Wynter B2B message testing
Diagnose and improve B2B page messaging using Peep Laja's Wynter method
Based on Peep Laja's public method.
Not affiliated with or endorsed by Peep Laja. Each step cites the public work it comes from: 7 sources.From poetic promise to a clear offer
Fixes category confusion before revising relevance, value, or differentiation.
The future of workforce agility
Discover the future for your workforce.
Learn moreAutomate payroll corrections for multi-state teams
Replace the manual spreadsheet process for payroll corrections. For HR operations leads at companies with 500 to 2,000 employees.
See payroll automationIllustration from the skill's worked example, not a recorded run.
What it does
When your agent should reach for this skill.
Use this skill to find why a B2B revenue page fails to communicate its offer or persuade its intended buyers, then produce an evidence-linked revision and retest plan. It fits homepage, campaign landing page, product page, and value proposition reviews before launch or after disappointing results. It is especially useful when internal copy debates lack buyer evidence. Message testing explains interpretation and motivation; conversion measurement requires separate behavioral evidence. 16
Start with buyer research if current priorities, buying triggers, alternatives, or decision roles are unknown. Existing recent research can supply this context without another survey. Choose a preference test for two or three viable directions and a pricing page test for price comprehension. Routine message diagnostics should focus on the persuasion path, excluding blogs, careers pages, support material, and late-stage competitor comparison pages. 4
The deliverable is a message diagnosis and revision brief, using the output template. If buyer responses are unavailable, produce a clearly labelled provisional expert review and a research plan. Never describe an untested rewrite as validated, invent respondent comments, or manufacture scores.
The method
11 steps, restated from Peep Laja's public work. Numbers link to the sources.
1Frame the decision
Read method: Scope and evidence. Name the page, intended buyer, next action, and decision the work must inform. Record the message hypothesis and observable success condition before inspecting feedback. Keep the scope to a concrete page or message direction.
2Check buyer context
Use method: Buyer context to distinguish known facts from assumptions about pains, priorities, triggers, buying roles, and alternatives. Reuse credible recent evidence. Where gaps would change the message, plan focused buyer research before writing the promise. 24
3Choose the study
Follow method: Study selection. Match the question to a message, campaign landing page, pricing, preference, or buyer-context study. If research cannot run, preserve the expert-review label and identify exactly what buyer evidence remains missing.
4Freeze the stimulus
Follow method: Stimulus and audience. Record the URL or faithful image, version, hero, important sections, and campaign context. Keep enough of the page experience to detect an explanation that arrives too late. Specify respondent criteria and separate prospects from knowledgeable customers. 56
5Plan the sample and questions
Use method: Questions and sampling for neutral comprehension prompts and ratings with explanations. Start with the playbook's 15-person qualitative default where appropriate. Treat sample counts as planning guidance, never a guarantee of saturation or statistical significance. 46
6Collect and organize evidence
Use the evidence fields to retain respondent IDs, segment, page location, rating scale, and feedback. Cluster underlying questions, preserve meaningful disagreement, and distinguish observed interpretation from the agent's explanation. AI summaries must remain traceable to the actual responses.
7Find the earliest failed layer
Apply method: Ordered diagnosis: clarity, relevance, value, then differentiation. Record the primary failure, evidence, causal hypothesis, and proposed revision. Do not let a favorable downstream score override basic category confusion. 1
8Audit proof and alternatives
Read method: Proof, friction, and sameness. Compare claims with alternatives buyers actually use, including manual work or inaction. Identify unsupported promises and consequential doubts. Distinguish defensible differences from unusual wording. 13
9Rewrite around the diagnosis
Follow method: Revision rules and consult the worked illustrations. Revise the failed buyer question first, connect capabilities to desired outcomes, place proof beside claims, and simplify repeated sections. Preserve strategic tradeoffs instead of implementing every suggestion. 35
10Review the complete brief
Run the pass/fail checklist. Confirm evidence supports the diagnosis, every priority change has a reason, and copy claims are verified or visibly provisional. Correct failed items or record the evidence needed to resolve them.
