For eight consecutive quarters, the numbers said everything was fine: 98% of customers loved the brand.
The numbers came from a tracking programme that cost $8 million a year and surveyed 42,000 people, and it did not waver. Over those same two years, the brand lost market share every quarter.
Nobody had fudged the sample or written a leading question. The tracker was working exactly as designed. It was just measuring love on a 5 point scale, on an annual basis.
We interviewed 118 insights and marketing leaders who buy and run brand trackers to find out if and why their trackers were falling short.
TL;DR, 1) brand tracker data is usually stale because the metrics take too long to collect, and 2) the aggregate metrics lack any real qualitative insight.
Key takeaways
Brand tracking is the repeated measurement of how a defined audience knows, considers and feels about your brand relative to competitors, measured identically each wave so movement reflects the market.
In our study of 118 brand tracking buyers, 81% said their tracker cannot explain why metrics change, and 93% wanted a deeper read on why.
The metrics most trackers lead with (brand love, favorability, consideration ) are trailing indicators. They move last, after the thing that moved them has been building for months.
42% of buyers described their tracker data as stale; 63% said it wasn't actionable.
In short: asking more people isn’t the solution - it’s asking the same people, in the same wave, to explain themselves.
What is brand tracking and what does it measure?
Brand tracking is the practice of measuring the same brand metrics, with the same method, against the same audience definition, at a regular cadence so that changes over time can be attributed to the market rather than to the survey.
Most programmes track some combination of the following. Consistency matters more than the specific list.
Unaided (spontaneous) awareness - "which brands come to mind when you think of [category]?" The hardest metric to move, and the most predictive of consideration.
Aided (prompted) awareness - recognition when shown the name or logo.
Consideration - would this person actively consider buying.
Preference - where you sit against a named competitive set.
Usage, lapsed usage and rejection - current, past, never, and never again.
Brand attributes - the associations attached to you: innovative, good value, for people like me.
Advocacy - NPS or similar, usually tracked alongside rather than as the headline.
Funnel conversion - the ratio between each stage, which is where diagnosis happens. High awareness with weak consideration is a very different problem from low awareness with strong conversion.
A brand tracker also needs a competitive set. A brand health metric with no comparison is uninterpretable: awareness falling three points during a category-wide contraction is a very different story from the same three points while a rival gains five.
For definitions and interpretation guidance on each of these, see our guide to brand health metrics.
Why most brand tracking disappoints
We used Listen to interview 118 people who buy and run brand trackers. Two issues came up repeatedly.
1. The metrics move late and last
Brand love, favorability and consideration are trailing indicators. They are the accumulated feelings from everything customers have experienced over many previous months. By the time brand love drops, whatever caused it has been building for a long time, and the fix is expensive and slow.
Our study results said the same. 42% of buyers called their data stale; a six-month lag from fieldwork to usable insight came up on more than one interview. Two quarters on a question the business needed answered that week. And then, worse still, 63% said the output wasn't actionable, and 51% said getting anything out of it took too much manual work. A quarterly tracker reports on a period that ended weeks before the readout, and then needs a person to spend a fortnight digging up context and turning it into something actionable.
2. The number arrives without a reason
It’s hard to get any granular or deep insights from quantitative 5-point scales. A lot of the time, it stays artificially high and captures a feeling closer to general positive regard than to any form of love that a marketer cares about. It just isn’t the language people use for brand love.
And then if the metrics start to move, 81% of the buyers we studied said their tracker can't explain why. As one put it: "It's good to know it went up or down, but you don't know why."
The usual remedy is to commission a follow-up qual study on what the root cause could be - the price change, the competitor’s campaign, the packaging refresh, the category. That takes weeks, costs even more, and arrives long after the decision it was meant to inform.
Solving stale data: Continuous vs periodic brand tracking
The alternative to large, stale quarterly waves is smaller, more frequent ones- roughly the same annual sample, distributed differently.
Periodic tracking gives you a large sample per wave, clean wave-on-wave comparison, and a reporting rhythm the organisation already understands. It also builds in the lag, and makes any single wave hostage to whatever was in the news that fortnight.
