Content strategy · 3 October 2026
You have useful work to show, but no verified revenue uplift—or permission to name the client. You can still publish a credible case study. Show what was done, attach evidence to the claim, and stop before turning an unmeasured outcome into a success story.
By TrendTransformers · Sources checked 3 October 2026. The worked example below is fictional, not a client result.
Repair the claim before designing the page
Imagine a consultancy that has reorganised an equipment supplier’s enquiry page. It can document the old and new page, the reasons for the changes and a successful submission test. It has not compared qualified enquiries before and after launch.
Unsupported version · fictional example
“Our redesign transformed lead generation and increased conversions.”
A completed redesign does not prove either outcome. Even an attractive screenshot cannot establish that more suitable customers contacted the business.
Defensible version · same fictional project
“We separated repair and replacement enquiries into two clearly labelled routes, added the information needed for each, and tested both submission paths after launch. The effect on qualified enquiries has not yet been measured.”
The second version gives a buyer something to judge: a recognisable problem, a specific decision and a defined verification boundary. To make it a useful project page, show an approved before-and-after crop with annotations explaining the two routes. Do not fill the missing results section with an estimated percentage.
More content is not the same as more evidence
In its 30 September 2026 content survey, Ahrefs reports 301 respondents, of whom 280 used AI for content creation. Among AI users, 85% said production had become faster. These are self-reported survey findings, not a representative census of Belgian firms or evidence that publishing faster improves rankings.
The practical inference is narrower: generating another explanation is increasingly easy. Documenting a decision you actually made adds something a generic explanation cannot. A case study should make that distinction visible.
Google’s people-first content guidance asks about original information, clear sourcing and first-hand experience. It does not promise a ranking reward for putting “case study” in a title. Treat the format as a way to substantiate expertise, not as an SEO shortcut.
Give each important claim its own evidence record
Before writing the story, complete a short record for its strongest claim. This is an editorial working method, not a certification or an official search requirement.
- Claim
- What exactly do you want the reader to believe? “Both enquiry routes submitted successfully” is narrower than “the website converts better.”
- Evidence
- Which dated artefact supports it: an approved screenshot, test record, interview note or comparable report? A screenshot of a chart without its definitions is incomplete evidence.
- Scope
- What was included and excluded? For a functional check, name the route, device or browser tested. For a business metric, state the period, metric definition and relevant denominator.
- Other explanations
- What else changed? A new campaign, seasonal demand or a price change can affect enquiries alongside a redesign. A before-and-after observation alone does not isolate the cause.
- Publication boundary
- What may be shown, who has approved it, and what remains unknown? Keep the supporting record even when confidential details cannot appear publicly.
A useful historical example is NN/g’s 2021 account of Marketade’s work for Baileigh Industrial. The team tested navigation before and after changing the information architecture, reusing the tasks with minor wording adjustments and recruiting a fresh participant group. The lesson is the connection between the intervention and a defined test—not a percentage to borrow for another company’s sales claim.
If the available evidence supports delivery but not impact, publish a delivery account. If it supports an observed improvement, describe that observation and its limitations. Reserve causal language such as “produced the increase” for evidence that can genuinely support attribution.
A hidden client name is not a licence to invent a client
An anonymised account still describes real work. Keep the problem, constraints, decisions and findings accurate, and say that identifying details are withheld. Confirm what can be published before using screenshots, quotations or business data. Removing a logo alone may not conceal identity; filenames, distinctive interfaces and unusually specific facts can reveal it.
A concept is different. A self-initiated design can demonstrate an approach, interaction or visual direction, but it cannot demonstrate a client relationship or a commercial result. Label it as a concept next to the preview—not only in a distant disclaimer. If there was no live deployment, say so. If a screenshot has been recreated or simplified, identify that too.
When confidentiality leaves too little evidence to explain the decision, a general method article may serve the reader better than an apparently authoritative but unverifiable success story. Do not turn several clients into one fictional “anonymous client”; label a composite as illustrative.
Questions before you publish a case study
Can I publish a case study without revenue figures?
Yes. Explain the problem, the work you actually completed and the evidence available. State that commercial impact has not been measured rather than implying a sales or conversion increase.
Can an anonymous case study still be useful?
Yes, if you can share enough approved context and evidence to explain the work. Keep the underlying facts intact, disclose that identifying details are withheld, and do not describe a fictional composite as a real client.
Does publishing case studies guarantee rankings or AI citations?
No. A case study can provide distinctive evidence and help buyers evaluate your service, but the format alone does not guarantee search rankings, citations or enquiries.
Make the proof page support the service page
A service page explains the offer; a case study shows how a particular problem was handled. Give each a different job. Instead of repeating the main service keyword across several near-identical stories, title the project around the actual problem and decision—for example, separating repair from replacement enquiries.
Link from the case study to the relevant service using a descriptive phrase, and place a contextual link back to the proof from that service page when appropriate. Keep the essential explanation in readable page text, not only in screenshots. Clear business, service and project names make the account easier to interpret without requiring invented claims or special “AI-friendly” language.
This is where SEO and GEO content strategy and AI-assisted website creation meet: the design should help a prospective customer inspect the evidence and recognise whether the service fits.
Start with one project, not a portfolio quota. Choose the claim you can prove today, show the decision behind the work, and name the limit. A modest, inspectable account is more useful than a spectacular result that nobody can verify.