How to write a case study that sells: the six-part shape

A case study is a short, structured account of one customer's outcome: who they were, what was in the way, what was done, what changed, and how that was measured. On a business website it is the proof unit, the thing that turns "we do this" into "we did this, here are the numbers, here is the person". It sits on the evaluation stage of the journey our journey mapping guide describes, where the buyer has stopped exploring and started deciding, and it is the page a comparison-stage buyer opens before the enquiry form.

It is not a testimonial, which is someone else's opinion, and it is not a project description, which is what you did without what happened. It is closer to a piece of evidence: a claim (the result), the conditions under which it was produced (the situation and the obstacle), the method (what was done), the measurement (the numbers) and the witness (the named person). Written that way it does three jobs at once: it convinces a buyer, it ranks for the problem it describes, and it gives an AI engine a sourced, dated, specific outcome to cite when someone asks who is good at this.

Three rules for the numbers. They come from an instrument you can name: organic sessions from Analytics, impressions and positions from Search Console, rankings from the tracker, leads from the form or the phone log. They cover a stated period with a stated baseline, because "traffic up 300 per cent" from a starting point of ten visits is a different claim from the same figure at ten thousand. And they are shown, not summarised: a before-and-after chart with the figures visible and the source under it, in the house style our data guides use, because a reader who can see the axis believes the bar.

Consent comes before any of it. The client agrees, in writing, to the numbers being published and to how they are described, whether or not they are named. Australian Consumer Law treats a misleading claim about results as misleading conduct regardless of intent, and a case study is a claim about results; the number on the page has to be the number in the instrument, for the period stated, with nothing rounded up. Our social proof guide covers the wider set of trust signals that sit around a case study, and the same rule applies to all of them: true, specific, sourced.

Many businesses cannot name clients: contracts forbid it, the client would rather not, or the house rule (as ours is) says never. A case study still works, because the reader is not buying the client's name, they are buying the specificity. Describe the client by type and place (a boutique cabinetry business in the Sutherland Shire, a Canberra lawn care company), keep every number exact and sourced, name the person who did the work on your side, and let the obstacle carry the credibility. Our own case studies section is written this way: five accounts, no client names, every result measured, and it is the section prospects most often mention on the first call.

Two things an anonymised case study should never do. It should never invent a quote from the client, because a fabricated testimonial is misleading conduct and, for a regulated health client, a breach of the advertising rules our SEO for doctors guide covers. And it should never blur the numbers to hide the client: a result that cannot be measured because measuring it would identify them is a result that should not be published.

The full study lives once, on its own page, with the question it answers in the title ("how a Canberra lawn care business grew organic traffic 14x" is a title a prospect in that situation might search) and the result in the first line. Everything else links back to it. That is the same one-canonical rule our topical map guide applies to every page: one home, many doors. And the client's own version, with the full data, belongs behind the login of the client portal, where it is the most persuasive retention page a business can have.

A case study is a day's work once the numbers exist: an hour with the person who did the work, an hour pulling the before-and-after from the instruments, the chart, the draft, the client's consent and corrections. The expensive part is the discipline of measuring from the start, so the baseline exists when the result does, which is why the data layer comes first in every engagement we run. Our Snowball SEO platform holds the rankings, citations and traffic for every client from day one, so the before-and-after is a query rather than an archaeology, and that is the heavy lifting other agencies bill by the hour. Writing them is part of our articles and written content service, and the measuring is part of the data and analytics service that makes them possible.