How Many Case Studies Does a B2B Site Actually Need?

How many case studies does a B2B site actually need?

Fewer than you think, and more specific than you have. Three genuinely well-made case studies covering your three main buyer situations will outperform twelve shallow ones. The number is not the variable. What decides whether they work is whether a reader can find one that looks like their own company.

I get asked this most often by founders who have two case studies and feel behind. Usually they are not behind, they are just uncertain, and the uncertainty gets expressed as a volume target. Twelve becomes the goal because twelve sounds like a company that has done a lot of work.

The better target is coverage. If a buyer from each of your main segments can see themselves in something, you have enough. If they cannot, you could have thirty and still be missing the one that mattered.

Why is more not better here?

Because every extra case study makes the shelf harder to search and each individual one less likely to be read. A visitor does not read your case study library. They scan for the one that resembles their situation, and a long grid of similar-looking cards makes that harder rather than easier.

There is also a quality floor problem. Case studies are expensive to make properly, because they need a real customer, real permission, real numbers and real time. Committing to a large number almost guarantees that most of them are produced cheaply, and a cheap case study actively reduces trust in the good ones beside it.

The third cost is maintenance. Every case study is a claim about a customer relationship at a point in time, and relationships change. A library of twenty means twenty things that could become stale or awkward, and nobody in marketing enjoys the conversation about removing one.

What makes one case study do the work of five?

Specificity that a stranger can verify. A named company, a named person, a described situation with enough detail that somebody in the same position recognises it, and an outcome that is stated plainly rather than implied. Everything else is packaging.

The detail that does the most work is the starting position, not the result. Buyers discount results automatically, because every case study has a good one. What they cannot fake is the before. If the before describes their exact mess, the after becomes believable by association, and the piece has done its job before it reaches any numbers.

The second strongest element is an honest complication. What was hard, what took longer than expected, what the customer was sceptical about. A case study with no friction in it reads like an advertisement, because advertisements are the only stories where nothing goes wrong.

How do you decide which customers to feature?

Pick the ones who look like the buyers you want more of, not the ones with the biggest logos. A recognisable brand impresses your team. A company that looks exactly like the reader persuades the reader. Those are different goals and the first one is usually chosen by accident.

Map your case studies against your buyer segments and look for the gap. If four of your five stories are enterprise and most of your pipeline is mid-market, you have a library that flatters you and does not serve your actual buyer. That mismatch is common and easy to fix once it is visible.

This is also where an unclear buyer profile shows up, because you cannot choose representative customers if you have not decided who you are for. I went through that diagnosis in the piece on an ideal customer profile that is too broad, and the case study shelf is one of the clearest symptoms of it.

Where should case studies live on the site?

Next to the claims they support, with the library as a secondary destination rather than the primary one. A case study linked from the exact service page it demonstrates is doing work. The same case study sitting only in a Customers section is waiting to be found by somebody who was already convinced.

That means the same story appears in more than one place, which some teams resist as duplication. It is not duplication, it is placement. The page a buyer is on when the doubt arrives is where the proof needs to be, and that is rarely the customers index.

Keep the library anyway, because some buyers do go looking, and because it is a useful thing to send a link to during a sales conversation. Just do not expect it to carry the persuasion on its own. It is a reference shelf, not a sales page.

What do you do when you cannot name the customer?

Publish the story without the name, and be explicit that you are doing so. An anonymised case study with a described company, a real situation and a real outcome is worth considerably more than nothing, as long as you do not pretend the anonymity is a stylistic choice.

What kills these is vagueness compounding. If the company is unnamed, the person is unnamed, the industry is approximate and the numbers are rounded to nothing, there is no story left. Anonymity costs you one element. Give everything else back in detail so the piece still reads as an account of something that happened.

Ask for the name anyway, and ask properly. Most refusals are about approval processes rather than reluctance, and offering to write it, to let them edit it, and to hold it until a quiet period converts a surprising number of them. I set out how I handle the unnameable version in the piece on writing a customer story when the customer cannot be named.

How often should they be refreshed?

Review annually, rewrite rarely. The situation described does not age, because the problem your customer had two years ago is still the problem your next customer has today. What ages is everything around it: job titles, product names, screenshots and whether the relationship still exists.

The check that matters most is whether the customer is still a customer. A prominently featured company that churned eighteen months ago is a liability, and it is the kind of thing that surfaces during a competitive deal at the worst possible moment. One annual pass over the list catches it.

Resist the urge to add a fresh date to keep things looking current. A case study with a specific timeframe in it is more credible than one that appears to have happened at no particular time. Buyers are not put off by a story from two years ago. They are put off by a story that seems to be avoiding saying when it happened.

What do buyers actually use them for?

Mostly to reduce the risk of a decision they have already half made. By the time somebody opens a case study, they usually believe you could do the thing. What they are looking for is evidence that a company like theirs survived working with you, and something they can forward to a colleague who was not in the room.

That second use is the one to design for. A case study is often read by one person and circulated to three. The version that travels well is self-contained, explains the situation without requiring context from your website, and can be understood by somebody who has never heard of you.

It also means the internal sceptic is a real audience. Write with them in mind and the piece gets better for everybody, because the things a sceptic wants are the things that make a story credible: names, numbers, timeframes and an admission that something was difficult. Logos do none of that, which is the argument I made in the piece on why customer logos are not proof.

What should a case study never do?

Claim a result the customer would not repeat out loud. If you cannot imagine the named person saying that sentence on a call, it should not be in there. Everything in a case study is borrowed credibility, and borrowed credibility is withdrawn the moment the lender is embarrassed.

It should also never be the only place a number appears. If a result is good enough to headline a case study, it should be consistent with what your sales team says, what your pricing page implies and what your onboarding actually delivers. Proof that contradicts the rest of your site does not raise trust, it raises questions.

And it should never be a thinly disguised product tour. The moment the story becomes a list of features the customer used, it stops being evidence and becomes marketing with a name attached. The customer is the subject. You are a character in their story, not the other way round.

What should you do next?

List your buyer segments down one side of a page and your existing case studies down the other, then draw the lines. The segments with no line are your gaps, and the gaps are the only case studies worth commissioning this quarter.

Then take your single best existing story and check whether it is linked from the page where its doubt arises. Most teams have a good case study sitting in a library that nobody visits, and moving that one link does more than producing a new one.

If you are trying to work out whether your proof is thin or simply badly placed, reach out. It is usually placement, and that is a much cheaper problem to have than the one people assume they have.