Rules
Part of Reading marketing collateral with a skeptical eye
How to repair marketing collateral: six passes teams can test
A six pass repair for marketing collateral: check truth, job, sequence, access, action, and control, then test a realistic task and hand off the fix.
What to take away
- Diagnose the reader journey before redesigning the page.
- Test comprehension, action, and delivery with intended users.
- Fix stale evidence and response failures before adding polish.
How to improve marketing collateral begins with watching the current asset in its real setting. Collect the file, brief, sources, versions, distribution points, response data, customer questions, sales notes, support cases, accessibility findings, production defects, unused stock, and staff workarounds.
Run a six-pass repair
| Pass | Question | Repair | Example |
|---|---|---|---|
| 1. Truth | Are claims current and bounded? | Correct, qualify, source, or remove | Example: a price table shows a discontinued tier. |
| 2. Job | Does the asset support one reader decision? | Narrow audience and purpose | Example: a flyer tries to sell and educate at once. |
| 3. Sequence | Can the reader find the answer quickly? | Reorder and label content | Example: eligibility rules sit after the call to action. |
| 4. Access | Can intended users perceive and operate it? | Fix format and alternatives | Example: a PDF has no tagged reading order. |
| 5. Action | Does the next step work end to end? | Repair terms, route, and ownership | Example: the QR code opens a retired form. |
| 6. Control | Can teams update and retire it? | Assign version and lifecycle records | Example: the folder holds three undated versions. |
CDC's Develop and Test Materials guidance addresses public-health information and emphasizes accurate, accessible, actionable material.
Test a realistic task
Section508.gov offers accessible presentation guidance for federal electronic presentations.
- Ask a participant to explain the offer in their own words
- Locate price, limits, evidence, and support without prompting
- Complete the response path on the likely device
- Test a printed proof at actual size and lighting
- Record wrong answers and hesitations, not only preferences
- Retest the corrected version against the same task
Prioritize defects by customer harm, factual risk, blocked action, frequency, and cost to repair. A broken qualification or inaccessible response route outranks decorative inconsistency. Release the smallest correction that resolves the material issue, then measure whether the real journey improved.
Test again with the approved production file, not only the editable draft. Export settings, image substitution, font changes, tags, folds, paper, redirects, and publishing systems can introduce defects after the content review is complete.
Print production specifics matter: paper stock, coating, fold type, trim size, ink coverage, and binding change how the piece handles and how long it lasts. Proof on the final stock and fold before release, and verify bleed, safety margins, and color against the approved sample.
Worked example shape: Asset, a four-page service brochure. Defect, the eligibility rule sits in paragraph three and testers miss it. Repair, move the rule under the headline and add a one-line eligibility check. Measured result, rerun the same task with the same tester pool and report the change in completion rate, time, and blocked steps.
Run a controlled handoff
The GOV.UK technology selection guidance recommends adaptable choices, data control, security review, and ownership-cost analysis.
The CISA software acquisition fact sheet covers development practice, supply-chain exposure, deployment, and vulnerability management.
Assign each step to a named role with a clear finish. Test one broken-data and one revoked-access case, then record the repair, time, and vendor help. Keep the evidence record beside the decision. Set the next review date and an earlier-check trigger. Record rejected options and the chosen path. Use one set of definitions across interface, export, notes, log.
Common questions
Should improvement start with a redesign?
No. Start with truth, reader job, access, action, and version control; design follows those decisions.
How many users should review a draft?
Use enough relevant participants and methods to expose important failures, and state the limits of what was tested.
What should be measured after revision?
Measure corrected comprehension, task completion, qualified progress, service defects, mature outcomes, and full cost.
