Incrementality startup INCRMNTAL only works with inbound leads.
That might sound like a humble brag, as in the company doesn’t need to market itself. But there’s a reason for it, said CEO and co-founder Maor Sadra.
Measuring the true impact of a marketing campaign, channel or tactic often reveals it’s not working – and some marketers just don’t want to hear that.
INCRMNTAL has lost customers in the past because “the truth hurts,” Sadra said. It’s not easy learning that a large chunk of your ad spend is ineffective or redundant.
But that’s the reality.
And for marketers that want to live in reality – because you can’t reallocate budget if you don’t know where it’s being wasted – INCRMNTAL released a new version of its platform on Thursday. The platform uses causal AI models to measure multiple channels simultaneously on an ongoing basis so that marketers can get a genuine understanding of campaign performance.
The real cost of acquisition
Causal AI algorithms use artificial intelligence to see the relationship between cause and effect. Correlation without causation (we’ve all heard that one before) is a common pitfall in marketing measurement.
Just because someone downloads an app after seeing an app install ad, for example, doesn’t mean they wouldn’t have taken that action organically or that the ad deserves credit for the conversion. (Looking at you, last click.)
INCRMNTAL’s algorithms analyze the marginal value of spend across channels, including mobile, web, streaming, TV, influencer and out-of-home.
Marginal contribution is related to incrementality, but there’s a nuance there. Whereas incrementality is a measure of whether there is any performance lift produced by a marketing activity, marginal value is a calculation of how cost-efficient the next dollar spent is at a marketer’s current spend level.
Or, more plainly: How much does it really cost to acquire your next user?
In many cases, the more money an advertiser pumps into a channel, the less efficient it gets – and therefore more expensive – as that channel becomes saturated. That’s why it’s important to measure total and marginal conversions, said Moti Tal, INCRMNTAL’s co-founder and CTO.
If a marketer can see that their marginal cost per install on a certain channel is getting too high, Tal said, they can consider scaling their campaign on other, less saturated channels.
Measure, lather, rinse, repeat, reallocate
But for marketers to get a complete view into their marketing performance, these calculations, calibrations and reallocations must happen constantly.
Marketers are in a perpetual state of motion. They make changes and daily tweaks to their campaigns – or even more frequently than that. Plus, marketing doesn’t happen in a vacuum because people don’t experience marketing messages or channels in isolation.
Say a marketer increases its bids on Facebook and simultaneously reduces its budget on Google. Did that contribute to conversions? Did it cannibalize organic results? Was there any impact on performance at all?
Those are the kinds of questions a causal AI model is designed to help answer. But every change a marketer makes is also an opportunity to run a mini incrementality test, Sadra said – minus the need for planning, A/B testing or holdout groups.
“Rather than a marketer having to pause all of their spending on a certain channel for multiple weeks to test whether it’s incremental,” Sadra said, “we measure and track all these changes, and each one is like a little live micro experiment.”
But saving ad spend from being wasted is only half the battle. That money needs to go somewhere else.
Now that INCRMNTAL’s new platform, which was in beta for the past three months, is fully launched, the company is working on developing a feature that will help marketers with scenario planning.
“Like oh, great, I saved a million dollars, but now where should I spend it?” Sadra said. “That’s coming soon.”
