Customer case study: the structure that convinces

A B2B buyer makes a good part of their decision without you: they compare suppliers, read their pages, ask colleagues. All the while, one question keeps coming back: “did this work for someone like us?” The customer case study is the document that answers it when nobody from your side is in the room. Built well, it serves the website, sales, the newsletter and trade shows for years; built badly, it reads like an advert and convinces nobody.

This guide describes the structure that convinces, for teams of 20 to 500 people. It extends our guide to marketing automation for SMEs, which places each marketing task alongside the others.

What a case study has to prove

Buyers do not read a case study to learn what you can do: your website already says so. They read it to check three things.

1. That the client resembles them. Same sector, same size, same market, or at least the same problem. A spectacular case at a 10,000-employee group does little to reassure a team of 40.
2. That something changed, and by how much. Not “better efficiency”, but what was done before, what is done now, and what the gap represents.
3. What the project asked of the client. Their team’s time, the decisions to make, what took longer than planned. It is often the question buyers do not dare ask in a meeting.

Google states a similar expectation for any content that wants to be judged helpful: show lived experience, not second-hand knowledge. Among its self-assessment questions is this one: does the content clearly demonstrate first-hand expertise and a depth of knowledge? A case study is exactly that, provided it stays precise.

The seven-block structure

1. 01The client profile.

   Sector, size, market, project period, status (in production, completed, ongoing). Five lines the buyer reads first to decide whether to keep going.

2. 02The starting point.

   The situation before, in the client’s words: what was eating time, what was getting lost, what was worrying. A dated number if possible.

3. 03The decision.

   Why act at that moment, which options the client ruled out, and why they chose you. This block speaks to the buyer weighing several solutions.

4. 04What was done.

   The steps, who did what, how long it took. No jargon: what the client’s team saw change.

5. 05The results.

   Measured, dated, compared with the starting point, with their source (the CRM, the accounts, the analytics tool).

6. 06The limits.

   What remains to be done, what was harder than expected, what the project did not solve.

7. 07The client’s words.

   One or two sentences of theirs, reviewed by them, and what comes next: what they plan now.

This order follows the buyer’s reasoning: “is this my situation?”, then “what was done?”, then “can I believe it?”. It fits on a web page, in a two-page document for sales, and in a ten-minute story during a webinar.

The numbers: measure before so you can tell the story after

Most failed case studies fail on the first day of the project: nobody recorded the starting point. Six months later, the team feels that “things are better”, but nothing says by how much. We turned it into a method rule: the “before” snapshot can never be taken again later. Every month without measurement is lost proof.

Before launch, note with the client three to five numbers they already track: response time to an enquiry, time spent on a task, number of enquiries received, error rate, cost of handling a case. Date them, and say where they come from. The same reading, repeated afterwards, gives a result nobody can dispute. These are also the numbers of automated marketing reporting: what is tracked every week compares effortlessly.

Limits make the rest believable

A story where everything went to plan reads like an advert, and readers know it. A paragraph on what resisted (data messier than expected, a team to win over, a missed deadline) does the opposite: it makes the results that follow believable, and it prepares the buyer for what lies ahead.

Advertising law points the same way. In the United States, the FTC’s guides on endorsements ask that an out-of-the-ordinary result come with what most customers can expect, and that any connection between the quoted client and the supplier be disclosed.

In practice: if the client received a discount in exchange for their testimonial, say so. And if their result is exceptional, present it as such.

The client’s approval, and their review

A case study is written with the client, never about them. Three rules avoid unpleasant surprises:

• Approval is asked for before writing, and spells out what will be published: name, logo, numbers, quote, photo.
• The client reviews everything, numbers included. Their words are their own: you may cut, you do not rewrite their quote.
• Without approval, the case stays anonymous but precise: the sector, the size and the city are often enough to make a story believable. On this website, an assignment whose client has not yet given approval is cited that way, without the name, and no result from it is published before the before-and-after measurements.

The best moment to ask is the end of a successful project, when the client sees the gap for themselves. The interview takes thirty to forty-five minutes; recorded with their approval, it provides the material for the page and its quotes.

Where to put it to work

A case study sleeping on a “References” page does no work. It gets reused:

• Sales send it at the right moment: the case closest to the prospect, right after the first meeting. It is their best argument when they are not in the room.
• The newsletter turns it into an issue, as described in our guide to the B2B newsletter.
• LinkedIn shares an extract, preferably from the leader’s profile.
• A webinar with the client as guest gives them a voice, as in our guide to the B2B webinar.
• A collection of cases by sector can become content offered for an email address, then sent only to contacts in that sector, following the logic of segmentation.

In a new market, a local case outweighs a translated one: a German buyer looks for a German client. Choosing markets is covered in our guide to international SEO, and the place of case studies in the year in the marketing plan for a 50-person company.

AI helps write, the client and a person validate

AI saves time on the mechanical part: transcribing the interview, suggesting an outline along the seven blocks, spotting the client’s strongest sentences, turning the page into a short post or an email. Whatever the model, it does not know what happened: each number is checked against its source, and each quote is reviewed with the client. The method is detailed in our article on content written with AI, and the split between tool and writer in AI and digital marketing: what changes for agencies. The writer remains the one who chooses the angle and holds the story together.

Knowing which case studies sell

A case study is measured by what it triggers: prospects who read it before signing, salespeople who send it, enquiries that mention it. The CRM can tell you, if each contact keeps its source and sending the case is logged there. How to connect these signals to a sale is described in our article on marketing attribution, and the moment a reader becomes a contact for sales in the marketing-to-sales handoff rule. To share a case by email, our method for running an email campaign lists what to check before sending.

Where to start

Pick the client whose project most resembles those you want to sign next year, not the most prestigious one. Ask for their approval, record their starting numbers with them if any remain, and write the seven blocks, limits included. For projects starting now, record the starting point today: they will be next year’s case studies.

To connect case studies, campaigns and CRM through to the sale, our digital marketing offer tracks each contact with its source.

Frequently asked questions

Sources

1. Google Search Central, creating helpful, reliable content (new tab)developers.google.com