B2B case study: how to write the one a buyer forwards to a colleague

A B2B case study that gets forwarded to a colleague has three properties, and none of them is stylistic. It describes a recognizable situation before it describes your solution. It carries at least one figure the client agreed to make public. And it lets the client speak, with a name, a title and a company attached. Structure, length and vocabulary have never stopped anyone from forwarding a case study. What stops them is that nothing in it can be verified.

So the hard part is not the writing. It is the getting. A case study is won at the authorization stage, not at the drafting stage, and that is where most of them die quietly, weeks after an interview you already paid for.

What a technical reader checks first

A plant manager who receives your case study does not read it as sales copy. He reads it like a spec sheet, and he looks for four things in the same order every time.

First, does the company named look like his. A 40-person fabricator south of Québec does not see himself in a project run at a 4,000-person multinational, even when the problem is theoretically identical. Second, is the constraint his own. A slipping delivery date, a line that stops, a bid lost to a cheaper supplier: those are situations, not themes. A piece that opens with your values has already lost that reader.

Third, is anyone named accountable for it. A case study with no name, no title and no company reads as fiction, however well it is written. Fourth, is the number bounded. A figure with no period, no comparison base and no mention of what else changed over the same months does not count. A technical buyer knows this, and treats a bare percentage as decoration.

There is a fifth thing he checks without admitting it: whether the project ever hit an obstacle. No mandate runs clean from start to finish, and a piece that says otherwise reads as an ad. The cost to you is simple. Your case study sits on your site, nobody forwards it, and you wrongly conclude that the format does not work in your industry.

The figures a client will approve, and the ones they never will

Most case studies die because they were built around a number the client can never sign off on. The line is clearer than people expect, and it is not arbitrary: a client refuses any figure that helps a competitor, their own procurement group, or whoever negotiates with them next.

Almost always refused: the price paid, the margin, production volumes, the name of their end customer, real contractual lead times, and anything that lets an outsider reconstruct their cost structure. This is not obstruction. A supplier who publishes what one client paid loses the next one, and the client who authorized it gets the number quoted back at them at the next renewal.

Far more likely to clear: a relative change rather than a dollar amount, an internal indicator nobody outside can cross-check, a duration, a scope (how many sites, how many lines, how many employees affected), and an operational fact anyone could verify on site. Which is why the first question is not what to write, but which number survives. Ask your client that before you mobilize anyone, and take the answer seriously. If nobody on their side can name a publishable figure, the case study does not exist, and it is better to learn that on day one than after the shoot.

The client quote carries everything else

All the text around it is you talking about you. The quote is the one part you cannot manufacture, and that is exactly why it holds up the credibility of the whole page.

A quote that says "great service, professional team" sits on every competitor's website too. It proves nothing, because it could have been written without the client.

A quote that travels names what was not working before, or what worried the client at the moment of signing. It is the only kind a buyer believes, because a buyer comparing two suppliers knows no project runs without friction. A quote that admits hesitation is worth more than a paragraph of praise, and it is usually the line an uncomfortable supplier cuts in the edit. Watch him do it: that is a signal about him, not about the client.

The rule that decides everything: never rewrite the client's words. A reworded quote gets bounced by their communications team, or worse, gets approved and reads as fake. That is also the line I hold in my own work. AI can produce the text around the quote. It cannot make a named person say something true about you with their title and their company printed beside the sentence. The proof is not in the words. It is in the fact that a real person said them.

The approval path, where most case studies die

In a structured company, the person who speaks is not the person who authorizes. The chain almost always looks the same: your project contact agrees, communications approves the principle, legal reviews the final text, and sometimes a parent company abroad has the last word. Every step adds business days, and every step removes something.

The classic trap is asking for authorization at the end. At that point you are asking three people to approve a text they did not commission, about a project they did not follow, built on a number they did not choose. The reflex will be to cut. What you need before you write a single line: who signs, what is excluded outright, and how long their usual approval cycle runs. Three answers, on the first call.

The decision belongs to the client, never to you. Your job is to hand them something easy to say yes to: a short text, a fallback version with no figures prepared in advance, and no gap between what they agreed to verbally and what they end up reading. On the project video series produced with Mecart, the work spanned 14 project videos across 10 states and provinces, and every project arrived with its own authorization perimeter. That is the part of the job that never shows in the finished piece.

When you will never get the authorization, and what to do instead

Say it plainly: in a share of cases the answer is no, and it is no for good reasons. A client bound by a confidentiality agreement cannot authorize you. A buyer who treats working with you as their own competitive advantage has no interest in announcing it to their rivals. An aerospace machining subcontractor, a supplier of controlled environments, an integrator working on sensitive processes: for them the refusal is the norm, not the exception.

Do not work around it. Half-anonymizing, along the lines of "a manufacturer in the Québec City region", saves nothing. A technical reader treats an anonymous case study exactly as he would treat an invented testimonial. You lose the credibility without gaining the permission.

Three things work instead. The project without the client: show the process, the installation, the physical result and the specifications, naming nobody, because the proof then sits in what can be seen. The private reference: a client who refuses publication will very often take a call with a serious prospect, which makes no web page but does close deals. And your own data, the part you can prove about your own operation without anyone's permission. Accept publishing few, too. Two real approved case studies beat eight vague ones, and your site repays it. It is the same logic that makes a manufacturer website credible to a technical buyer.

Where this goes next

When the client agrees to speak on camera, the constraints change and so does the preparation. That is the subject of the page on the B2B client testimonial video. The full logic, from the proof itself through to putting it in front of decision makers, is set out on the B2B marketing page. On our side, the person who shoots is also the person who runs the ad accounts, so a case study is designed from day one around what it will have to do once it is live.