Panel Data vs. Controlled Experiments: Choosing the Right Tool for a Pricing or Packaging Decision
On this page
A brand manager deciding whether to change a price, a package, or a claim usually starts with the same question: what do we already know? The answer sits in three different places, and each one answers a different part of the decision.
What syndicated panel data tells you
Numerator's purchase-history dashboards help a team examine what customers bought and how patterns shifted. Those observations supply a baseline. Numerator's full-service research also includes pre-launch price, message, concept, and pack studies. Compare the proposed study rather than assume every panel product only describes the past.
For a pricing or packaging question, ask whether you need existing transaction data, new descriptive research, or a randomized comparison. Recruitment, instrument development, and analysis determine the study's timeline; obtain a scoped estimate.
What's the gap between what happened and what would happen?
Purchase history can show that share dropped after a competitor cut price, but the timing alone does not identify the price effect. A randomized study can separate specified attribute contrasts within its design, using either recruited people or generated responses.
Two errors can leave a decision unsupported: shipping a change from descriptive trends alone, or commissioning a survey whose design never compares the relevant alternatives. Use the existing data to define the question, then choose a study that identifies the required contrast.
Three ways to answer a pricing or packaging question
| Evidence | Question it answers | Respondent or data source | Where it fits |
|---|---|---|---|
| Purchase-history dashboard | What customers bought and how that changed | Observed panel transactions | Establishing the baseline and trend |
| Descriptive survey | What respondents report | Recruited people or generated profiles, disclosed separately | Exploration and measurement of stated responses |
| Randomized choice study | How specified attributes change choices inside the study | Human panel or synthetic population | Testing alternatives; check transfer to actual purchases |
Respondent source and study design are separate decisions. A commissioned human-panel survey can contain a randomized conjoint experiment. Numerator's pricing guide describes choice tasks for comparing products and prices. Inspect assignment, alternatives, estimator, and outcome rather than infer identification from the word 'panel'.
Testing the action, not reconstructing the market
Subconscious runs controlled discrete-choice experiments with generated responses to estimate effects on choices within the simulated study, with uncertainty where supported. There is no published case study for this exact vendor comparison. Review the research evidence and plan a matched human check before treating simulated choice effects as purchase effects.
For a CPG brand comparing package designs or price points, transaction data supplies context. A randomized human or synthetic choice study addresses specified trade-offs. A market test or other relevant validation is needed to assess actual purchase performance.
What does a controlled choice study leave to other sources?
The choice study described here does not itself supply a transaction-history baseline. Obtain that evidence from a suitable data source, such as a panel data provider, and confirm its coverage for your category and period. The controlled study then tests the next decision against that baseline.
Before committing the budget
Ask which of the three questions above is still open. If it's "what has been happening," panel data answers it. If it's "which specific action will change what happens next," that calls for a controlled test of the actions under consideration before they ship. Book time to scope a test for a specific pricing, packaging, or claims decision.
