Case study example #1. Bigger product image

Skinner Auctions, of Antiques Roadshow fame, ran an A/B test blowing up the images in their online catalogue by 28%, from 250 pixels across to 350 pixels across. 

Unfortunately, as you can see here, the bigger image meant much of the page content was pushed below the fold for typical visitors (fold mark ours) forcing visitors to scroll.  

Responsive web designers know that on many sites, as few as 20% of visitors will bother to scroll down.    

Was it worth the risk?     

Skinner Version ASkinner Version B

Images from WhichTestWon, copyright protected. 

Absolutely! The larger image enticed 63% more visitors to click to start the bidding process. Even better, a whopping 329% more visitors who started bidding actually filled out all the online forms required to place a bid.

So, the larger image helped fickle bidders maintain their initial excitement as they worked their way through the bidding process.

Case study example #2: The attack of the 50 foot button

We think of this multivariate test as ‘the button that wandered near the nuclear power plant.’  The testing team for Profile Pimp, an online service targeting young adults, tried 45 different variations of its ad campaign landing page. 

All the versions used the same offer and much of the same copy. 

You can see the control version the team started with here. As you may notice, the button on the control was pretty big already. In fact, it’s larger than most marketer’s response buttons!

Profile-Pimp Version A

Images from WhichTestWon, copyright protected. 

But, the winning creative seen below, had a MASSIVE button. In fact, we’ve never seen a bigger button before or since.

Profile-Pimp Version B

Images from WhichTestWon, copyright protected. 

Measured test results were equally massive. The winning version had 135% higher immediate clickthroughs. Plus, the service gained 51% higher earnings per click as visitors made their way down the conversion path to purchase. 

Could you make your buttons bigger? It’s an easy change and worth testing.

Case study #3. Testing a mega-sized background image

This test shows that bigger images can work as well for B2B as they do for consumer marketers. 

Dell’s testing team wanted to increase lead generation. Their original landing page wasn’t bad at all. It was designed using best practices including no distracting navigation, useful bullet points, and easy-to-use form and no scrolling required.

Dell Version A

Images from WhichTestWon, copyright protected. 

Dell’s testing team tried a radical revamp for their test, using what we call the ‘mega-sized background image’. Basically, instead of using an image as one of many page elements, they used the image as the background that the rest of the elements sat on.

So, the background becomes your ‘white space’.

This is a responsive design test we’ve seen many other marketers using recently, including ecommerce sites in Europe. Did it work? 

Dell Version B

Images from WhichTestWon, copyright protected. 

The mega-sized image lowered visitor bounce rates by 27%, plus it increased leads generated by 36%

Dell was so impressed by these results that the testing team ran out similar redesign tests for other B2B product lines… and so far they’ve all raised lead generation and contact form conversions as well, sometimes into the triple digits.

Warning: we can’t guarantee larger images will always increase conversion rates. No one can tell you that. Your brand, web pages and site visitors are all unique, so what will work can’t be predicted with a simple best practices guideline. 

Also, as we’ve seen across many, many tests, the wrong image (and sometimes any image at all) can reduce conversion rates, often dramatically.  If your image isn’t compelling for your audience, making it bigger probably won’t help.

Our advice is to reduce your risk by testing. Always run an A/B or multivariate test until you have conclusive results or until 21 days have passed (whichever comes later) before you roll out a design change.    

Just use this article to inspire your design team (or the HIPPO) to let you get started testing.

Justin Rondeau is the producer of WhichTestWon’s The Live Event, an intensive two-day conference held each May featuring Case Studies and Workshops on how to optimize for better conversions.  Econsultancy readers can get $50 off their tickets by using the coupon code Econsultancy at