Skip to Main Content
Haley Adams
Haley Adams | Research Operations Lead, Civis Analytics

Advertising is omnipresent in American life. Advertisers spent a total of $49.1 billion on television commercials in 2021, up 6 percent year over year, pitching everything from cars to crypto to cable TV itself. We all watch these commercials, we all think about them, and we all talk about them: some of them make us laugh, and others make us a bit misty. But none of that necessarily makes these ads persuasive — and isn’t persuasion what marketing’s all about?

Civis Analytics believes that persuasiveness — an ad’s fundamental ability to deliver relevant, resonant messages that inspire audiences to take action — is absolutely paramount, regardless of whether the message is a multi-million dollar commercial aired on Super Bowl Sunday or a simple text sent to donors on an average Tuesday afternoon. Emphasizing persuasiveness may seem obvious, but too many marketers continue to focus on how their ad makes the audience feel, not on what it inspires that audience to do, whether that’s purchasing their product, supporting their cause, or voting for their candidate. 

An ad that’s a crowd pleaser might ultimately do nothing for your organization’s bottom line; at the same time, an ad can be dull as dishwater, but the substance of its message can still strike a powerful chord with its intended audience. You won’t know your ad’s fate unless you test it for persuasiveness. Here’s how we do it at Civis.

How message testing works at Civis

Creative Focus, Civis’s online message testing tool, scientifically determines which ads have the greatest impact on audiences. Creative Focus utilizes randomized control trial experiments (widely considered the gold standard in social science) to precisely quantify an ad’s persuasiveness among key audience subgroups: advertisers can run multiple image, message, and/or video ad tests, and in a matter of days receive clear, concise data on which ads are most convincing, as well as which audiences are most convinced. 

Creative Focus also measures an ad’s potential negative impact, eliminating the risk of wasted spend on messaging that jeopardizes campaign goals or brand image. And if none of your ads has any effect one way or another, we’ll tell you that, too, because that level of transparency is just as important. 

Creative Focus testing follows a three-part process: 

  1. Sampling: We identify and survey a sample of the U.S. population representing the people you want to learn about. 
  2. Weighting: Within these samples, we correct for biases that inevitably come up when you conduct an online survey. For example, women are more likely to take surveys than men, so every female respondent in a sample gets down-weighted, and every male respondent gets up-weighted. These steps get us to a place where our data is truly representative of the general American population. 
  3. Modeling: The modeling process takes all the survey data and sorts out the random noise to determine the ad’s actual impact among the different subgroups in your audience. It might reveal whether an audience member likes a particular ad because it’s effective, or because she’s a woman, or because she’s a Democrat, or it might pinpoint different, more random reasons.

Developing subgroup models that accurately gauge how ads perform among different types of people is critically important. Even if you know an ad is 5 percent effective overall, without modeling you don’t know if it’s equally effective among men and women, or if it has no impact among women but a powerful impact among men. The more you drill down into specific subgroups, the more salient persuasion becomes.  

When Creative Focus testing is complete, Civis synthesizes the findings and delivers clients a deck depicting how their ads fared among different audiences — men versus women, Democrats versus Republicans, etc. Clients can use that information to figure out where their ads are most salient, and proceed accordingly. 

Consider the alternatives

Compare the Creative Focus approach to testing ads on Facebook and then measuring which one receives more clicks, or which generates more page views or video views. Here at Civis, we refer to those as “vanity metrics” — things that have no correlation with whether a message actually persuades a person to do something. (We know that clicking on ads is something nobody does in real life, for example, greatly diluting the inherent value of this tactic.)

Pre- and post-design surveys are another common methodology that Civis eschews. Here a survey might ask how likely you are to purchase Nike sneakers, then show you an ad, and then ask “Now how likely are you to buy Nike sneakers?” to measure the ad’s incremental impact. Sure, it’s advantageous over the likes-and-clicks approach because you actually get some connection back to what you want the ad to do. But people are smart, and research shows that when you put them in this kind of survey environment, they know they’re expected to say they’re now more likely to purchase Nike gear; others might think to themselves “Ehh, that ad wasn’t that good, so I’m going to say I’m not likely to buy Nike.” 

Creative Focus also sidesteps the conflicts of interest implicit in message tests conducted by publisher partners. It’s common for Facebook, YouTube, and other publisher giants to run message tests for advertisers spending above a certain threshold: they might say “We can do a brand lift study for you — we’ll set up an experiment inside our platform and show some people ads, withhold them from others, and then measure how effective the ads are.” The problem is that publishers have a deeply vested interest in wanting your ads to look effective, and that can sow doubts about the validity of their findings. 

Still another Creative Focus differentiator: Civis brings to the table a wealth of in-house survey science expertise, which is important because how you set up the survey — and how you word each question — can dramatically skew the results. In addition, we regularly research the survey panels we work with to ensure their consistency and quality, and we closely track the latest research on survey science to guarantee we’re following the latest best practices. 
Don’t treat persuasiveness as an afterthought as you build out your organization’s next marketing campaign. Put your ads to the test — the right test — and know with total confidence that you’re sending messages that matter to the people who matter.