11Retest and decide
Apply method: Retest and maintenance. Compare buyer interpretations using a comparable audience, stimulus scope, and questions. Recommend ship, revise, or gather context against the declared decision rule. Specify a follow-up trigger; avoid spending on another message test for typos or design-only changes. 4
Worked example
How the method plays out on one case.
1: Fix category confusion before polishing value
Brief and weak first draft
ClearLedger sells payroll correction software to HR operations leads at companies with 500 to 2,000 employees operating across multiple US states. The decision is whether to launch its new homepage. Assume the exercise brief verifies that its product checks corrections against payroll rules and exports approved changes to an existing payroll system.
Hero headline: The future of workforce agility.
Subheading: Unlock seamless operations with one intelligent platform built for tomorrow.
CTA: Discover the future.
Page outline: Vision statement, three repeated sections on regional teams, employee growth, and operational transformation, then a payroll correction explanation near the bottom.
Illustrative evidence and diagnosis
In a fictional 15-prospect test, nine initially describe staffing or scheduling software; six find the payroll explanation only lower on the page. Several like the idea of less manual work but cannot say what the company sells. These overlapping counts are not market estimates.
- Observed within the illustration: Wrong categories and late comprehension.
- Inferred: The abstract hero and delayed product account cause clarity failure.
- Primary layer: Clarity, despite some favorable value reactions. 1
- Secondary issue: Repeated use-case sections delay the explanation. 5
Improved result
Hero headline: Payroll correction software for multi-state HR teams.
Subheading: Check payroll corrections against your rules and send approved changes to your existing payroll system, without managing another correction spreadsheet.
CTA: See a correction workflow.
Page outline: Hero, short correction workflow, supported integration details, one consolidated multi-state use-case section, implementation questions, demo CTA.
Proof block: Show a product walkthrough using fictional demonstration data. Add named payroll integrations only after verification. Omit a time-saving percentage because this exercise contains no measured result.
What changed and why
Naming the category answers the first buyer question. The workflow explanation connects the capability to recognizable work. The spreadsheet comparison supplies a relevant current alternative, but remains a differentiation hypothesis until buyers explain why it matters. Consolidation removes repeated explanation without imposing a universal short-page rule. 135
Retest and decision
Retest the same ICP definition and page scope with neutral paraphrase questions. Check whether prospects identify payroll correction software before being told the category, locate the explanation early, and understand what remains in their existing payroll system. Recommend launch only when the primary misclassification is resolved under the team's declared criterion; otherwise revise. The improved copy is proposed, not validated. 4
1 more worked example in examples/worked-examples.md.
What's in the package
7 files, pinned at d5668a9. Open any file on GitHub.
SKILL.mdEntry point the agent loads: when to use the skill and the procedure.6,262 bytesLICENSEApache-2.0 license.10,756 bytesreferences/method.mdThe method in full, step by step, with citations.15,070 bytesreferences/sources.mdEvery public source the method is built from.2,000 bytesexamples/worked-examples.mdWorked examples showing the method applied end to end.7,006 bytestemplates/output-template.mdThe output format the agent fills in.6,218 byteschecklists/review-checklist.mdChecks the agent runs before handing back the result.5,665 bytes
Sources
The public work this skill is built from. Read Peep Laja's originals.
- Peep Laja, B2B Message Layers frameworkwynter.com
- Peep Laja, How to use buyer intelligence to win the marketwynter.com
- Peep Laja, How to compete and win on messagingwynter.com
- Wynter, The Wynter Research Playbookwynter.com
- Peep Laja, Qualified, public written notes accompanying Do you even resonate?wynter.com
- Kate Meda, The definitive guide to message testingwynter.com
- Wynter, Wynter example testswynter.com
The skill file also covers: Files in this skill. Read the full SKILL.md.
Edge wrote this skill from Peep Laja's public writing and talks listed above. It is not affiliated with or endorsed by Peep Laja. If something misrepresents the method, write to [email protected] and we will correct or remove it.