled CTR would be doubled revenue, but it could still give a nice bump in overall revenue.
Added to list of things to test.
I like these kind of ideas! Haha never tried it, but would love to see more data on this. The theory makes sense.
That looks like a pretty neat idea, but not always the results are good.
I tried the same thing in mobile , putting a cat in the animation for a split second.
The CTR increased a lot by conversions were way lower than the original one.
Curious to see how your ad did in terms of revenue.
Yep, I've tried this approach too. Whilst I haven't tested it a lot, my experience was similar to bbrock's - high CTR, not so much on the CVR. It's possible that more optimisation would solve that, though - very curious to see your results after a bit more traffic!
Yea, still testing.
@BBrock/caurmen - i'm curious about CR also. Whether people are just clicking because of a 'WTF?' reaction...or because of actual interest.
But early on the most god awful images (just for a second or 2 - either end or mid animation) seem to get the most clicks.
Will update once I have some serious data.
EDIT:
Also TJ has animation restrictions - like 2 second minimum frame transition on certain placements.
I think this could be even more effective if you decreased how long the weird/interrupt image was displayed...like 0.5 seconds.
Ah, interesting - I've used frames with about a 0.3 second duration for this trick in the past. That works very well to give the "TV channel change" flash effect.
I was BIG on this type of approach on Facebook. Sometimes it worked, sometimes it didn't. It ALWAYS got a lot higher CTR.
My best tip is to try this concept with some more relevant. Fake subliminal msg thats easy to catch, etc..