• by Nathan Hugenberger , July 24, 2026

Everyone loves "better" audience data. The fancier the identity spine, the more exotic the data source, the more impressive the pitch deck — the more confident everyone feels that this is the audience strategy that's going to change the game.

But here's the thing: careful analysis of new client accounts frequently finds that the audiences that came most highly recommended by the prior agency weren't actually the highest performing ones. Sometimes they weren't even close. That's worth considering.

Here are seven reasons why "better" audience data might not actually be better for your business.

1. Just because your agency makes more money on it doesn't make it better. Over and over, careful audits of new client accounts have shown that the audiences the prior agency loved most — the ones built from the fancy identity spine, the ones that came highly recommended — weren't the top performers. It makes you wonder: Why were those the favorites? Agency incentives matter, and they don't always align with yours.

2. The CPM tax is just too high. Many specialized audiences and match processes add CPM costs to the media in activation. Sometimes those costs are significant. Even if the audience is genuinely better, the incremental CPM can drown out the performance gains. Better targeting that costs 3x the CPM isn't better if it only performs 1.5x as well.

3. Meta knows more than you do. Using first-party data for social targeting usually sounds like a smart idea. But sometimes the platform's own affinity audiences know more about people's interests, preferences, and propensities than your CRM file does. The walled gardens have an extraordinary amount of behavioral signal. Don't underestimate what their algorithms already know. Try out both and see how they work, possibly together.

4. You've over-used and over-exposed it. We've seen examples where first-party and CRM-based audiences were being hit by almost every part of the marketing — email, display, social, search, you name it. At some point, marketing more to this audience isn't creating incremental returns. It's just annoying people you've already converted or saturated.

5. It's too narrow. Your first-party audience or that highly specified third-party segment might sound great in a planning meeting, but it might only cover a small fraction of your total addressable market. If you're over-indexing on precision and under-indexing on scale, you may be leaving enormous growth on the table.

6. Match rates don't work. Challenges with match rates can wreak havoc on an audience strategy.  By the time you've matched your CRM list across platforms and partners, you may have lost a significant portion of the audience you thought you were buying. What's left might not be representative of the original segment at all.

7. Look-alike may not mean act-alike. Brands sometimes don't scrutinize what "lookalike modeling" actually means inside a given platform or data partner. Finding people who look like your best customers demographically or behaviorally is not the same as finding people who will act like them..

The bottom line: audience strategy deserves the same rigor and skepticism you'd apply to any other part of your media investment. Don't fall in love with the pitch. Fall in love with the performance data. Test everything, question the recommendations that happen to benefit the recommender, and never assume that more sophisticated automatically means more effective.