A prospect told me he was going into surgery. His side comment became my best article and my AI roadmap.
I run a bootstrapped work management SaaS. A few weeks ago a warm prospect told me he could not evaluate the product because he was going in for major surgery. Then, almost as an aside, he described how he manages his projects today: every week he feeds an AI model his charter, plan, risk register, tracker and six weeks of reports, under a prompt he iterated for months. Three A4 pages, font size 10.
That aside did three things. It handed me the clearest articulation of the problem my product solves that I have ever heard (his words: the challenge is obtaining the information, not analysing it). It handed me a ranked AI roadmap, because he had hand built the five exact workflows PMs will pay for. And it handed me an article, which I published today:
https://medium.com/p/4e3927c005a0
The lesson I keep relearning: your best marketing copy is sitting in your DMs, verbatim, written by the people you are trying to sell to. I did not invent the phrase "three pages of prompt". He said it. I just noticed it.
Happy to share how the article performs across channels if anyone is interested.
This matches what I've noticed too, the line that converts is almost never the clever one, it's the one a real person said in their own words about their own problem. Writing copy from scratch tends to sound like marketing, writing it from a transcript tends to sound like a person. Do you keep a running doc of these lines as you hear them, or did this one just happen to stick?
This is a really interesting type of signal.I think the tricky part is not noticing the comment, but knowing how much weight to give it.A single prospect can sometimes reveal something important, but it’s also easy to overfit to one person’s situation.Curious how you’re deciding whether this becomes a bigger product direction vs just a valuable one-off insight.Have you seen similar patterns from other prospects, or was this the first signal that pushed the roadmap this way?
This lands hard. I run a work management problem too, just across three of my own businesses instead of one company. The pattern you noticed is exactly what I built FounderFlow around, the signal is almost never missing, it is just buried under everything else competing for attention. Watching for it costs more than most founders admit.
The 'challenge is obtaining the information, not analysing it' line is the kind of thing you can only get from a real conversation, never from a survey or interview script. Sales calls are the most underrated research tool most founders have and almost nobody mines them properly. I keep a note of exact phrases prospects use, especially the throwaway ones like this, because they're almost always sharper than anything the marketing team would write. His verbatim description of the problem is probably also your best headline if you A/B tested it. Would be curious how the Medium article performs on SEO vs. the IH post in terms of inbound over the next month.
The line about getting information rather than analyzing it is the one that matters. People can spit out a decent summary now, but getting the update out of someone's head and into the system still takes a dozen tiny interruptions. That's part of why I built DictaFlow: hold a key, say the project update while you're working, release, and it types into the field. A better prompt won't fix that capture step.
The article insight is solid, but the roadmap is the part I'd pressure-test. A guy who hand-built a three-page prompt and iterated it for months is your most extreme user, not your median one, and the five workflows he built are the five he needed. Power users tend to want things the next fifty PMs never touch, so it's worth checking those five hold up across a few colder prospects before you rank a roadmap by them. And yes, I'd take you up on the channel breakdown: if 'copy from your DMs' converts, the real question is whether it's the words doing it or just the audience each channel sends.
This is the good stuff — and the surgery-aside is what makes it real. Those unguarded asides are exactly where the honest problem statement lives; the moment someone stops reciting requirements and just describes what they do on a Tuesday, you get the truth.
The "three pages of prompt, font size 10" detail is carrying a lot here too. It's not just a nice quote — it's a signal of intensity. Someone who hand-built a 3-page prompt and iterated it for months isn't a casual want; that's real pain, already self-treated. Those are the exact words you put on your landing page, because they've done your positioning for you.
One thing I'd add: capture the phrasing AND the friction that produced it. "Obtaining the information, not analysing it" is gold, but so is the weekly ritual it took to get there. The pain is the setup; the quote is the punchline — ship both.
Curious what the article's early numbers look like. "Your best copy is in your DMs" is easy to believe and hard to operationalize, so real conversion data would be genuinely useful.
The part that stuck with me wasn't the prompt—it was the sentence, "the challenge is obtaining the information, not analysing it."
That changes where the product's value lives. If AI is becoming good at reasoning, the competitive advantage shifts toward reducing the effort required to assemble trustworthy context in the first place.