Measure The Lift Before You Run It
Co-founder and CEO of Swayable, the data platform that measures whether creative will actually move people. His case: click-through rates are a poor proxy, and the time to know is before the campaign runs, not after.
"Swayable measures persuasion."
James Slezak is co-founder and chief executive of Swayable, the data platform that measures whether advertising creative will cause lift, used by some of the world's largest brands, their agencies, and strategy consultancies. His argument is that click-through rates and view counts are poor proxies for the ideas that actually move people, and that measuring persuasion before a campaign runs, rather than after, is what lets marketers act on the evidence. Swayable's clients span CPG, technology, telecoms, and Hollywood studios, including Paramount Pictures across its major theatrical releases.
Swayable exists to answer a question most marketers have very little real data on: will this creative actually make someone want the product. The familiar metrics, click-through rates, view counts, engagement, are poor proxies, and James argues they can actively lead you toward the ideas that do not work. The brand lift study tries to answer it, but arrives too late, after the campaign has already run. Swayable's whole proposition is to deliver that measurement of lift before the spend, in a test that returns in around 24 hours, so the decision can still be made.
The company is deliberately a product company, not an agency. The majority of the team and the investment is engineering, PhDs with backgrounds in published experimental science, building randomised controlled survey experiments into a cloud product. James's analogy is handing a marketer the keys to a Ferrari rather than the engineering manual: the statistical machinery is sophisticated, but the marketer should be free to focus on the story, not on how to run the trial. Because the test happens in Swayable's own first-party environment, it is platform-agnostic, the same method works for a Facebook ad, a Google ad, or an Airbnb billboard.
The deeper shift James points to is in how marketers choose who to reach at all. The segment with the highest existing demand is not always the one a campaign can move most, and marketing, in his view, is about increasing demand, not just harvesting people who already have it. Brands that build iterative test-and-learn loops on this see the biggest gains: a 2x improvement in messaging effectiveness is common, and in the strongest cases, with the discipline to feed each result into the next campaign, he has seen as much as 10x.
"Ideally before, because then you can do something about it."
"You can't just point at click-through rates and view counts and engagement and say, well, that's a proxy. It's really not."
Most big-idea marketing is meant to make people want the product in the first place, to love the brand or trust it. Engagement metrics do not measure that, and James argues they can steer brands away from the ideas that actually cause lift.
"The limitation is you get it too late. You've already run the campaign at that point."
The brand lift study is worth doing, but it grades homework after the fact. Swayable's whole reason to exist is delivering the same measurement before the spend, in around a day, so the decision can still change.
"When people do those kinds of iterative test and learn processes, we've seen increases as high as 10x."
A single test commonly doubles messaging effectiveness. Brands that build the discipline to feed each result into the next campaign, the way Grupo Bimbo built learning labs, see the gains compound far beyond that.
Stay close to the conversation.
New leaders and insights. No noise.
Thanks, you're subscribed.
Look out for the next edition in your inbox.