How to optimize display ad performance with AI pre-testing

How do you measure performance of display ads?

The performance of most display ads is measured based on a very simple formula and three key performance indicators (KPIs):

• Total amount of impressions
• Click-through rates (CTR)
• Cost per mille (CPM)

The biggest problem with this system of performance measurement is just as simple: the data being used is almost exclusively gathered after the ads have gone live. And then there’s the fact there are several ways to determine what constitutes an ‘impression’.

Most DSPs and other publishers ‘count’ an ad to be seen when it’s been 60% in-screen for X amount of seconds. An impression is considered a ‘viewable’ impression when it has appeared within a user’s browser and had the opportunity to be seen. Not the purest metric to base your ad strategy on, right?

There are definitely better ways in which you could be measuring performance and consequently optimize your display ads. But first, let’s break down whether display ads as we see them today are effective, at all.

‍

How effective are display ads?

Display ads can be a very effective way to boost brand awareness, drive website traffic and increase conversions. So the short answer is: yes, display ads are an effective way to promote your products and services.

While display ads aren’t always portrayed in the most favourable light, they do often get a positive ROI for most advertisers. However, statistically, only 0.06% of people click on display ads, according to Google’s benchmark report. Not an amazing result, especially compared to the very first display ad, launched in 1994.

This banner ad racked up a solid 44% CTR. Long gone are the days of those results, so what happened?

Differences between attentive reach and low-cost reach

Over the years, the low-cost, immense reach of programmatic display advertising led to a broken advertising model, one that’s coincidentally also very prone to fraud - according to the ISBA, only 51% of advertiser spend actually makes it to publishers. 

Here’s what happened: 

Nowadays, most display advertising focuses on getting the lowest possible cost per thousand impressions (CPM). This type of advertising resulted in a slippery slope, where proper attribution fell into the background. This low-cost impression mindset with a strong disregard for attribution makes most modern display advertising practices incredibly wasteful.

The solution? 

Instead of blindly focusing on CPM, display advertisers should switch from a ‘low-cost reach’ to an ‘attentive reach’ model. As a matter of fact, even the World Advertising Research Center (WARC) has noticed how measuring attention to ads can bring more fairness and transparency and how it’s helping brands and publishers improve ROI.

And it’s not just neutral research associations like these that support a shift to an attentive reach model. Even international advertising agencies like Dentsu are trying to make a change in media trading in what they dub the ‘Attention Economy’. 

‍

Attention is the new name of the game, basically. Focusing on attention instead of low CPM will help you become a more effective advertiser.

When you focus on attention as a result for your display ads, you pave the way for your audience to have an emotional response. This, in turn, can actually trigger memory, drive behavior, and with it, desired outcomes for your display ads.

What makes a successful display ad?

When it comes to successful display ads, high quality content is king. Display ads are mostly shown in crowded digital spaces, where quality can really make you stand out.

Of course, you should also make sure your targeting is on point. Reaching the right people increases the relevance of your display ads drastically. The more relevant the ad, the more successful your display campaign will be.

We also suggest experimenting with the following display ad formats:

• Rich media to encourage interaction with the ad
• Video to boost overall CTR (up to 18% higher than other formats)
• Dynamic creatives that personalize your ads, depending on who’s seeing them

Regardless of form, at the end of the day, it’s all about making display ads that stand out. Optimizing display ads for success means designing ads that truly capture the audience’s attention. This triggers memory, and drives clicks, conversions, and increases in LTV.

‍

How can I improve display ad performance with AI pre-testing?

Nowadays, there are highly accurate AI-based visual attention tools on the market. Some of these deliver a staggering accuracy compared to real-life eye tracking. Basically, without using real-life test subjects and tracking their eye movement, you can get results that are 94% accurate, instantly. 

We already mentioned that a big problem with measuring display ad performance is due to the fact marketers only work with data gathered after the ads go live. Pre-testing your display ads can show you whether certain aspects of it will generate valuable attention.

A tool like Brainsight even provides ‘ad templates’, where your ad creatives are placed inside the context of a specific category (display, social, OOH). The attention your ad creative generates is then benchmarked against 10.000+ similar ads, resulting in an ‘attention score’.

Is your ad’s attention score higher than 65? This will lead to an average 59% increase in clicks, because people shown the ad to are proven to actually ‘see’ it. This shifts display ad metrics from ‘impressions’ to ‘valuable attention’ - attention scores become the new display ad KPI.

Besides, you do this research instantly, giving you the opportunity to optimize your display ads before they go live. This preemptive AI pre-testing will improve your display ad campaign performance, drastically. 

‍

How does an AI pre-testing tool like Brainsight work? 

The workflow is incredibly simple. You create a free account, upload your ad creative and select ‘display banner’. That’s it. Artificial intelligence will now generate an eye-tracking heatmap with 94% accuracy, and show you the ad’s attention score.

Is your attention score lower than 65?

Consider redesigning the ad until you raise that number. Use the AI-generated heatmap to find out which areas of the design need improvement. 

There’s even the option to use the tool’s API to automate this process even further, which could help DSPs, ad platforms, or creative studios with their own software, to increase the performance of their ads and creative content.

‍

What is the average ROI of AI pre-testing display ads? 

The short answer: it depends. First of all, AI pre-testing is, by default, a lot cheaper than using real-life testing services. So switching from real-life tests to an AI pre-testing tool like Brainsight will already effectively lower the costs of your display advertising campaigns.

But mostly, the return on investment you get from AI pre-testing depends on the budget you’re working with. 

Look at it this way:

Statistical analysis shows that ads with a Brainsight attention score score of 65+ reach almost 60% higher conversion rates on average than ads with score <65.

This means the ROI of using Brainsight to optimize your display ads is potentially 60% of your ad budget. Imagine your ad campaign delivers 10k clicks. By using AI pre-testing, this could’ve been 16k clicks. 

A CRO tool like Brainsight is priced at €179 per month (billed annually). This means the higher your display ad campaign budgets, the bigger the ROI you’ll get out of using AI pre-testing. Sounds like something you’d benefit from? Get in touch, we’d love to show you what’s possible.