CRO Case Study: How to achieve a 53% increase in conversions in 30 days thanks to banner testing
Are you struggling to maximize the conversions of the banners on your website?
This is exactly what one of our clients was experiencing.
In these cases, our experience tells us that banner testing via A/B tests is the technique you need to clear things up.
Below, we'll dive into a remarkable case study, led by Fran Devesa, who has led the CRO team at iSocialWeb to deliver an impressive 53% increase in conversion rates in just thirty days through efficient banner testing.
Intrigued already?
If so:
Keep reading to discover how we pulled it off, don't miss a thing.
But first, here are the keys to this case study:
Key takeaways
- A/B tests on banners can deliver significant increases in conversions.
- In this case study, our client achieved an impressive 53% increase in conversions in just 30 days through a simple strategic test on one of their banners.
- Banner testing is widely underestimated today by professionals, despite being a great opportunity to improve conversion rates and digital assets' ROI.
- Choosing meaningful success metrics and ensuring statistical significance to properly structure A/B tests are essential steps for effective banner testing.
Starting point: how to attract more advertisers from the product listings
This case study is even more notable given that our client, a multilingual classifieds website, competes against much larger, more established players in the European market.
If you run a portal like this, this story will sound familiar:
The business model of classified ad portals, like Milanuncios or Wallapop, is based on building a platform that lets individuals and businesses post ads for goods or services they want to sell or buy.
These portals connect buyers and sellers, and the revenue model is based on advertiser subscriptions, generating recurring revenue streams.
Therefore, one of the biggest challenges for a classifieds site is convincing potential advertisers to subscribe and pay a monthly fee.
Without them, it's impossible to generate ad listings or revenue.
Hence the priority request our client made right after we helped them optimize the page and form:
Attract as many advertisers as possible in under 30 days.
The question was how to do it.
The problem: where to act to test a conversion element
From the start, the problem was obvious.
The advertiser capture zones had already been optimized and the rest of the conversion elements were aimed at potential buyers on the platform.
So, where to act in this case to capture potential advertisers?
For that, it was essential to:
- Find a placement on the site with enough visits for the results to be significant.
- And at the same time not impact the other conversion elements.
Consequently, we needed to pick an event that could yield a large sample and generate significant results for the client within a very short time frame.
Quickly, Fran came up with a simple solution to kill two birds with one stone.
He put forward the following hypothesis to confirm:
"Most advertisers enter the ads listing to check out the competition and compare offers against their own.
So, it's logical to think that if they see a competitor on the listings, it's because it's profitable and they're making money.
And that's a reason to also appear on the listing with their own offer.
Well, that's exactly the moment when we can get them to sign up on our site.
And reel them in."
How?
Very simple:
Creating banner-style ads to show on the product category sections where all advertiser offers are listed.
The solution: A/B test for the banners on product and service listing pages
Fran's proposal wasn't without risks.
But it was worth running a test to clear things up.
The chosen placement, product pages, concentrated a lot of traffic on the site.
And if we used a banner, these usually go unnoticed by users, so we had the advantage of minimally affecting the user experience.
So that's exactly what we did.
Leaning on Fran's reasoning, we decided to build an experiment for this type of page, proposing two banner versions to be served to users through an A/B test and, in the process, optimize conversions.
The process: how we did it
Following the A/B-testing best-practice manual, we opted to create banners with the message:
"Want to appear on this listing?"
Two different banner versions were created for this case.
The first version was more eye-catching, while the second was more subtle.
We decided to use the more subtle version, since most page visitors were customers looking for classified ads.
The more subtle version was designed to blend in among the real photos and illustrations of the other listings.
And two different sizes.
1. A rectangular one for the desktop version
We chose to show the banner in horizontal format after the first 4 items of the grid-style listing. The position was made clear to the client and it may vary depending on the page's needs.
And the other square-shaped one for the mobile version of the page, following the same layout as the previous scheme in figure 1.
It's important to note that the client's site is in 7 languages, so 7 banners were created in each language for the desktop version and 7 banners for the mobile version.
A total of 14 designs to complete the test in the different languages.
2. The tool: Google Optimize
The test was built using the Google Optimize tool, showing 100% of visits one of the two designed versions to first measure the impact on conversions and decide which design was more effective.
At this point, the most important thing was to answer and set up the A/B test quickly and with assurances.
Remember we had a very ambitious mandate from the client:
Attract new subscribers in under 4 weeks.
Well, thanks to Fran's experience and the client's team readiness, we were able to set up the banner testing very quickly and effectively.
By the end of the experiment, we'd managed to display the banner on more than 9,000 occasions and captured a total of 38 new subscribers in 30 days.
More than one new advertiser every day.
But you're probably eager to know which banner won the test.
And by how much margin.
Great!
Jump to the next section to find out and satisfy your curiosity.
The results: 38 new advertisers subscribed in 1 month
As you know, banner testing consists of creating multiple banner variations and running them simultaneously to see which one performs best.
In this case:
The banner test identified a variation that resonated with our client's audience much more than expected.
As you can see below:
The truth is we didn't expect significant differences between the two banners.
Since the design was similar and we didn't expect much of the impact to depend on design.
But the reality is that variant 1 delivered 53% more conversions.
And had we used that version, instead of getting 38 new clients, we would have gotten 50.
In any case, we exceeded the client's expectations by a wide margin.
Delivering a significant impact on their bottom line.
In conclusion: banners still deliver value when it comes to conversions
Our case study on how to achieve a 53% increase in conversions through banner testing highlights the huge potential this format has to attract customers.
And, of course, it showcases:
The impact of A/B testing on advertising performance improvement.
From our experience, this advertising resource has a very bad reputation and is underutilized.
But in this specific case, it has served to:
- Attract a significant volume of new subscribers.
- Improve the ROI of product listings through a new revenue stream.
By analysing and optimizing key elements like images, call-to-action buttons and landing pages, it's possible to effectively and quickly improve conversion rates.
But, as we've shown in this case.
Other elements also play an important role in conversion rate optimization.
Bringing all of this together is only possible with the right focus, experience and tools.
At iSocialWeb we have extensive experience optimizing web projects to deliver consistent results.
Don't miss the chance to improve your ROI and grow your revenue.
Write to us!