Application Download Offer Test
| VERSION A & B | |
 
How site visitors voted:
VERSION A (54%)
VERSION B (46%)
|
Actual Test Results:
Congratulations! No matter whether you chose Version A or Version B, you won either way. The difference between the versions’ initial clickthrough rates were nearly negligible. However, by the end of the somewhat lengthy conversion funnel, Version A pulled firmly ahead with a 9.38% better EPC (estimated Earnings Per Click) than Version B. So, just because initial clicks were about the same, didn’t mean final outcome was. WiderFunnel, a full-service conversion testing firm, ran this A/B test as one of a series of PPC landing page tests for client W3i, an application network for Windows apps and plug-ins, which owns Music Oasis. The tests were conducted on W3i’s proprietary testing platform. Luckily, this platform was able to measure not just initial clicks, but also EPC based on how many additional download offers for apps such as Yahoo! Toolbar and WeatherBug that visitors responded to during their initial conversion process. We suspect Version A ultimately won because the landing page had slightly longer copy, setting visitors’ expectations a little more clearly about what would occur after they clicked on the download button. WhichTestWon.com is sponsored by: WiderFunnel Conversion Optimization, the A/B and multivariate testing firm that guarantees Conversion Rate Lift. Click here for info on improving your site’s results – risk free – today! |
I have a feeling that version B didn’t do as well because the blonde looks like that girl on sites that the url doesn’t exist. You know the girl with the backpack in front of a college….
Interesting study.
@Ryan Graves LOL!
The second one looks a lot more “stock” and is overall not as nice or credible of a design IMO. Interesting test.
Version A won because of the look-and-feel similarity to Vista, and the glossy-plastic feel that most post-2003 users have come to expect from apps. It’s a hardware-software branding thing in look-and-feel that many webdesigners have not yet 100% capitalized on. The color scheme and layout of Version B harks back to table-driven html design era, something you don’t see too often anymore in webworking.
The “size” of the brunette is much larger while the blonde, is smaller and wearing a fade-into-the-background red singlet. I wonder what would have happened if she was the same size and wearing a colour that popped?
Arghh – I meant she was wearing a GREY singlet. What if she worn red instead!
I think it’s girl A because she looks more like a rocker chick and the ad is for downloading music. The other girl looks like she should be selling clothes, flowers, or something boring. The girl with black hair seemed more edgy.
“A” won because it looks like a legitimate offer for a legit service. Version “B,” despite the cute blonde, looks far more “scammy.” In addition, “A” tells you HOW it works. Doing so makes the service seem more substantive, as if there’s actually a well-backed company behind it. Verison “B,” by contrast, with its lack of details, does little to ease the minds of those leery of malicious downloads and other come-ons. Basically, “B” ANSWERS NO QUESTIONS. This LEAVES questions in the readers’ minds, resulting in less click-through.
Another fun one… thanks Anne!
Drew Eric Whitman, D.R.S.
Direct Response Surgeon(tm)
http://www.Cashvertising.com
Author of:
“CA$HVERTISING:
How to Use More Than 100 Secrets
of Ad-Agency Psychology to Make
Big Money Selling Anything to Anyone”
I choose Version A because the blonde looked washed out and blended in with the background. The brunette, on the other hand, stood out more and vibrated more energy.
“We suspect Version A ultimately won because the landing page had slightly longer copy, setting visitors’ expectations a little more clearly about what would occur after they clicked on the download button.”
How can we verify this?
That remembered me of an advice I’ve heard before.
“10000 clicks in an ad that says GET RICH NOW and lead to a page that sells acne products isn’t worth nothing.”
That, not to that extreme extent, applies here, maybe page A was more closer to what people would have to do/get after the click and since people would click anyway (having to pay or not) for this, A was more successful.
I’m just a beginner, cannot guarantee that, but I really believe that.
Anyway thank you for your website.
[...] This test resulted in something very interesting that remembered of an advice I’ve heard a while ago, which says: If you get 10000 clicks from a GET RICH QUICK ad and your page sells something different from a get-rich-quick solution, then all those 10000 clicks were useless. [...]
Girl A is listening to the music. Girl B looks like she works in a call center.
I do think it’s interesting that neither one outperformed the other initially, but I wonder about the lengthy funnel and why it led to differences in conversion rate.
The blonde in B is looking directly at you, she’s too engaging. You’re forced to look at her which distracts from the process of getting free music. In version A the girl is focused on the offer and more passive, so you’re less likely to study her face and instead follow her gaze. Engagement is good for banner ads but once on landing pages focus should be on the offer and action. Landing pages don’t always have to look like ads since the user is already there.
Also the colors in A are softer and can hold the eye better than the black and white contrast in B which makes the eye want to dart around slightly also.