Message testing asks a defined audience to evaluate candidate claims before an agency commits to creative production or media. It can show what people understand, find credible, and consider relevant, with open-ended reasons that help the team revise the copy.
A message test is a research input to the campaign. Its ranking does not establish which ad will convert best under live delivery conditions. Testing early is useful precisely because the team can change a claim before building the whole execution around it.
Why message comes before creative
Separate the claim from the execution when the claim itself is uncertain. A design treatment, celebrity, or animation can draw attention without making the proposition believable. Testing comparable text first lets the team investigate the proposition before interpreting responses to a finished ad.
The practical question is whether the intended audience understands and believes this particular message. Concept testing covers related proposition and claims decisions; creative testing assesses the execution around them.
How a message test actually runs
Define the audience and decision before writing the questionnaire. Choose a manageable set of candidate messages, keep stimulus length and information comparable, and decide whether each respondent sees one option or several. Random assignment helps balance cells in a monadic design; randomized order helps manage order effects when people evaluate several messages. Repeated ratings from one person are not independent observations.
Use consistent questions on clarity, credibility, relevance, and differentiation. Ask what the respondent thinks the claim means and why they gave their rating. Avoid leading wording and keep the response scales the same across options.
The MX8 Labs Research Platform can support programming and reporting production. Review eligibility, logic, exposure order, quotas, and output before fielding. Turnaround depends on questionnaire readiness, audience incidence, usable sample, and review; a same-day result is not a guarantee for every design. The automation workflow describes the production steps.
Different segments, different winners
The overall winner may not be the useful decision. Predefine segments that matter to the campaign and size them for the planned comparison. Report the raw and weighted base, missing responses, and the exact threshold behind each percentage.
Here is an illustrative monadic test of three subscription messages. Each message is shown to a separate group of 100 eligible respondents; all figures are invented.
| Claim | Clear | Credible | Relevant | Base |
|---|---|---|---|---|
| A: More choice every month | 81% | 77% | 65% | 100 |
| B: Cancel whenever you need | 88% | 86% | 74% | 100 |
| C: The lowest price anywhere | 83% | 42% | 79% | 100 |
C attracts interest but raises a credibility question. The next action is to read the reasons, verify whether the claim is supportable, and rewrite or remove it. B is a candidate to carry forward, not a statistically proven sales winner. A small niche segment within these cells may need a larger follow-up sample before its ranking informs targeting.
See sample size and precision for planning meaningful differences. A narrow score gap without enough effective base is a reason to collect more evidence, not force a winner.
Feed the test, don't replace it
Turn the findings into a documented shortlist: message, audience, evidence, open questions, and proposed execution. Retest a material rewrite because the earlier score belongs to the wording respondents actually saw.
The subsequent live A/B test evaluates the selected variants under delivery and landing-page conditions. It can include predefined segment analysis where its design and traffic allow. Survey research supplies a different kind of evidence, including explanations that behavioral logs may not contain; it does not replace observation of conversion.
Find it first
Agencies can use this sequence in a brief, pitch, or client recommendation. The team owns the research decision and interpretation while the platform automates routine production. Use the agency testing workflow to connect the shortlist to the next campaign step, and the creative-testing checklist when the executions are ready.
Request a study quote using the actual audience, survey length, respondent count, and review needs. The pricing page explains those factors.