How to Measure Purchase Confidence in Consumer Research

Most concept and product tests ask whether people would buy. Far fewer ask how sure they are. If you are working out how to measure purchase confidence, the short answer is this: ask it as a separate question, use a clear and fully labelled scale, and place it at the right moment in the respondent journey.

This guide is written for FMCG insight managers and for product and concept testing teams. It covers definitions, question wording, timing in an in-home use test (IHUT), how to read confidence alongside intent, and what to agree with your fieldwork partner before launch.

What purchase confidence is and how it differs from purchase intent

Purchase intent measures the stated likelihood that a respondent will buy a product. Purchase confidence measures how certain the respondent is about that answer. It is a statement about the strength of the decision, not its direction.

The two can move apart. A respondent can say they would definitely buy a new snack after reading a concept, yet feel unsure because they have not tasted it, cannot picture the price, or doubt it will be on the shelf in their usual shop. Another respondent may give a modest intent score but be very certain that the product is not for them.

That is why confidence is useful. Intent scores are known to overstate real behaviour, and confidence gives you a way to separate firm interest from polite interest. It also helps explain why a product that scores well in a concept test can disappoint after use.

Keep the definitions clear in your report. Confidence is not satisfaction, and it is not the same as certainty about product quality. It is certainty about the purchase decision.

How to measure purchase confidence: question wording and scale choices

Good measurement starts with a question that respondents read the same way. A few principles help.

Ask intent first, then confidence

Capture the intent answer, then follow with a confidence question that refers to it directly. For example: "How confident are you in your answer about buying this product?" Asking the two in a fixed order, on separate screens or lines, avoids blending them into one muddled response.

Use plain, single-idea wording

Avoid double-barrelled phrasing such as "How sure and how likely are you...". Avoid jargon. Keep the wording consistent across every product and concept in the study, and across markets, with careful translation checks for multi-country work.

Choose a balanced, labelled scale

Label every point, not just the ends. A typical confidence scale runs from "not at all confident" to "completely confident", with clear middle points. Make sure the steps are balanced and equally spaced in meaning. Decide in advance whether you want a neutral midpoint. An odd scale gives respondents a genuine middle option. An even scale pushes them to lean one way. Either can work, but the choice must match your reading plan.

Keep the scale the same as your benchmarks

If you compare results with earlier studies or norms, do not change the scale or the labels. A small wording change can shift results enough to break the comparison.

Add an open-end where it matters

A short follow-up such as "What would make you more sure?" gives rich diagnostic material. It points to price, availability, trial or claims. Open-end coding turns those answers into themes you can quantify.

Measuring confidence after real product use in an IHUT

An in-home use test gives respondents something a concept board cannot: real experience. That makes it the best setting to see whether confidence holds, grows or falls once the product is in their kitchen or bathroom.

Timing matters. Many teams measure purchase intent at two points: after a concept exposure and after use. The confidence question should sit with each of them, so you can see how experience changes certainty. A sequential monadic design lets you compare products while limiting order effects.

• Before use: capture intent and confidence after the concept or pack is shown. This is your baseline.
• After first use: ask again once the respondent has tried the product as intended.
• After the full use period: for repeat-use categories, measure once more. Confidence often reflects whether the product fits into a routine.

The shift between the baseline and post-use scores is often more informative than either score alone. A product where intent holds but confidence rises has earned trust. A product where intent holds but confidence drops may be living on its concept.

For more on running these studies well, see our IHUT playbook and our product placement and in-home use test services.

Using confidence with intent to read concept and test results

Read the two measures together. A simple grid helps teams make decisions.

Intent	Confidence	What it usually suggests
High	High	Firm interest. A strong candidate to progress.
High	Low	Appealing idea with open doubts. Look at price, claims, trial and availability.
Low	High	A clear rejection. Check whether the product is reaching the right buyers.
Low	Low	Undecided. Often an awareness or understanding gap.

Report top-box and net scores for both questions, then show the cross-break. Many teams create a "confident intenders" group: respondents who say they will buy and are highly sure. That group is a more cautious basis for volume estimates than intent alone, although any forecast also needs the right calibration and category knowledge.

Avoid inventing a conversion factor from confidence unless you have your own validated data linking stated answers to in-market behaviour. If you do not, use confidence as a ranking and diagnostic tool, not as a sales prediction.

Your tabulation plan should include the cross-break from the start. Our data processing and tabulation team can build banner tables that show intent by confidence, split by audience and product.

Sample size, quotas and fieldwork pitfalls

Confidence measures are only as good as the sample behind them. Watch for these issues.

• Cell sizes after the cross-break. Splitting intent by confidence by audience shrinks bases quickly. Plan the sample around the smallest cell you need to read, not only the total.
• Category relevance. Recruit real category buyers or users. People who never buy the category will report low confidence for the wrong reasons. Our respondent recruitment services screen on purchase behaviour, not just demographics.
• Quota balance. Set quotas on the variables that drive the category, such as age, household type, shopping channel and brand usage. Keep them identical across products in a monadic or sequential design.
• Order and fatigue. In a multi-product IHUT, rotate product order. Long questionnaires reduce attention, and confidence ratings tend to flatten when respondents tire.
• Drop-out between waves. If respondents leave between pre-use and post-use stages, the people who remain may be more engaged. Track completion and check whether the final sample still matches your quotas.
• Straightlining and low-quality responses. Scales invite pattern answers. Build in attention checks and review open-ends. Our guide to data quality benchmarks lists what to ask your partner.

How to brief your fieldwork partner for confidence measures

A clear brief prevents most of the problems above. Share the following at the start.

1. The decision the study supports. Go, no-go, optimise or rank? This shapes how confidence is read.
2. The exact question wording and scale, with labels, plus any benchmark you need to match.
3. The measurement points. State whether confidence is asked after the concept, after first use, after the full use period, or all three.
4. The target audience and quotas, including category behaviour and any exclusions such as recent trial of competitor test products.
5. The design. Monadic, sequential monadic or paired, and the rotation plan.
6. Data outputs. Specify the tables, cross-breaks and open-end coding you need.
7. Quality and compliance expectations. Cover consent, data protection and the standards your project must meet.

Ask your partner to flag any wording or sample risks before launch, and to run a soft launch so you can check how respondents read the confidence question. If your studies cover several markets, agree translation and back-translation steps early.

If you are planning an FMCG test and want to talk through the design, our team can help. See our product testing services or get in touch through the site.

Frequently asked questions

Is purchase confidence the same as purchase intent?

No. Intent is how likely someone says they are to buy. Confidence is how sure they are about that answer. Reading them together gives a more careful view of a product's potential.

When should I ask the confidence question in an IHUT?

Ask it wherever you ask intent: after the concept or pack exposure, after first use and, for repeat-use categories, after the full use period. Comparing these points shows how experience changes certainty.

Can I use confidence to forecast sales?

Not on its own. Use it to rank products and diagnose weaknesses. Any forecast needs validated links between stated answers and real behaviour in your category.