Most global brands do not run one brand tracker. They run four or five, commissioned from different agencies in different regions, each with its own scales, bases and calendar. Global brand tracking then turns into a debate about whose number is right. This guide shows how to harmonize brand trackers into one view without re-fielding them. It applies data harmonization, making data from different sources comparable, to brand trackers.
Key takeaways
- Global brand tracking often rests on several trackers: regional agency choices, acquisitions and method changes each add one.
- Before you combine brand trackers, align five things: brand lists, KPI definitions, scales, sample bases and wave timing.
- Convert every tracker to a common net, such as top-2-box consideration among all buyers.
- When a method or agency changes, run a parallel bridge wave and calculate an adjustment factor before you join the trend lines.
- Flag markets that cannot be compared instead of forcing them into a global total.
Why do global brands end up with several brand trackers?
It rarely starts as a plan. A regional team commissions its own tracker because the global one skips its market. An acquisition brings a legacy study with years of history. A move from phone to online interviewing, or a new sample source, breaks a single tracker in two.
The cost shows up in the global review. Each region presents its own numbers and footnotes, and the meeting debates whether a 45 in one market beats a 7.2 in another. In the programs we see, nobody is wrong about their own market. The numbers were never built to sit side by side, so there is no global brand tracking view at all. If you need the basics of the method first, see what brand tracking is.
What has to be aligned before you can combine brand trackers?
Five elements decide whether two trackers measure the same thing.
| Element | What typically differs | What to agree |
|---|---|---|
| Brand lists and naming | Sub-brands listed separately in one market, merged in another; local spellings | One master brand code list; local brands mapped to it |
| KPI definitions | Unaided awareness as first mention or all mentions; consideration asked of everyone or only of those aware | A KPI dictionary with wording, base and net for each metric |
| Scales | 5-point, 7-point or 0 to 10 for consideration and preference; NPS on 0 to 10 in one study, 1 to 10 in another | A common net, such as top-2-box, with a documented conversion rule |
| Sample frames and weighting | All adults vs category buyers; different weighting targets | One reporting base, or a flag where bases cannot be matched |
| Wave timing | Monthly continuous fieldwork vs quarterly bursts | A common reporting period, such as calendar quarter |
NPS deserves a note. The standard version uses 0 to 10, with promoters at 9 to 10 and detractors at 0 to 6. A 1 to 10 version has no clean conversion, so treat it as a separate measure.
How do you harmonize brand trackers without re-fielding?
You work from the brand tracking data you already have, ideally respondent-level files from each agency rather than summary tables.
- Inventory every tracker. Record market, agency, method, sample source, base, scales, weighting targets, wave dates and questionnaire version.
- Build a KPI dictionary. Define each metric once: question wording, answer scale, base and net.
- Map questions and codes. Link each tracker’s question IDs and brand codes to the dictionary and the master brand list. In mTab, you can import tracker and survey data into one governed layer, where it is harmonized so reports answer from all of it.
- Rescale to a common basis. Recompute nets on a shared definition and re-base metrics to one population.
- Bridge method changes. Where an agency or method changed, use overlap waves to estimate the size of the break.
- Flag what is not comparable. If a judgment call moves a result materially, report the range and mark the market.
What does a worked example look like?
Here is an illustrative case. Two agencies measure consideration for the same brand in three markets, all with 500 respondents.
- Market A, Agency 1: 5-point scale, asked of all category buyers. Top-2-box (“definitely” or “probably would consider”) is 42%.
- Market B, Agency 1: same 5-point scale, but asked only of people aware of the brand. Top-2-box among the aware is 55%, and aided awareness is 80%.
- Market C, Agency 2: 0 to 10 scale, asked of all category buyers. Scores of 10, 9 and 8 are 9%, 8% and 13%; a score of 7 is 15%.
Market B needs re-basing. People who are unaware of a brand cannot consider it, so consideration among all buyers is 0.55 × 0.80 = 0.44, or 44%.
Market C needs rescaling. Mapped onto 0 to 10, the 5-point scale’s points fall at 0, 2.5, 5, 7.5 and 10, so top-2-box covers 7.5 and above. Scores of 8 to 10 qualify: 9 + 8 + 13 = 30%. A 7 converts to 1 + 0.4 × 7 = 3.8 on the 5-point scale, which rounds to 4 but sits below the cut-off. Including it gives 30 + 15 = 45%.
| Market | Agency | Raw result | Conversion | Harmonized consideration (all buyers) | Flag |
|---|---|---|---|---|---|
| A | 1 | 42% top-2-box | None | 42% | Comparable |
| B | 1 | 55% of aware; 80% aware | 0.55 × 0.80 | 44% | Comparable (re-based) |
| C | 2 | 30% at 8 to 10; 45% at 7 to 10 | Rescaled, boundary unclear | 30% to 45% | Not directly comparable |
A and B can sit side by side. Market C cannot: depending on one boundary decision, it is well below or slightly above the other two. Report its trend against its own history, and flag its level as not directly comparable until a bridge wave settles the conversion.
How do you handle a vendor switch with a bridge wave?
When a market moves to a new agency, run the old and new trackers in parallel for at least one wave, with the same questionnaire and base. The gap estimates the method effect.
Say Market A runs a bridge wave with 500 respondents per study. The old agency reads consideration at 42%; the new one reads 34%. First check that the gap is real. The 95% threshold for a difference between two independent samples is 1.96 × √(0.42 × 0.58 / 500 + 0.34 × 0.66 / 500) = 1.96 × 0.0306, about 6.0 points. The 8-point gap clears it.
The ratio adjustment factor is 42 / 34 = 1.235. If the new agency reads 36% next wave, the series-consistent figure is 36 × 1.235 = 44.5%. An additive adjustment, 36 + 8 = 44%, gives a similar answer. Pick one rule and document it. If the gap is inside the threshold, do not adjust; mark the agency change on the chart.
How do you report global brand tracking by market and region?
Show harmonized market results first, each with its flag. Build regional and global totals only from comparable markets, and state the weighting basis; global brand tracking across many markets covers market-size versus equal weighting in detail. Flagged markets appear beside the total, not inside it.
The same view often needs other sources by type, such as retail scan and consumer panel data, covered in our guide to syndicated data, or CRM and CX survey data, the subject of customer data integration. The aim is one agreed version of each number that every region reads from. Crosstabs and banner tables with weighting and significance testing make the market-by-region cut practical on a wide tracker.
What are the common mistakes?
- Comparing nets with different bases. Consideration among the aware almost always looks higher than among all buyers.
- Hiding the conversion rule. If nobody can see how a 0 to 10 score became top-2-box, every review challenges the global number.
- Joining trend lines across a vendor switch without a bridge. The break looks like a brand change.
- Forcing every market into the total. One non-comparable market can move a regional figure more than any real change.
- Mapping brands by name alone. Sub-brands and local names need explicit codes in the brand tracking data, or share metrics double count.
What should you do next?
Start with the inventory: one row per tracker, with agency, method, base, scales and wave dates. That table shows where your brand trackers already line up for global brand tracking and where a bridge wave or a flag is needed. Once the harmonized view exists, you can track brand health and consideration continuously and see what changed and why across every market.