Your reps have learned, one wasted call at a time, which leads look like the deals you win. Write that knowledge down and the argument gets much shorter.
Picture the monthly pipeline meeting. Marketing brings a slide: 140 leads handed to sales last month, up on the month before. The sales manager brings another: 11 meetings booked. Someone says the leads are rubbish. Someone else says half of them were never even called.
Both people are probably right. Nobody leaves the room knowing what to change, and next month the same two slides come back.
Why reps skip leads
Watch a good rep go through a fresh batch of leads. They don't work it top to bottom. They scan the company name, the website, the headcount, the industry and maybe the job title, and inside ten seconds each lead lands in one of two piles: looks like the people who buy from us, or doesn't. The second pile gets a token email, or nothing.
That scan is a fit check, run from memory of the deals the rep has won and lost. Someone five years into the seat is usually right. Someone in their first quarter is mostly guessing, and copying whatever the senior rep next to them does.
Marketing tends to be measured on a different idea of a good lead. A lead counts as qualified because it downloaded the guide, came to the webinar or crossed a score threshold built on page views and email clicks. All of that tells you someone is paying attention. None of it tells you whether their company looks like your customers.
So the two teams work from two definitions of a good lead, and only marketing's is written down. Sales keeps its definition in people's heads, where marketing can't see it, can't test it and can't aim campaigns at it.
The pizza van that once landed in my own lead allocation got there through exactly this gap. Every step of the process worked. Nobody had asked whether a one-person catering business could ever buy what we were selling.
How to write one definition of fit that both teams can check
You can do this in a spreadsheet in an afternoon. All you need is a CRM export and someone from each team in the room.
- Export every deal you closed in the last 12 to 24 months, won and lost. For each one, pull the company facts you actually hold: industry, employee count, country, lead source. You need the losses as much as the wins. A trait that shows up in most of your wins and most of your losses tells you nothing about fit.
- Put each fact into a few coarse buckets. Employee count in bands (1 to 10, 11 to 50, 51 to 200, 201 to 500, over 500). Industry in the eight or ten groups your team really uses, not the 140 categories LinkedIn offers. With small numbers, fine buckets give you noise.
- For each bucket, count wins and losses and work out the share won. Compare that with your overall share. Buckets well above it are your fit. Buckets well below it are where reps are right to hesitate.
- Write the definition as one short paragraph, with the counts beside it. For example: "Our best fit is a financial services firm with 51 to 500 staff (based on 23 closed deals in financial services, 14 won). Retail and companies under 50 staff rarely buy from us." The counts let either team check the claim without taking anyone's word for it.
- Put it where both teams work. Add a simple fit field to the CRM (good, partial, poor) that marketing fills in before handover, using the paragraph. Agree that a poor-fit lead with lots of engagement still goes to sales, labelled, so the rep knows what they are picking up. Put a review date on the paragraph.
Here is step 3 on a small book of deals.
| Bucket | Won | Lost | Share of closed deals won |
|---|---|---|---|
| 11 to 50 employees | 6 | 22 | 21% |
| 51 to 200 employees | 21 | 18 | 54% |
| 201 to 500 employees | 9 | 10 | 47% |
| Financial services | 14 | 9 | 61% |
| Retail | 2 | 15 | 12% |
| UK | 25 | 30 | 45% |
Illustrative data, made up for this example. Across all 86 closed deals in it, 36 were won: 42%.
The 51 to 200 band and financial services sit well above 42%. Retail and small companies sit well below. The UK is close to average, so country is doing little work here, even if everyone in the office assumes it matters. That last line is often the most useful one in the whole exercise, because it removes a rule nobody needed.
If you want to go a level deeper on the analysis itself, I've written up how to analyse your closed-won deals step by step. To check a single company against the same counts, see how to tell if a company is a good fit.
Where this usually goes wrong
- Looking at wins only. Most of your customers probably have a website and a UK address. So do most of the companies that said no. Always put the losses next to the wins.
- Reading too much into small numbers. Two wins out of three in a sector is a coincidence until proven otherwise. As a rough rule, set aside any bucket with fewer than ten closed deals and say so in the definition.
- Treating each trait on its own. Company size and industry often travel together, so a strong showing in one can be the other in disguise. A spreadsheet won't untangle that easily. Say so in the paragraph rather than stacking percentages.
- Writing it once and forgetting it. As you sell into new markets the wins change, and a year-old definition quietly starts sending reps the wrong way. The review date matters.
- Using it to keep score on people. The definition is there to decide which leads get worked first. If marketing reads it as a verdict on their lead count, or reps read it as an audit of their call logs, both teams will stop using it within a month.
What to do with the answer
The monthly meeting now has something to look at. The question changes from "are the leads rubbish?" to "how many of last month's 140 matched the definition, and what happened to those?"
If good-fit leads went uncalled, that's a follow-up problem, and sales owns it. If most of the 140 were poor fit, that's a targeting problem, and marketing owns it. Usually it's some of each, and now you can see how much.
Marketing can then move spend towards the campaigns that bring in good-fit leads, rather than the ones that bring in the most leads. New reps get on day one the pattern the senior reps took years to pick up.
Where Telepath fits
Telepath does this job for HubSpot teams, from your own data. It learns what your closed-won deals have in common and gives every open deal a T-Score from 0 to 100: how much it looks like your wins. The T-Score is a fit score. It measures resemblance to past wins and makes no forecast about any single deal.
Each open deal also has a "Why this score" panel. It lists every trait the score uses, how well the deal's value fits your wins and how much of the score each trait accounts for, so a rep or a marketer can see the reasoning instead of arguing with a number. For more on how deal scores work in HubSpot, read HubSpot deal scoring explained. The wider idea, using your own wins to decide who you sell to, is what we call ICP intelligence.
If you'd like to see what your won deals have in common before you open the spreadsheet, the free ICP report does a first pass from a HubSpot connection or a CSV.
