Most SEO case studies are not read as evidence. They are read as adverts, because that is what most of them are: a traffic graph with no axis labels, a percentage with no starting point, and a client name that cannot be checked. A good case study for SEO works the other way round. It gives a sceptical reader everything they would need to judge the result for themselves, including the parts that make it look less impressive.
That is not just good manners. The rules the US Federal Trade Commission applies to results claims in advertising set out almost exactly what an honest case study should contain. You do not need to be a lawyer to use them as a checklist.
Why most SEO case studies are not believed
The common faults are easy to list. The chart starts at the low point after a migration, so recovery looks like growth. The headline percentage is calculated from a tiny base. The time window is chosen to end at a seasonal peak. The work described is vague, so nobody could repeat it. And the result shown is the agency’s best of the year, presented as if it were routine.
Experienced buyers spot these and discount the whole page. A case study that names its own weaknesses is rarer, and far more persuasive.
The principle the FTC applies to results claims
The FTC staff guidance FTC’s Endorsement Guides: What People Are Asking, dated June 2023, is written about endorsements and testimonials in general. It never mentions SEO or agency case studies, and a case study is not automatically an endorsement. But it is the clearest statement of how a US regulator reads a claim that someone got a specific result, and that is exactly what an SEO case study makes.
The starting point is that “endorsements must be honest and not misleading.” The key passage for results is this: “Endorsements claiming specific results usually will be interpreted to mean that the endorser’s experience reflects what others can also expect. Statements like ‘Results not typical’ or ‘Individual results may vary’ won’t change that interpretation.”
So the guidance gives advertisers two choices. Either “Have adequate proof to back up the claim that the results shown in the ad are typical”, or “Clearly and conspicuously disclose the generally expected performance in the circumstances shown in the ad.” Where an endorser had “exceptional, or even above average, results”, the advertiser without proof of typicality must “make clear to the audience what the generally expected results of following that same regimen are.”
Three more points carry straight over. On age: “If you’re using endorsements that are a few years old, it’s your obligation to make sure the claims still are accurate.” On relationships: an endorser who “has been paid or given something of value” usually needs that connection made clear, which covers a client who got a discount in exchange for being featured. On selection: “Only asking for reviews from customers who you think are more likely to be happy with your product would be misleading if it substantially skews the favorability of the reviews.”
What a trustworthy SEO case study contains
Translate those principles into page elements and you get a checklist. Every claim on the page should come with the evidence a reader needs and the disclosure that stops it misleading.
| Claim on the page | Evidence to show | What to disclose |
|---|---|---|
| Traffic grew | Baseline figure, start and end dates, the source tool | Seasonality, migrations or tracking changes in the window |
| We caused it | The specific work done, in order, with dates | Other changes the client made at the same time |
| This works for clients like you | How results ranged across comparable clients | If this result was unusual, say so and give the typical outcome |
| The client recommends us | A quote the client approved | Any discount, fee or favour given for featuring them |
| It is current | The date the study was written and the period covered | Whether the site or your service has changed since |
The baseline and dates matter most. Show them in the format your reader already trusts, which usually means screenshots or exports from analytics or Search Console, and keep the definitions consistent with your SEO reporting. If the case study claims organic growth, state how organic was measured; our guide to organic traffic in GA4 covers where that number can drift.
Then pick the metric that matches the claim. Sessions are easy to inflate and rarely what a buyer cares about. Leads, revenue or the SEO KPIs agreed at the start of the engagement say more. If you talk about return, show the working, as set out in our post on SEO ROI.
A worked hypothetical
Here is how the same result reads written two ways. The site is invented: call it example.com, a shop selling garden tools, and the figures are illustrative only.
The advert version says: “We grew organic traffic 300% for a garden retailer.” The trustworthy version says: “Between March and September, organic sessions to example.com went from 2,000 to 8,000 a month, measured in GA4. The site had lost rankings after a redesign in February, so part of this is recovery. We fixed redirects, rewrote category pages and built links to the top ten categories. The client also started an email programme in May. Across our other retail clients that year, gains were smaller. The client received a reduced fee for agreeing to be featured.”
The second version is less exciting and much more believable. A buyer can see what was done and judge what might be repeatable. The disclosure about the fee follows the same logic as sponsored post disclosure: say it plainly, near the claim, not in a footer.
B2B and AI SEO case studies
A B2B SEO case study faces a longer path from search to sale, so traffic alone proves little. Show which pages influenced pipeline, how long deals took, and what share of leads were attributed to organic search and how. If the evidence is partial, say which part you could not measure.
An AI SEO case study, whether about AI-assisted content or visibility in AI answers, follows the same rules. No special standard applies because a tool was involved. State the baseline, the dates, the method and the typical result, and be clear about what your measurement could and could not see. If you are tempted to project the result forward, separate it clearly from what happened, as you would in SEO forecasting.
Testimonials that sit alongside the study deserve the same care. Ask all comparable clients, not only the happiest, for the reasons the FTC gives about review selection; our post on Google reviews covers collecting them fairly.
The limitation is important. This is US guidance from FTC staff, and the page itself says it “doesn’t provide a safe harbor from potential liability”; whether a claim is deceptive “will depend on the facts of the specific case.” The Guides were revised in 2023 and may change again. Treat this as a writing standard, and take legal advice if a case study is central to your marketing.
Frequently asked questions
What should a case study for SEO include?
A baseline, start and end dates, the data source, the work done in order, other changes in the same period, how typical the result was, any relationship with the client, and the date the study was written.
Does “results may vary” protect an SEO case study?
Not on the FTC’s reading of results claims. Its guidance says statements like “Results not typical” or “Individual results may vary” won’t change the impression that the result shown is what others can expect.
Should I disclose that a client got a discount for a testimonial?
Yes. The FTC guidance says the connection usually needs to be made clear when an endorser “has been paid or given something of value”.
Can I keep using an old case study?
Date it and check it. The FTC says that for endorsements a few years old, it is your obligation to make sure the claims are still accurate.
The takeaway Write the case study a sceptic would accept: baseline, dates, method, what else changed, what was typical and any relationship with the client. The less you hide, the more the result is worth.
