Most case studies read like a press release with a logo stapled on. They list features, thank the client, and forget to tell a story anyone needs. A good one does the opposite: it shows a buyer who looks like them solving a problem they recognize, and it earns belief instead of asking for it.
This is a guide to writing the version that actually works. What a case study is for, how to choose a story worth telling, the structure that carries it, the proof that makes it credible, and the line you should never cross to get there.
A case study is a transfer of belief
Before the format, the purpose. A prospect reading your case study is asking one quiet question: has this worked for someone like me? Everything in the piece either answers that or wastes their time.
That reframes the job. You are not writing a victory lap for your own team. You are lending a future buyer the confidence of a past one. The hero is the customer, the stakes are their problem, and your work is the turning point in their story, not the subject of it.
A case study isn’t proof that you’re good. It’s proof that someone like the reader got what they came for.
So the test for every sentence is simple. Does it help a stranger picture themselves making the same decision and getting the same result? If the line only flatters you, cut it.
Choose a story worth telling
The biggest quality lever is upstream of the writing. It’s selection. A well-written case study about a forgettable project will always lose to a plainly told one about a project that mattered.
Pick stories that do three jobs at once.
- They match a buyer you want more of. The customer in the study should resemble the prospect you’re trying to convince: same segment, same scale, same kind of pressure. A reader pattern-matches on the logo and the problem before they read a word of the body.
- They have a real before and after. There has to be tension. Something was broken, slow, ambiguous, or at risk. If the situation was fine and got slightly nicer, there’s no story, just a testimonial in disguise.
- They have an outcome you can stand behind. You need an end state that’s true, specific, and ideally measurable. Not a feeling. A change.
When you have several candidates, choose the one where the stakes are highest and the result is cleanest. A messy win you can prove beats a tidy win you can only assert.
The structure that carries it
A case study is a narrative, not a spec sheet. The shape that works almost every time has five beats, and they run in this order for a reason.
1. The customer and their situation
Open on the customer, not on you. Who are they, what do they do, and what was the world like before the work started? Keep it tight, but make the reader recognize the setup. This is where they decide whether the story is about someone like them.
2. The problem and its stakes
Name the problem in concrete terms, and say why it hurt. A slow site is a fact. A slow site costing them signups before launch is a stake. Stakes are what make a reader lean in, because stakes are what they’re carrying too. If you’re documenting web work, the cost is usually measurable, which makes the problem easy to state in the customer’s own business terms rather than yours.
3. The approach and the decision
Now you enter the story, but as a guide, not the lead. What did you do, and more importantly, why? The reasoning is the part that builds trust, because it shows judgment rather than just activity. Skip the exhaustive task list. A reader doesn’t need every sprint. They need the two or three decisions that changed the outcome, and the trade-offs you weighed to get there.
4. The result
This is the beat readers skim to first, so make it land. State what changed, in terms that matter to the customer’s business, not your craft. Pair the number with the meaning: not just a faster load time, but the conversions that followed. If you can show a before and after side by side, do.
5. What it meant
Close on the human end of the story. A quote from the customer in their own voice, a line about what the change made possible, a sense of where they are now. This is where a transaction becomes a relationship in the reader’s mind.
You can bend this order, but you can’t skip the tension. A case study with no problem is a brochure.
Make the result credible
A claim a reader doesn’t believe is worse than no claim, because it taints everything around it. Credibility is built from specifics.
- Use real numbers with context. “Faster” means nothing. “Page load dropped from 4.1 seconds to under one, and signups rose over the next quarter” means something. If you’re going to talk results, learn to frame them like business outcomes, which is the whole discipline of measuring design ROI.
- Quote the customer saying something only they could say. Generic praise reads as invented. A specific quote about a specific moment reads as true. Push past “they were great to work with” to the detail that proves it.
- Show, don’t only tell. Screenshots, a before-and-after, a real artifact from the work. Evidence the reader can see does more than an adjective ever will.
- Name the constraints. A study that admits the timeline was tight or the scope was hard is more believable than one where everything was effortless. Honesty about friction is a credibility signal.
Where you can’t share a number, share a mechanism. Explain why the work produced the result, so a reader can trust the logic even when the metric is confidential.
The line you don’t cross
Here’s the trade-off that decides whether your case studies build trust or quietly erode it. The temptation is always to round up: to borrow a metric, to imply a result the work didn’t cause, to make the before worse than it was. Don’t.
Inflated case studies are a short-term gain and a long-term liability. Buyers compare notes. Customers read the study about themselves. The moment a number can’t survive a follow-up question, the whole document, and your credibility with it, collapses. A modest, true result will out-sell an impressive, shaky one every time, because the reader can feel the difference even when they can’t name it.
If the only way to make a story impressive is to stretch it, you picked the wrong story. Go back to selection.
How Strynal approaches case studies
We treat a case study as part of the brand, not a sales attachment written after the fact. It has to carry the same voice as the site and the messaging around it, because a reader who feels a seam between your polished homepage and your thin proof stops trusting both. The standard is the same one we hold for all editorial that carries a brand: every claim earns its place, and nothing ships that we couldn’t defend in the room.
That works because the people who scoped and built the engagement are the ones who help shape the story, so the reasoning in the study is real reasoning, not reverse-engineered marketing. The before and after are accurate because we were there for both. And the writing reads like the rest of your site, since the same care goes into a case study as goes into the copy on every page.
If your wins aren’t pulling their weight on the page, the fix is usually a better-chosen story told with more honesty, not more superlatives. See how we build proof that persuades in motion and content, or tell us what you’re trying to prove and we’ll help you turn the work you’ve already done into the case for the next one.