Pick your best segment from won and lost deals

What it does

Most founders pick their market by guessing at job titles and industries. Your closed deals already know better. Group them into segments you could build a list from, rank each one on how fast it buys, how much it pays, how long it stays and whether it brings you more, and your aim falls out of the data.

Aim at buyers who close faster, pay more and stay longer, and three of the four numbers behind your ceiling move at once.

What you need

• Every deal closed in the last 12 to 24 months, won and lost, from your CRM
• For each customer: whether they’re still with you, whether they’ve grown, and whether they’ve sent you anyone
• A spreadsheet, or the free template on this page
• Claude or ChatGPT, if you want help with the grouping

How it works

1. 1

   Export your won and lost deals

   For each deal: the company, its industry and size, the buyer’s role, where the deal came from, the dates it was created and closed, its annual value, and the loss reason if you lost it. For customers, add whether they’re still with you after 12 months, whether they’ve grown, and whether they’ve referred anyone.

2. 2

   Group them into segments you could find on a list

   A segment is something you can filter for: an industry, a size band and the role that buys. “Logistics companies with 50 to 200 people, where the head of operations buys” is a segment. “Companies that value efficiency” isn’t. If you can’t build a list of it, it isn’t a segment yet.

   Prompt for Claude or ChatGPT, with your deal export attached

   I'm attaching my CRM's closed deals from the last [12 OR 24] months, won and lost.
   
   1. Group the deals into 3 to 6 segments. Each segment must be something I could filter for in a contact database: an industry, a company size band and the role that buys. Name each one like "logistics companies with 50 to 200 people, where the head of operations buys".
   2. For each segment, give me: deals won, deals lost, the median days from created to won, the average annual value of the deals won, the share of customers still with us after 12 months, and the number of referrals.
   3. List the deals you couldn't place in a segment, and why.
   
   Use only what's in the file. If a number can't be worked out, write "missing" and say what's missing. Don't guess.

3. 3

   Rank every segment on four things

   • How fast they buy: the median days from first call to signed. Fewer is better.
   • How much they pay: the average annual deal size.
   • How long they stay: the share still with you after 12 months.
   • Whether they bring more: referrals and expansion.

   Rank each segment on each one, add up the four ranks, and sort. The lowest total is your best segment.

   Google Sheets: paste into A1, then fill row 2 down, one row per segment

   Segment	Median days to close	Average deal a year	Still with you at 12 months	Referrals	Speed rank	Size rank	Stay rank	Referral rank	Total, lowest wins
   Segment 1	21	12000	70%	2	=RANK(B2,$B$2:$B$7,1)	=RANK(C2,$C$2:$C$7,0)	=RANK(D2,$D$2:$D$7,0)	=RANK(E2,$E$2:$E$7,0)	=SUM(F2:I2)

   With fewer than about 20 closed deals, treat the ranking as a hunch, and test it on your next calls.

4. 4

   Read the lost deals

   Lost deals tell you who not to chase. Group the loss reasons by segment. A segment that keeps losing on price may not value what you sell enough to pay for it. Deals that go quiet after a great demo point at your sales process, not your aim. And a deal lost to budget, timing or a missing feature that has since changed goes back on your list: see reopen lost deals.

5. 5

   Pick your aim, and check you can list it

   Your aim is the top one or two segments. Before you commit, check two things: that you can find them in a contact database, and that there are enough of them to hit your numbers. Count the companies that fit, then write the sourcing spec someone else could build the list from.

What good looks like

You can name your aim in one sentence, with the numbers behind it: “Logistics companies with 50 to 200 people, where the head of operations buys. They close in about six weeks, pay $30K a year, and nine in ten are still with us after a year.”

Mistakes to skip

• Aiming at the segment with the most deals instead of the best ones. A crowded segment in your CRM often just means that’s who you’ve been chasing.
• Segments you can’t build a list from, like “fast-growing companies” or “teams that care about data”.
• Ranking on deal size alone. A big deal that churns in eight months is worth less than a smaller one that stays.
• Leaving out lost deals. They tell you who not to chase, and which reasons might have changed.