Switching brand tracker: what the time series is worth

Many companies keep their brand tracker because of the time series, the years of comparable data that a switch would break. Our comparisons suggest that the part a switch breaks, the absolute level of each KPI, was never very portable. The parts most decisions rely on, direction over time and rank against competitors, can be carried across.

Breaking a series has real costs, so the argument deserves an answer based on data.

Why companies keep a brand tracker for its time series

A tracker that has run for five or ten years carries a lot of weight inside a company:

Targets are set as levels, such as reaching 60% awareness by the end of the year.
Board decks show the same line chart every quarter.
Bonuses are sometimes tied to it.

A new supplier will almost certainly report different levels, and a step in the chart is hard to explain to people who were not part of the decision.

Those concerns are fair. They rest on the assumption that the old levels described the market in a way a new tracker cannot reproduce, and that assumption can be tested.

Three brand trackers measuring the same brand

Earlier this year we compared three independent trackers that all measure the same well-known confectionery brand in Germany, from 2022 to early 2026. One of them is ours. The other two are run by different agencies, with their own samples and questionnaires.

On aided awareness, the three reported 79 to 84% (ours), 74 to 84% and 88 to 94% over the period. At one point in time, two of them were 14 points apart.

None of the three is wrong. As the table shows, they define the category differently, screen respondents differently and word the purchase question differently, and each of those choices moves the level.

The direction was the same in all three. Awareness fell by 4 to 7 points between 2022 and 2026, and consideration and purchase fell too. The purchase figures moved together across every pair of trackers, with correlations of 0.59 to 0.72.

Awareness was less consistent. Two of the trackers moved closely together, with a correlation of 0.84, while the third, measured once a year on a broader category, followed the other two only weakly from quarter to quarter.

A brand manager reading any one of the three would have drawn the same conclusion about where the brand was heading, while quoting a different awareness figure depending on which one they read.

The comparison also showed a drop among younger consumers in every source. Consideration fell by 9 points in one tracker's 18 to 29 group, and by 16 and 17 points in our 16 to 24 and 25 to 34 groups.

A drop that appears in independent trackers with different designs is much harder to dismiss as noise between waves than a move in one of them. That is a practical reason to run two sources side by side for a period during a switch, because a movement both sources show can be trusted more than either one alone.

A shorter brand list changes levels inside one tracker

Changing supplier is not the only way to break the levels. In one B2B tracker we run, the list of brands shown to respondents was cut from around 25 to 11 between two years. Aided awareness rose for every brand that stayed on the list:

by 12 to 17% in relative terms for the two best-known brands
by 40% on average for the rest

Read at face value, the client's own awareness went from 39% to 57% in a year. After calibrating against the brands measured in both years, the part that reflects a change in the market was close to 3 points, and the rest came from the shorter list.

We reported both series side by side and marked the year as a method break. The raw history was left as it was, because it records the market as it was measured at the time.

What switching brand tracker breaks, and what it keeps

Levels break. They depend on who is asked, how respondents are screened, how questions are worded and how the funnel is routed, so a new design will move them.
Direction and rank carry over, as long as the competitor set and the target group stay close to the old ones.
Levels can be bridged with a parallel run of both trackers or a calibration on brands measured in both, with the break marked in every report.

This is also why rank is a more robust way to communicate position than a level. Two trackers can disagree by ten points on a brand's awareness and still agree on where it sits among its competitors.

When keeping the existing tracker is the better choice

There are cases where I would keep the existing tracker:

A decision is formally tied to an absolute level, such as a target written into an agreement or a figure reported externally year after year. A switch then has a cost that calibration only partly covers.
The new design cannot reach the same target group, so the comparison breaks for reasons that have nothing to do with levels.
The main motive is price. The work of bridging two series then belongs in the calculation.

What to do before switching brand tracker

List the decisions the tracker feeds, and which numbers each one uses. Decisions that rest on direction or rank will survive a switch, while targets set as absolute levels will need to be reset.
Compare the old and new designs line by line before reading any number: who is asked, how they are screened, question wording, funnel routing, how often you measure, sample size and the time window for purchase. In the German comparison, these choices explained almost all of the differences.
Bridge the levels where the budget allows. Run both trackers in parallel for one or two waves, or calibrate on brands measured in both, and report the raw and the calibrated series side by side.
Mark the method break in every report and leave the raw history as it is.
Move targets from levels to rank and to the gap to named competitors where you can, because those figures stay comparable across trackers.

In my view, a time series earns its place through the decisions it supports, and most of those rely on direction and relative position. Both can be carried across a switch with some care.

In our comparison, absolute levels were the least comparable part of the data, so on their own they are a weak reason to keep a tracker.

The time series is a weak reason to keep a brand tracker