Stop targeting who you think should buy

We built the audience carefully. Demographic filters, behavioral signals, geographic radius all pointed at the right person. The campaign ran, the results came back, and the people who actually converted looked nothing like the profile we'd built.

Sound familiar?

It happens across verticals and budget levels, and most marketers have seen it at least once. You go back to the deployment data, you look at who engaged, who clicked through, who showed up, and there's a gap between that group and the audience you were convinced you needed to reach. Sometimes it's a small gap. Sometimes it's wide enough to explain why a campaign that looked well-built came in under target.

The gap isn't usually a media problem or a creative problem. It's an audience assumption problem.

The customer you want is not always the customer you have

Every brand carries a picture of its ideal buyer. That picture often reflects who leadership wants purchasing the product—a certain income bracket, a certain lifestyle, a certain set of values that aligns with how the brand sees itself. There's nothing wrong with having that picture. The problem is when it drives targeting decisions.

We've watched this play out in automotive more than anywhere else. A dealership wants to reach a certain kind of buyer, so every campaign gets built around that image. Meanwhile, the people actually buying the cars aren't the ones being reached. They're financing a used SUV, responding to a clearance event, driving 40 minutes from outside the target radius.

The gap between those two groups is where money gets wasted. And the fix isn't a better creative or a bigger audience. It's taking the data from who actually bought and building your next campaign around people who look like them.

What the conversion data shows

One of the most clarifying things a marketer can do is run a matchback. Take the post-campaign sales file and compare it against the original data. Who was in the audience you targeted? Who ended up converting? Where do those two groups overlap, and where don't they?

When we run this kind of analysis for our customers, the results are often humbling. The people who converted were frequently present in the data, but they weren't the ones the campaign was built around. The targeting logic pointed the budget somewhere else.

That gap is where spend leaks. And the fix isn't a bigger audience or more impressions. It's building your next campaign around the people who looked like your actual converters, not the people you expected to see.

If you can bring your cost per acquisition down by tightening the audience to people who genuinely look like buyers, you either spend less or you reach more of the right people with the same budget. Neither outcome is bad.

‍The data has to be trustworthy first

None of this works if the underlying data is stale or shallow. Before you can model off of conversion behavior, you need data that's worth modeling from.

There are four things that make data actually usable. 

• Accuracy: Is the information correct?
• Freshness: Is it still correct? Someone who was in the market for a car two years ago has almost certainly already bought one. Marketing to them as if they're still shopping means paying to reach someone who's already moved on. 
• Depth: How many people does the data cover? 
• Breadth: How much do you actually know about each person beyond a name and an email address?

When all four of those are in good shape, you can start to do meaningful things with audience modeling. When even one of them is off, you're building on a shaky foundation, and the matchback results will tell you so.

‍Where email fits in

Email plays an important role in helping brands stay in front of the right audience. When a campaign runs, identity graph matching helps carry the same customer-selected audience across additional channels, creating more opportunities for people to encounter the message beyond the inbox.

That consistency matters. The more often the right audience sees a message across different environments, the easier it becomes for brands to stay visible throughout the buying process.

The mindset shift

Most marketers already know their conversion data exists. They just rarely go back to it once the campaign hits its numbers. Hitting the acquisition goal feels like the end of the story, so the audience assumptions that drove the campaign never get questioned.

That works until it doesn't. At some point a competitor with better audience data starts reaching your actual buyers before you do, or the cost per acquisition quietly drifts up because you're spending to reach people who were never going to convert.

The marketers who catch this earliest are the ones who routinely go back to the conversion data and ask who actually bought, whether they matched the original audience, and what a campaign built around the real buyer would look like instead.

It's a harder question to ask when things are going fine. It's also the question that separates marketers who consistently perform from ones who are surprised by their own results.

We built Site Impact around marketers because we understand how they think. And the best ones we've worked with, across two decades and thousands of campaigns, share one habit: they follow the data even when it contradicts the plan.