Five Questions Your Case Studies Keep Dodging (And Why the Answers Sell the Work)
Photo: Elshad Iman (Elşad İman), CC BY-SA 4.0, via Wikimedia Commons
Most portfolios are a highlight reel. Big hero image, clean type, a few screens, a green number at the end. They look great in a tab on a laptop at a conference.
They also answer almost nothing.
If you've ever sat on the other side of the table — the side where someone is deciding whether to trust you with six figures and a launch date — you know what that reel feels like. Shiny. Slippery. You can't get a grip on it.
Case studies aren't proof that you did something. They're proof that you can think. And thinking is what clients are actually buying. So here are five questions your current write-ups probably aren't answering, and why answering them is the cheapest sales work you'll do this quarter.
1. What was true at the start?
Every project begins inside a set of constraints, and most case studies sand them off. 'The client wanted a modern refresh.' Okay — and what else? What was the budget? What couldn't you touch? Was there a legacy CMS held together with plugin tape? A brand book signed in blood by a founder who's still on the board? A hard date because of a trade show in October?
Constraints are where craft shows up. Anyone can design something nice with unlimited time and no stakeholders. Naming the box you had to work inside tells a reader what kind of problem-solver you are. 'Six weeks, a locked brand book, and a payments system nobody was allowed to open' is a sentence that sells. 'We partnered to reimagine the digital experience' is a sentence that floats away.
2. What were the two or three decisions that actually moved the work?
Not your process. Not the phases. The forks in the road. Where did the project go one way instead of another, and who decided?
This can be three sentences. 'We argued about whether the homepage should lead with product or proof. We tested it with twelve users. Proof won, and it changed the whole nav.' That's a case study. It tells me you run arguments to ground instead of letting them run forever.
If you can't name the decisions, you probably weren't driving the work — which is a different problem, but worth knowing.
3. What did you try and throw away?
This is the question almost nobody answers, and it's the one that separates an honest portfolio from a brochure.
Rejected options are evidence of depth. The abandoned homepage direction. The name that tested badly. The feature you cut after round two because the data said no. Show one. Show the sketch, even. It costs you nothing and it signals that your first draft isn't your final answer — which is exactly what a nervous client wants to hear.
4. What number moved, and what number stubbornly didn't?
Outcomes matter. But a case study where every metric improved is a case study nobody believes. Real work has mixed results, noisy attribution, and at least one thing that didn't budge.
So say that. 'Checkout completion went up 19% over two months. Newsletter signups stayed flat — the placement was wrong and we knew it by week three.' That last clause does more for your credibility than the 19% does. It tells the reader you'll tell them the truth when the truth is inconvenient, which is the single most valuable trait a partner can have mid-project.
Keep the numbers honest and scoped. Don't claim credit for a quarter of revenue because you redesigned a footer.
5. What would you do differently?
The retrospective question. Not self-flagellation — just a signal that you've thought about the work since you shipped it. 'We'd have brought the content team in a month earlier.' 'We over-designed the settings page.' 'We should have pushed harder on the CMS choice.'
One or two lines. Done. It shows a team that keeps learning, and it quietly reassures the reader that if something goes sideways on their project, you'll notice.
Why the flawless version loses
There's a persistent belief that case studies are a beauty contest and the cleanest one wins. In practice, buyers are pattern-matching for risk. They're trying to figure out what you'll be like in week nine, when the stakeholder email lands at 11pm.
A story with constraints, tradeoffs, and a dead end reads like a real working relationship. A story where everything went smoothly reads like marketing. Marketing is fine. Trust is better.
And honestly? The messy version is more memorable. Clients repeat your war stories in their own internal meetings. 'The team that killed their own homepage concept because the data said so' is a sentence that travels.
Run the audit this week
You don't need a new site. Pick your three best case studies and add a short paragraph under each of the five questions. Twenty minutes apiece. If you can't answer one, that's information too — either dig it out of the archive or stop featuring that project.
Then look at what you've got. It'll read less like a portfolio and more like a working method. That's a much harder thing to comparison-shop.