Before freestyle, I had to learn to blow bubbles. Then kicks. Then arm strokes. Each one felt almost too basic to bother with on its own. But every layer only worked because the layer under it was solid. Nobody skips straight to butterfly.

Here’s the part I didn’t expect: the fundamentals don’t go away. They just go quiet. You stop noticing them right around the time they start doing the most work.

Brand health measurement runs on the same logic, and I think a lot of research teams are getting the lesson wrong.

A senior insights leader told me recently that brand trackers are done. Obsolete. A relic of the deck-and-fieldwork era, replaced by always-on AI listening. I’ve been chewing on that ever since, because I think the opposite is true. Trackers matter more than ever. We’ve just been asking them to do the wrong job.

The argument that “trackers are dead” isn’t about brand tracking as a concept.

It’s targeting one specific, tired version of it: the slow, expensive, 50-slide deck that lands weeks after fieldwork closes, vendor costs that only go up, and findings that die somewhere between the insights team’s Slack channel and an actual decision.

Search “brand tracking” and the headlines write themselves—Is brand tracking irrelevant? How to revive your dying tracker. Read past the headline, though, and none of them actually argue that measuring brand health over time has stopped mattering. Every complaint is about the method. Not the underlying need.

Meanwhile, AI listening and CX signals have genuinely expanded what’s possible—always-on data at a scale no survey panel could touch. That’s a real conversation worth having about modern research infrastructure. But speed without calibration is just confident confusion, dressed up in a dashboard. More data doesn’t automatically mean more understanding.

We’re diagnosing the wrong problem. It’s not that brand trackers exist. It’s that most organizations still treat them as a standalone line item competing for budget against listening tools, instead of one layer in a system that was never designed on purpose.

The future of brand health measurement isn’t a sleeker tracker or a smarter listening platform. It’s a measurement architecture where every instrument has one precise job and does it so well that the others become more valuable.

Treat these as competing line items, and you’ll always be defending the budget. Treat them as a system, and each one starts making the others more valuable. It’s the same as learning to swim: the layer underneath has to be solid, or nothing built on top holds.

AI listening has a structural weak spot that doesn’t get talked about enough: the data isn’t stable.

Platform algorithms shift. Conversation volume spikes and dips for reasons that have nothing to do with your brand. You can see what people are saying right now but comparing that to six months ago is shakier than it looks, because the platform, the participants, and the conversation dynamics are all moving independently of brand performance.

That instability matters the most when the stakes are highest. Attitudes toward AI are a live example—trust in AI is shifting fast and unevenly across markets right now.

If you’re a brand operating in this climate, the hard part isn’t spotting the conversation. It’s knowing whether a shift in sentiment is a real, durable move in brand equity or noise that’ll settle back down next quarter. Without controlled measurement, you genuinely can’t tell the difference.

A single data point tells you almost nothing on its own. A trend tells you whether you’re actually moving, and in which direction. One-off research captures a moment. Tracking reveals momentum.

This was never “trackers versus AI.” It’s trackers doing the one job only they can do, while everything else does what it does best.

Back to the pool.

When I finally got freestyle down I didn’t stop blowing bubbles, I just stopped thinking about it. Breath control went automatic, the invisible foundation everything else got built on. Skills compound. Each layer makes the next one possible, and eventually, you forget you ever had to learn it separately.

That’s the model for brand measurement, too. AI flags the anomaly. The tracker sizes it. The diagnostic layer explains it. And increasingly, an LLM sits across all three, surfacing connections across data sources that would take a human weeks to spot manually.

When the underlying architecture is solid, that synthesis is genuinely useful—anyone in the organization can ask a plain-language question and get a coherent, defensible answer back. When the architecture isn’t solid, you get confident-sounding noise wearing a research team’s credibility as a costume.

I’ve seen this pattern emerge repeatedly across B2B organizations: they start with a well-built tracker, layer in AI listening, and commission focused diagnostic research whenever the tracker raises a flag. Each layer compounds the value of the one before it.

The insights leader who told me trackers are dead wasn’t entirely wrong. The version they meant (slow, siloed, generating decks nobody opens) deserves to get killed off.

But the fix isn’t abandoning controlled measurement. It’s giving it a smaller, sharper job: the validation anchor inside a system that’s gotten faster, noisier, and more complex than it’s ever been.

The future isn’t trackers versus AI. It’s a measurement system built on purpose, where each instrument does exactly one thing well—detect, measure, explain, and increasingly, synthesize.

In that system, the brand tracker isn’t obsolete. It finally has the job it was always meant to do.