Claims Testing Research Brief

Example output

Claims testing research brief — Larkspur Dairy probiotic yogurt drinks, 'clinically shown to support digestive comfort'

Why this claim needs perception testing specifically
'Clinically shown' is a high-commitment phrase — it promises a specific evidentiary weight, and shoppers who read label claims closely will react differently to it than to softer language like 'may support.' The research question isn't just 'do shoppers like this claim,' it's whether they find it credible at the specific strength you're proposing, and whether they even parse 'clinically shown' the way you intend. Test appeal and believability as two separate scores — a claim can score high on appeal and low on believability, and that gap is exactly what should stop it from reaching pack.

Perception and believability measures
- Appeal: "How appealing is this claim to you?" (1-5 scale) — captures whether it's a claim shoppers want to see, independent of whether they trust it.
- Believability: "How believable do you find this claim?" (1-5 scale) — the core measure for this test; report separately from appeal, never blended into one composite score.
- Purchase relevance: "How much would this claim influence your decision to buy?" — ties belief back to actual behavior relevance, not just attitude.
- Overreach check: "Does this claim feel like it's promising more than a yogurt drink realistically could?" (yes/no + open-end) — a direct overreach probe, since 'clinically shown' is exactly the kind of phrase that can trigger skepticism if shoppers feel oversold.

Comprehension checks
Before scoring appeal or believability, ask respondents in their own words: "What is this claim telling you the product does?" Code the open-ends for whether "clinically shown" is understood as "a study proved this" versus vaguer interpretations like "doctors recommend this" or "it's healthy." If a meaningful share of respondents overinterpret the claim as a stronger promise than intended, that is itself a finding — even a technically defensible claim can mislead through the reader's inference, which is a different risk than the claim being false.

Head-to-head comparison structure
Test 'clinically shown to support digestive comfort' against at least one softer alternative — for example, 'formulated to support digestive comfort' — using a monadic or split-sample design, identical exposure format for both, differing only in the claim wording. This isolates whether the specific word 'clinically' is doing the appeal work, the credibility work, or actually working against you by inviting skepticism.

Scoring and interpretation plan
Report appeal and believability as separate top-line numbers for each wording tested, cross-tabbed against the comprehension-check codes. A claim wins only if it scores acceptably on both appeal and believability — a high-appeal, low-believability claim is a liability, not a finding to celebrate. Set your pass/fail threshold before fielding, not after seeing the numbers.

Screener criteria
Recruit yogurt or cultured-dairy category buyers specifically, skewed toward shoppers who report reading ingredient or claim language on pack — general grocery shoppers who never read labels will underweight the comprehension risk this claim actually carries in market.

Wording bias flags
Do not precede the belief question with the product's other benefit claims — stacking claims primes generosity toward all of them. Keep the scale balanced (not "how much do you love this claim" framing) and randomize claim order if testing more than one version per respondent.

Important distinction
This research tests how consumers perceive and believe the claim — it does not confirm the claim is scientifically true, and it does not write the final pack wording. Route the actual evidentiary question to whoever owns your substantiation work, and route the winning wording to whoever writes your final pack copy.

Want me to draft the full field-ready questionnaire, formatted for a survey tool?