Most founders blame the writer when a case study falls flat. The real damage happened three months earlier, in a pilot kickoff call nobody thought to record properly.
By the time you sit down to write the story of your first real result, the decisions that would have made it easy are already gone. Nobody captured the baseline number. Nobody got the practice's sign-off to use their name. The clinician who said the one sentence you'd love to quote said it on a call that wasn't recorded, and now you're paraphrasing from memory and hoping legal doesn't ask for a source.
This is the part of proof-building that has nothing to do with content strategy and everything to do with it. A case study is only as good as what you collected while it was happening. Fix the collection, and the writing takes an afternoon. Skip it, and no writer, however good, can invent a before-and-after number that was never measured.
This lands on Felix's desk, whether he wants it there or not
Felix's fundraising story next year runs on evidence he decides to gather this quarter, long before the pitch deck gets built or the narrative gets workshopped with advisors. The raw material: a number before, a number after, a name willing to be quoted, a date attached to both.
Investors have heard "we're seeing strong early signal" a thousand times. What they haven't heard a thousand times is a specific practice, a specific metric, a specific quarter. That specificity is what separates a slide that gets skimmed from one that gets forwarded to a partner. And specificity isn't something a writer conjures at the end. It's something an operator decides to preserve at the start.
Ask a sharper question instead: what are we doing right now, this pilot, that we'll wish we'd done when we finally sit down to write it.
What actually gets lost
The failure mode is almost always the same four gaps, and they're boring enough that nobody flags them until it's too late.
No baseline. A pilot goes live, everyone's focused on onboarding, and three months in someone asks "so what was the no-show rate before we started?" Nobody knows. Nobody wrote it down on day one, because day one felt like a launch, not a measurement moment.
No consistent metric across pilots. Pilot one tracks patient response time. Pilot two tracks staff hours saved. Pilot three tracks something else entirely, because whoever ran onboarding picked whatever seemed most relevant that week. A year later, there's no clean comparison across sites, just three disconnected anecdotes.
No consent captured early. The practice loved the results. Nobody asked, in writing, at the time, whether their name could be used in a case study. Six months later, when marketing finally asks, the practice manager who championed the rollout has moved on, and the new one isn't comfortable signing off on something she wasn't part of.
No record of what people actually said. The best quote of the entire pilot happened on an unrecorded call, in passing, and everyone in the room remembers it a little differently now.
None of these require more resources to fix. They require deciding, before the pilot starts, that proof is a deliverable, not a byproduct.
A pattern worth recognizing
Picture the version of this that plays out at most seed-stage health tech companies. A pilot launches with a practice that's genuinely excited. Three months in, the practice manager mentions offhand that no-shows have dropped noticeably. Everyone in the room nods. Nobody writes the number down.
Six months later, marketing wants a case study. The founder calls the practice, asks for the number, and gets a shrug: "it felt like it improved, we didn't track it that precisely." The founder ends up writing something soft, "practices report meaningful improvements," because that's the most honest sentence the data supports.
The product might have genuinely worked. The pilot might have been a real success. But the proof of it evaporated the day nobody wrote the baseline down, and no amount of good writing later can put it back.
Four things to start tracking this quarter
You don't need a research team. You need four habits, applied consistently, starting with the next pilot you launch.
A baseline snapshot, taken on day one. Before anything changes, write down the number that matters most for this specific pilot: no-show rate, staff hours on scheduling, patient response time, whatever the pilot exists to move. One paragraph, one number, timestamped. It takes ten minutes and it's the one artifact in the entire pilot you can never recreate later. Assign it to a real person on the kickoff call, not a vague "someone should probably note this down."
One metric, held constant across every pilot. Pick a primary number that applies across sites, even if each pilot also tracks something specific to its context. A common thread across three or four practices is a pattern. Three disconnected numbers are just three anecdotes that happen to sit in the same folder. Put the metric in the pilot agreement itself, so it's not something customer success improvises site by site.
Consent, requested at kickoff, not at exit. Add one line to the pilot agreement: permission to reference the practice, anonymized or named, in future marketing material, with final approval sitting with them before anything publishes. Ask when the relationship is warmest and the paperwork is already open, not months later when you need a favor from someone who's since moved on.
A running log of what clinicians actually say. After every check-in call, whoever's on it writes down one sentence: a direct quote, in the clinician's own words, with the date. Most of it won't be usable. Some of it will be the exact line that makes the case study feel real instead of written by marketing. A shared doc with a running list beats trying to remember it eight months later.
Who actually owns this
The reason these four habits don't happen on their own is that they don't obviously belong to anyone. Product is focused on the build. Customer success is focused on keeping the pilot happy. Marketing hasn't been invited to the kickoff call yet, because there's nothing to market.
That gap is exactly where the baseline number gets lost. Fix it by naming an owner before the pilot starts, even if it's the founder himself for the first few. One person, one shared doc, four fields to fill in as the pilot runs. It's a smaller lift than it sounds, and it's the difference between a case study that writes itself later and one that has to be reverse-engineered from memory.
What this changes about the writing itself
When these four things exist, the case study stops being a creative writing exercise and starts being an editing job. The baseline and the current number give you the shape of the story before a single sentence gets drafted. The quote gives you the voice that isn't yours. The consent means nobody's scrambling for legal sign-off the week you want to publish.
Compare that to the alternative: a founder trying to reconstruct a narrative from memory, guessing at a number that sounds directionally right, hedging every claim because nothing was actually recorded. That's where the "we believe" and "early indications suggest" language creeps in, the hedging that makes a founder sound unsure of a result that might genuinely be strong. That hedge is a data problem dressed up as a writing problem.
The version investors actually remember
A partner sitting through a pitch has heard "significant improvement in patient engagement" from every company in the category this quarter. What they remember afterward is different: a name, a number, a before and after, a sentence a real clinician said out loud.
The reason is simple: it's checkable. An investor can picture calling that practice. A vague claim, however confident, can't be checked by anyone, and experienced partners have learned to discount exactly the claims they can't verify.
The founders who walk into their Series A conversation with this kind of proof aren't the ones who hired the best writer at the last minute. They're the ones who decided, two pilots ago, that the number and the quote were worth capturing the moment they happened.
If you're already six months in with nothing captured
Maybe this reads a quarter too late. The pilot's been running since spring, nobody's been tracking a baseline, and the case study you wanted for this fundraising cycle isn't coming together.
You've got two options, and only one of them is honest. The first is to write around the gap: soft language, no real number, a quote you're reconstructing from a call three months back. That's the version that reads as unsure, because it is.
The second is to start today, on a second pilot or a new phase of the same one, and be straightforward with the practice about why. "We didn't capture this as well as we should have last time. Can we set a proper baseline now and check in again in eight weeks?" Most clinicians respect that more than they'd respect a polished case study that quietly skips the specifics. And it gives you something real to publish two months from now instead of something vague to publish this week.
Either way, the fix looks the same: stop trying to write proof you never collected, and start collecting the proof for what comes next.
Where to start this week
Look at whatever pilot is currently live. Do you have a baseline number written down anywhere. Has anyone asked for permission to use the practice's name yet. Has a direct quote from a call been written down in the last month. Is one person actually responsible for any of this.
If the answer to any of those is no, fix it this week, not at the end of the pilot. The baseline you don't capture today doesn't exist next quarter when you need it.
Everything else about proof, the writing, the framing, the placement in a fundraising deck, is downstream of that one decision. Get the collection right, and the case study you write in month twelve reads like it was easy. It was, because the hard part happened months earlier, one habit at a time.