How to Switch Brand-Tracking Providers Without Losing Your Trend Line
A provider switch can move a KPI even when the market stands still. Run both trackers together and set an explicit cutover rule for each series.
Run the old and new brand trackers in parallel before you end the old contract. A same-period comparison gives you a chance to distinguish a provider difference from a genuine move in the market. Set the acceptable difference for each KPI before you see either report.
Imagine a regional meal-kit company that has tracked unaided awareness and first choice for three years. A new provider offers faster reporting and broader market coverage. The team uses a three-point change in unaided awareness to open a budget review. If changing providers can create a gap of that size, connecting the old and new points without a break could change the budget decision.
Freeze the part of the tracker you want to carry forward
Export the old questionnaire, exact answer options, display logic, eligible population, quota definitions, field dates, sample source, exclusions, weights, and KPI calculation rules. Include reporting scripts or formulas when available. Have each provider calculate the KPI on a common set of example records to catch a denominator difference before fieldwork.
In the overlap wave, match what you can and document what you cannot. Match wording, question order, brand lists, survey language, mode, market boundaries and field dates. Prevent duplicate participation where suppliers can do so appropriately. If the new provider also proposes a better questionnaire, test that change separately. Otherwise, differences in supplier and instrument arrive together.
Copy this parallel-wave protocol
Item | Decision to record before fielding |
|---|---|
Metric and use | Unaided awareness by market. A change of three percentage points would trigger a budget review in this example. |
Acceptable difference | For each carry-forward KPI, choose a gap small enough to preserve its business use. Keep market-specific tolerances where justified. |
Comparable wave | Same field window, target, questionnaire and reporting rules. Document unavoidable differences in source or execution. |
Sample plan | Plan precision around the acceptable gap and the markets that matter. Specify how weighting and survey design enter uncertainty estimates. |
Readout | Report each provider's estimate and base, the provider difference and its interval, plus a breakdown of composition and exclusions. |
Cutover | Continue, bridge with another wave, or reset the baseline. Assign a decision owner for each metric and market. |
Make the decision series by series
Suppose the example company requires the awareness difference to stay within two percentage points. An estimated difference of one point can still leave a wide range of plausible gaps. If the interval extends past the tolerance, the team lacks the evidence to join that series confidently. If the planned comparison places the plausible differences within the tolerance, the series has evidence for continuity under the conditions tested. Use an equivalence procedure appropriate to the design and choose it up front. Equivalence testing primer
For a stable gap that matters, check whether it persists across another wave and relevant segments before applying any adjustment. Preserve the raw series, label any adjusted view and date the provider change. For an inconsistent gap, start a new baseline and mark the break on the dashboard. A global average can hide a consequential break in a particular market.
Put the acceptance decision in the handover record beside the first live report. Each KPI should have a stated baseline, cutover date and named owner. Sapien's brand-tracking guide
