EIGHT PAGES
ONE WORKSHEET
Build an ideal customer profile out of your own record instead of a workshop.
01
THE
PROBLEM
Most ideal customer profiles are produced in a workshop, contain a persona with a stock photograph, and are never opened again by anybody who has to make a call.
The format is not the problem. The input was. Sales believes one segment, marketing believes another, product believes a third, and the workshop resolves it by averaging the three. What comes out is a profile that nobody actually wins with.
Your closed-won and closed-lost record already contains the answer. It has simply never been read as a dataset.
02
You are not looking for what your customers have in common. You are looking for what your winners have that your losers do not.
02
Twenty won, twenty lost, most recent first. The discipline is in what you leave out.
| COLUMN | WHY IT EARNS ITS PLACE |
|---|---|
| Industry, sub-vertical | Usually too crude on its own, but it is the filter everyone starts with. Test it, then move on. |
| Revenue, headcount | The same. Check whether it separates before you build a target list on it. |
| Business model | Corporate or franchised, single or multi-site, direct or channel. Often where the real signal lives. |
| Buyer, and signer | Frequently different people. The distance between them predicts your cycle length. |
| How it started | Inbound, referral, outbound, partner. Source correlates with win rate more than teams expect. |
| Cycle, deal size | Outliers in either direction are worth reading individually rather than averaging away. |
| Trigger in 12 months | New exec, funding, acquisition, incident, new locations. The most under-recorded and most predictive column here. |
| Still a customer? | A closed-won that churned in eleven months is not a win. It is a lesson. |
03
EXPECT
Three separating attributes. Occasionally four. If you find nine, you have found noise rather than a profile.
READ IT LIKE THIS
An attribute that separates puts your wins on one side and your losses on the other. An attribute that does not is a description, not a filter.
04
THE TRAP
Panel three is the one that fools people. Every customer is in that vertical, so it feels like a finding. It only tells you where you have been selling.
03
Most companies track fit and ignore signal, then wonder why timing always feels wrong.
| DIMENSION | WHAT YOU ARE MEASURING | WEIGHT |
|---|---|---|
| Structural pain | Does the problem you solve exist at their stage and model | 30 |
| Firmographic fit | Revenue band, headcount, geography, growth rate | 25 |
| Access | Whether you can reach the economic buyer at all | 15 |
| Technographic fit | Systems already in place that you complement or replace | 15 |
| Timing signal | A trigger inside the last ninety days | 15 |
THEN TIER
Tier 1 · top 10%
Full research, multi-threaded, human-led. Twenty-five to forty accounts per rep, by name.
Tier 2 · next 30%
Insight-level personalisation, lighter touch. Run as a programme, not by name.
Tier 3 · the rest
Exists to catch timing, never to force it. Worked only when a signal fires.
04
Tier 1 gets worked whether or not a signal fired. Tier 3 only when one does. Most teams invert this, then wonder why their best-fit accounts went untouched for three quarters.
04
Twenty minutes each. Never run by the rep who owned the deal.
The sheet tells you what correlates. Only the interviews tell you why.
What was happening in the business that made you look at this at all?
Who else was involved, and what did each of them care about?
What nearly stopped it?
What did we say that mattered, and what did we say that you ignored?
If you had chosen differently, what would have made the difference?
ON FIVE
The most valuable question and the least asked. In a loss it gives you the real reason rather than the polite one. In a win it tells you which part of your pitch was load-bearing.
Record the answer in the buyer's own words. A reason code is a category. A sentence is evidence, and it will still be useful in a year.
05
Ten interviews is a day and a half of work. It is the highest-return day and a half available to a revenue leader.
05
Present in most wins. Absent in most losses.
Write each as something a rep could check in two minutes, rather than judge over the course of a call.
ONE
HOW A REP VERIFIES IT
TWO
HOW A REP VERIFIES IT
THREE
HOW A REP VERIFIES IT
STOP
What we will stop targeting as a result
06
The last box is what makes this real. A profile that removes nothing from your target list has not been applied. It has been filed.
FREE
UNGATED
NO FORM
The Stage Definition Rewrite
OPERATIONS
The Ramp Plan
TEAM
The Five Questions
ENABLEMENT
The Expansion Motion
EXPANSION
The Outbound Playbook
STRATEGY
If you run this read yourself and never contact me, that is a good outcome.
WRITTEN BY
Josh Dempsey
CONTACT
josh@stageonegtm.com
EDITION
First, 2026