On this page
The first campaign we ran for a beta user reached about a thousand people. Within four days, 450 of them had accepted his connection request and written back. His message to me was close to panic: my LinkedIn has blown up, I can't tell which of these people is worth a real conversation and who's just building their network.
The system worked. That was the problem.
That week taught me the principle I now run MyPip on, and it's the one most founders have exactly backwards: your market is never the hundred thousand people who could possibly buy. It's the tiny slice of them who are ready right now. Everything else on this page falls out of that.
Quick answer
The 95/5 rule says that at any given moment, only about 5% of your target market is actually ready to buy; the other 95% aren't looking. For a solo founder that means your real addressable market today is the 5%, so the work is finding them, not building a funnel to catch everyone. MyPip exists to do that finding: a short daily list of the people closest to buying, so five minutes of effort lands where the readiness already is.
What is the 95/5 rule?
Take your target market, everyone who could possibly buy your product as defined by your ideal customer profile. Of that whole group, only around 5% are ready and in a position to buy right now. They're already looking for a solution, or they've decided they need one and are weighing options, or they know roughly what they'd pay. They are most of the way down the path before you ever say a word.
That 5% is your actual addressable market today. Not the 100%.
This isn't just my anecdote. John Dawes at the Ehrenberg-Bass Institute published the research version in 2021: businesses change providers of things like software, banking and legal services roughly every five years, so in any given quarter about 95% of buyers are simply not in the market for what you sell. The number is a heuristic, not a law of physics. The direction is what matters.
What most founders get backwards is where they aim. They try to speak to the whole hundred percent and build a complex marketing funnel to slowly drag strangers toward a purchase they weren't considering. That's the expensive, slow way. It's dramatically easier to find the people already walking toward you and open the door. Conversion stops being persuasion and becomes timing.
If the 5% framing feels small, run the other direction: a market of 100,000 still holds 5,000 people ready to buy this quarter. You need ten of them. The constraint was never the size of the pool. It's knowing which faces in it are looking back.
Why did 450 replies break everything?
Because a reply is not a result. A conversation is.
When our beta user's inbox filled up, the automated system had done its job perfectly and made his actual job impossible. He couldn't research 450 people. He couldn't write anything true to each one. The volume that looked like success on a dashboard was noise at the level where revenue actually happens, one founder talking to one person.
The lesson I took from watching it: sometimes you need to constrict the supply of your own output so you can serve each conversation properly. And you can only afford to send less if you can rank quality first, if you can look at a thousand people and say these four are in the 5% today. Volume is what you do when you can't tell who's ready. Precision is what you do when you can.
Isn't personalisation at scale the fix?
I don't buy it. Personalisation at scale is the same ten thousand sends wearing a better costume.
Look at how the word has degraded. Personalisation used to mean putting a first name at the top of an email. Then it became "I noticed you work at X company." Then it became template variables filled from a data provider. Now it's an AI filling the template variables instead of a script. Same skeleton, shinier flesh.
And AI has a property nobody selling it mentions: it averages toward the mean. No AI sales company has a special model the others don't. So everyone's "hyper-personalised" email regresses to the same email, and the person receiving four of them a day can tell. When AI sales tools first appeared a couple of years ago they worked, because they were new and different. Novelty was the personalisation. That's gone.
The economics are just as ugly. To do ten thousand sends with genuinely deep research on each person, you're looking at $5,000 to $10,000 on data enrichment alone. And since almost everyone buys from the same handful of data providers, you've spent that money to sound exactly like everyone else who did. No differentiation, no edge, a large invoice.
The honest fix runs the other way: drop the volume until real understanding of each person becomes affordable. That's not a limitation of the small approach. It's the whole advantage of it.
Can a handful of messages a day beat 10,000 sends?
A growth marketer once framed it to me as a dashboard: 10,000 sends, a 2% reply rate, 200 conversations a month. Beat that with your five a day.
Fine. Count everything.
To run those 10,000 sends you buy email domains, warm them up for weeks, pay for scheduling software and data tools. Call it $1,000 a month before a single reply, and that's the lean version. Your 200 conversations cost about $5 each, if every reply were a real conversation, which it isn't. Say 10% of them ever become customers: 20 customers, $50 of spend per qualified person, plus the weeks of setup, plus a burned domain when the sender reputation catches up with you.
Now the founder's version. A short list each morning, chosen because those specific people show signs of being in the 5% today. You send them yourself, written like a human wrote them, between coffee and your first meeting. When the people are already close to buying, the conversion on each conversation is a different sport from the conversion on a cold blast. You don't need 200 conversations. You need the right eight.
And the mass approach carries a cost the dashboard never shows: markets are finite. Blast your whole addressable market this quarter and you've spent your 95% before they were ready to hear from you. You get one first impression per person. The volume game sells it for $5.
How do you find your 5%?
You can't see readiness from a job title. That's the uncomfortable part, and it's why most people give up and go back to volume.
The 5% only exists inside a sharply defined ideal customer profile. Get that wrong and you're ranking the wrong thousand people. Get it right and readiness starts to show: someone describing your exact problem in a post this week, a team that just hired for the job your product does, a founder asking for tool recommendations in a community. The signals are public. Reading them for a thousand people is the part that doesn't fit inside a human morning.
This is the actual shape of my own routine now. MyPip sends me an email every morning with the day's actions and the specific people behind them, chosen overnight because something about them says now, not someday. My job is the five minutes of judgment and the send. The reading happened while I slept.
What the 95/5 rule stopped us building
The biggest product decision the 95/5 has forced at MyPip is the one users ask about most: automated sending. People want it. We keep saying no.
Automation only makes sense if you're trying to appeal to the 95, because automation needs volume to justify itself; you have to hit a quantified number of people to make the machine worth switching on. A short daily list of genuinely ready people cannot feed a machine, and it doesn't need to. If they're in the 5%, a handful of real messages is what it takes.
Has the refusal cost us anything? Some sign-ups, probably. It's also the only reason the beta user with the flooded inbox is still a user. That trade is the principle in one line: we'd rather help you have four real conversations than automate four hundred fake ones. It's one motion inside the full first-customers guide, and everything else in that guide gets easier once you accept who you're actually talking to.