But continuous tracking gives you a rolling read: campaign effects can be measured against a live baseline rather than a stale one. The trade-off is smaller bases per wave, which means being disciplined about what you read weekly versus what you only interpret on a rolling basis.
Note: continuous is not automatically better. A stable category on an annual planning cycle may be fine with quarterly waves. But it is definitely the superior method when the category moves fast, when you spend meaningfully on campaigns you want to evaluate, or when your organisation has repeatedly been surprised by a shift it should have seen.
Cadence is one of six decisions that determine whether a tracker survives its second year; the rest are covered in our guide to building a brand tracker.
Adding back the why: conversational brand tracking
The more consequential change is question type.
Open-ends are already known to be the most useful data in the tracker - they’re just also the most time-consuming and expensive to process. Traditional trackers minimize open-ended questions because somebody has to read and parse the answers. A thousand free-text responses per wave is a coding project, so the open-ends get cut back to one or two, asked generically, and analysed thinly. AI moderation and analysis removes that constraint, and makes it possible to get qualitative insights in days.
More significantly, conversational brand tracking adds more qualitative context and depth.
Firstly, the follow-up is conditional. A respondent who says they love a brand gets asked why and then gets probed on the evidence in their actual answer rather than a generic prompt. The score and the explanation come from the same person in the same session, so the why attaches to the what at the individual level rather than being inferred across aggregates.
Second, the coding happens after collection rather than before. Nothing is predefined, so the themes that matter surface on their own instead of being limited to the answer options someone from the brand thought to write. That's how you find the brand love, or the concern, that you didn't know to ask about, and it's why open-ended themes tend to move before the structured KPIs do.
This also makes it possible to accurately track things that resist a scale. “Which of these brands do you love”, "to what degree does this brand represent your cultural values" or "how much do you trust this institution" produce noisy 1-to-5 data, because every respondent reads the scale differently. Asked as a question people answer in their own words, then coded afterward, they become trendable.
*See this post for more details on our study on 615 customers and how the research method of 5-point scale inflated brand love by over 5x.
Where Listen Pulse fits in
Pulse is a new kind of conversational tracker that blends quant and qual: it measures the KPIs you already report, how they've changed, and the reasons why, in the same wave.
You can run the same study with the same screeners wave after wave, so the results line up. You can also add timely questions any time, with no new study needed.
With Pulse, the themes that will move next quarter's KPIs show up in conversations now, Pulse catches emerging themes early, so clients have time to act while it still matters. When KPIs do shift, Pulse reports those changes and the reasons why.
Other features of Listen Pulse:
Longitudinal by design - the same core study, audience spec and screening wave after wave, at whatever cadence fits your category, with a rotating flex section so you can add timely questions about a campaign or a competitor without breaking the trend line.
Mixed methods in one study - awareness scales, NPS, multiple choice, Likert, MaxDiff and ranking questions sit alongside AI-moderated open-ends.
Quantified qualitative - open-ended answers are coded into themes, sentiment and drivers after collection, so emergent concerns like "price concern" or "trust barrier" become metrics you can trend.
Traceable - every metric traces back to the moment in the interview behind it, with the participant's own words and the clip attached.
Layered onto what you run today - Pulse connects to Qualtrics and Decipher, routing the same respondents from your existing survey into a Listen conversation and matching them by ID. You keep your tracker, your KPIs and your history, and add the depth. Existing tracker data can be imported so Listen becomes the analysis layer.
Recruiting built for quality over volume —-30M+ verified respondents across 45+ countries and 100+ languages, with quality scoring that screens out fraud and low-effort responses, plus human review. Fewer, better conversations rather than thousands of thin ones.
Conclusion
Trackers used to primarily be quantitative because reading and coding open-ended answers at scale was prohibitively expensive. That constraint is gone - you can now, within a reasonable timeframe and cost, get both continuous live metrics and in depth qualitative reasoning that you can actually do something with.
Somewhere, there's a team looking at 98% brand love right now and feeling reassured. Someone needs to suggest they Listen a little closer.