Schedule a demo with Ad-Lib.io today

Have questions about how we make creative production easier? Fill out the form and an expert will be in touch shortly.
Join top global advertisers already using Ad-Lib.io

Beating Fatigue to the Punch

Ad-Lib.io
on
Apr 27, 2022
Blog

How AI Can Help You Anticipate a Decline in Campaign Performance

In our series, Engaging Creative, we dig into questions we get from clients about key levers of creative effectiveness and apply analytics to test hypotheses about the drivers of performance.

Advertisers frequently spend both time and money to optimize their audience targeting, yet, ironically, they let the same ad creative run on and on...and on and on…

The reality is that all that media optimization doesn’t matter if your audiences are just bored to death with seeing the same ad over and over. The result is fatigue, and it will crush the performance of your campaign. 

What exactly is ad fatigue? 

Ad fatigue is a decrease in campaign performance due to audiences becoming too familiar with your advertising. The more an audience sees the same version of your ad, the more likely they are to tire of it.

What impacts the fatigue of a single ad? 

We analyzed variant-level data to measure the interaction of impression volumes on declining CTR over time. As the chart below shows, the longer a single ad runs, the lower its predicted performance, particularly at high levels of daily impressions.


Interestingly, at the outset, there is negligible difference in CTR impact between ads served at different impression volumes. Over time, however, the more the daily impressions being served by a single variant, the more detrimental ad fatigue becomes. This challenge is especially prevalent for evergreen campaigns, like remarketing programs, that run for long periods of time and are seen by diverse audiences.

Impact from volume differences can also grow quickly. Using an average of all CTRs included in the analysis as a reference, we can measure the impact at different points of time. For example, after 100 days, a campaign whose CTR started at ~ 0.5% would have seen a ~30% decline in performance at a high impression volume, and only a ~20% decline at a low impression volume.

What are my options to slow the rate of fatigue?

Ad fatigue is typically reduced with three tactics: 

  • Frequency capping. Advertisers can restrict the number of times a single user sees an ad within a certain time window to avoid fatigue. This tactic saves media dollars by only delivering impressions while an ad is fresh and people are most likely to click. Frequency capping also preserves people’s experiences as they browse the web by reducing the number of redundant ads they see.

  • Scaling the number of variants in a campaign. People are less likely to tire of ads they see if they see lots of different iterations of them. Starting a campaign with many versions of an ad keeps content fresh and allows the best performers to rise above the rest over the course of the campaign to inform future marketing initiatives.

  • Mid-campaign refreshes. Frequently, once a campaign is live, marketers let the campaign run its course and perform a post-campaign analysis to generate learnings for the next program. However, when signs of campaign fatigue emerge, steps to refresh must be taken to boost performance.

But do campaign refreshes really recover a fatiguing campaign?

Our further analysis shows that an ad campaign refresh can have a significant impact on reviving interest, extending the attention-getting lifetime of the campaign, and, as a result, greatly improving the overall ROAS.

As the chart above shows, in-flight changes can recover and restore CTR, extending the lifetime value of the campaign.  Ad-Lib.io’s platform is designed to help advertisers easily create and approve refreshed variants of an ad. As one example, Estée Lauder partnered with Ad-Lib.io to launch creative refreshes in a timely manner that personalized content across four audiences with nine creative refreshes over the four-week holiday season, resulting in a 69% higher conversion rate.

However, it takes some time for an in-flight refresh to change the trajectory of a fatiguing campaign. So just imagine if you could optimize the timing of each in-flight change to your campaign, capturing the perfect moment to make the change? Not too soon - so that the effort and time to activate a new variant is not prematurely spent - and not too late - to avoid the tipping point of fatigue? 

Enter the role of AI - to predict fatigue before it happens.

What if you could be warned about performance drop-off before it happens? So you can get ahead of optimization and the creative sign-off process while your performance isn't trending down? Machine learning and AI models are the answer to predicting performance decline before it happens, allowing you time to optimize, but ensuring optimization doesn't happen too soon, wasting valuable assets and retiring creative before it's had a chance to deliver to its full potential. In fact, Ad-Lib.io models can anticipate fatigue in advance of it happening! 

For one customer in the beverage industry, refreshed creatives at the predicted moment created spikes in performance such that overall campaign CTR achieved a 44% improvement over the control and contributed to a 83% improvement in CPC versus the control. 

Net, net - not only is it important to optimize the performance of your campaign with a refreshed variant or set of variants, but the timing of that refresh can be optimized as well, to make sure you don’t either miss the window - and suffer a loss of interest from your audiences - or expend creative costs and energies too soon while your audiences are still with you. 

OK, so anticipating fatigue is now possible - but what about the work to create and activate the variants?

This is where Ad-Lib.io’s intelligent creative automation platform can help. Not only does our platform provide the creative intelligence that will predict the timing of fatigue at the campaign level before it happens, but it also leverages AI and automation to enable just one set of creative concepts, built with our Fix and Flex elements, to be expanded to hundreds of on-brand formats and messages - quickly and easily. With this creative intelligence directly integrated into Ad-Lib.io’s creative management tools, users can interrupt the effects of ad fatigue by taking optimisation actions based on the performance of element-level creative content. 

See the power of fatigue deterrence and creative variants in action!  Get in touch for a demo.

Footnote for the Analytics Techies

Our fatigue modeling (first graph) was fitted to variant-level data from April 2020 - April 2021 (N=8747) and filtered to exclude extreme CTRs (>2%) and variants with low serving (IMPRESSIONS < 1000).

The data found a significant negative main effect of daysServing (p<0.01), and significant negative interaction effect between Days Served and Impressions/Day (p<0.01).

Like our content?
Stay connected via the box below!
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Related Resources

Industry Awards