The Case Study Format That Actually Gets Read
This is for marketers and founders who have real results and a case study that nobody reads past the first paragraph. The single most important decision is where you put the number. Get that right and the structure around it is easy.
Decide this first
Put the outcome in the headline and the first two sentences. Not the setup, not the client's industry, not "the challenge." The number.
Every case study I've seen die a slow death on a company blog made the same choice: it opened with context. "Acme Co. is a mid-sized logistics company facing challenges in customer retention." Nobody keeps reading that. I wouldn't.
When I ran marketing on a DTC brand from $100K to $3M in revenue, the case study that actually got shared started with: "We took a skincare brand from $100K to $3M in 18 months without raising ad spend past $40K a month." That's it. That's the hook. Everything else in the piece exists to back that sentence up.
Once you commit to leading with the result, the rest of the format follows. You're not writing a story that builds to a reveal. You're writing proof that unfolds in order of how skeptical your reader is.
What to look for
A headline number, not a headline theme
"How We Improved Conversion Rates" is a theme. "How a 2-person team took checkout conversion from 1.8% to 3.4% in 90 days" is a number. The second one tells a skimmer whether to stay. The first one asks for trust you haven't earned yet.
If you only have one good number, use it in the title. If you have three, pick the one that's hardest to fake. Revenue and conversion rate beat "engagement" and "brand lift" every time, because readers know how to judge them.
A before-state that's specific enough to hurt
This is the part most people rush. They write "the client was struggling with lead generation" and move on. That's not specific, it's generic enough to apply to any business.
On the $2.2B infrastructure project I marketed, the before-state wasn't "low public awareness." It was: 62% of residents in the affected zip codes couldn't name the project when surveyed, and local opposition had delayed two prior permits by a combined 14 months. That's a before-state with teeth. It tells the reader what was actually at stake and gives the later numbers something to beat.
Readers stick around when the problem feels real, not when it feels like a template. Use a number in the before-state too, not just the after.
A method section that reads like a recipe, not a résumé
This is where most case studies either keep people or lose them. Don't write "we developed a comprehensive multi-channel strategy." Write what you actually did, in order, like you're handing someone instructions.
- What channel or lever you pulled first and why
- What you tested, and what the losing version looked like
- What you changed when the first approach underperformed
- The timeline, in weeks or months, not "over time"
On that DTC brand, the method section said we cut Facebook spend by 30% in month two because CAC had crept from $22 to $41, and moved that budget into SMS and email flows, which were converting at 11% versus paid social's 2.3%. That's a recipe. A reader can picture doing it. "We optimized our channel mix" is not a recipe, it's a sentence designed to avoid saying anything.
What to ignore
Skip the client testimonial quote in the middle of the piece. It interrupts the logic and almost always says something vague like "the team was a pleasure to work with." Save a quote, if you need one, for the very end, and make sure it includes a number or it's dead weight.
Skip the long bio paragraph about the company's mission and values. Nobody clicked on a case study to learn your client's brand philosophy. One sentence of context is enough. If you need three paragraphs to explain who the client is, the result probably isn't strong enough to stand on its own.
Skip stock photography of handshakes, whiteboards, or people pointing at screens. It signals the piece is filler before anyone reads a word. A chart of the actual before-and-after number does more work than any photo ever will.
Skip the "about our methodology" section that reads like a capabilities deck. If your case study starts sounding like a sales pitch for your agency rather than a record of what happened, you've lost the thread. The result should sell the service. The service shouldn't need to sell itself.
Common mistakes
The biggest one: burying the number. I've read case studies where the actual result, the thing that should be the headline, shows up in paragraph nine. By then, 80% of readers are gone. Scroll depth data from every blog I've run backs this up: you lose the majority of readers in the first 15 seconds, which is roughly two paragraphs. If your proof isn't in those two paragraphs, it doesn't exist for most of your audience.
The second mistake is writing the case study for the client instead of the prospect. I did this early in my career, at Petrobras, where PR materials were built to make the subject look good rather than to make a future client understand what was possible. That instinct carries over into marketing case studies constantly. People write to flatter the logo they worked with, not to convince the next one. If you're choosing words because they'll make the client happy rather than because they're the clearest version of the truth, you're writing the wrong document.
The third mistake is false precision without context. "We increased engagement by 340%" means nothing without a baseline. If engagement went from 4 people to 17.6 people, a 340% increase is true and also worthless. Always pair a percentage with the raw numbers behind it. "Email open rates went from 11% to 34%, worth about 2,400 extra opens a month" tells a real story. "211% increase in open rates" by itself tells the reader you might be hiding something.
FAQ
How long should a marketing case study be?
Long enough to prove the number, short enough that nobody has to skim. Most good ones run 600 to 1,000 words. If you're past 1,500, you're probably explaining instead of showing. Cut background, keep mechanics.
Should I use a template with fixed sections like Challenge, Solution, Results?
You can, but don't let the labels do the thinking for you. "Challenge" sections tend to drift into vague scene-setting because the heading gives people permission to ramble. If you use that structure, hold yourself to the same rule: specific number in the before-state, specific actions in the solution, specific number in the results. The labels are furniture. The numbers are the point.
What if the result wasn't dramatic?
Then lead with the constraint instead of the outcome. "How we grew organic traffic 22% with a $0 content budget" is a fine headline even though 22% isn't huge, because the constraint makes the number meaningful. A modest result with a hard constraint beats a big number with no context.
Do case studies need a client quote to be credible?
No. A specific number with a believable before-state is more credible than most quotes, because quotes are easy to doctor and numbers invite scrutiny. If you have a quote with real specifics in it, use it. If it's generic praise, cut it.
Where should the case study live, blog or dedicated landing page?
Landing page if it's meant to close deals, blog if it's meant to build search traffic and trust over time. The structure doesn't change. What changes is the call to action at the end, and how hard you push it.
If you take one thing from this: write the number first, then go back and build the proof underneath it. Everything else is detail work.