What a CRM tells you that a deal memo does not
The memo is the seller's story; the CRM is the company's diary. Nine things a four-quarter extract shows that no deck can, and what each means for the deal
By Semir Jahic
/ 10 min read
An information memorandum is written to be read once and believed. Every number in it was chosen, rounded and arranged by someone who wants the deal to close. The CRM was not written for you at all. It was filled in by sellers on Friday afternoons, one deal at a time, over years, and nobody tidied it for the sale. That is what makes CRM data in due diligence the most honest evidence in the data room. The memo tells you what management wants the pipeline to be. The records tell you what it is: how old each deal is, how often its close date moved, who carries the number and which customers stayed. Here is what a four-quarter extract shows that the memo cannot, and what each finding means for price, conditions and the first 100 days.
Why is the CRM better evidence than the memo?
Because the memo is an argument and the CRM is a record. A deck states 3.0 times pipeline coverage, a 40% win rate and 110% net revenue retention in three bullet points. The CRM holds every opportunity those numbers were built from, with the dates, amounts and owners that produced them. You recompute the claim instead of taking it.
Records are also hard to polish. A slide can be rewritten in an evening; four quarters of stage history and activity logs cannot be rewritten without leaving traces. What GTM due diligence checks explains how records and interviews are read together; this post stays with the records.
What does a four-quarter extract show that the memo cannot?
Nine things. For each, the memo makes a claim, the extract answers it, and the answer maps to price, a condition or the 100-day plan. Four of them (coverage, stage ageing, slipped deals and forecast against actual) form the forecast test covered step by step in an earlier post, so they get one line each.
| Test | The memo says | The data shows | What it means for the deal |
|---|---|---|---|
| Stage ageing and stale deals | "3.0 times coverage" | Open value idle for 60 days or far past the stage median of won deals | Coverage restated; price follows the real figure |
| Close dates pushed | "Strong second half" | Deals pushed three or more times; few closed in their original quarter | Forecast haircut; a condition that commit criteria are written |
| Win rate by segment | "Our ideal customer is mid-market" | 40% inside that segment, 22% outside it, most new logos outside | Growth planned where conversion is worst; re-base the model |
| Discounting | "Strong pricing power" | Discount spread by rep; last two weeks of the quarter against the rest | Revenue quality; a 100-day action on approvals |
| Attainment spread | "70% of reps hit quota" | Top three sellers' share of bookings; ramped reps at 100% | Key-person risk; retention terms for named sellers |
| Ramp of recent hires | "Productive in six months" | Months from hire to the first quarter at 80% of quota, by cohort | Plan capacity overstated; hiring plan re-based |
| Forecast against actual | "Accuracy above 90%" | Week-six forecast against closed, by quarter, with bias | Weight the next four quarters can carry; size of the earn-out |
| Pipeline created per month | "Marketing delivers demand" | Created pipeline by month and source against what the plan needs | Whether next year exists yet; a pipeline-generation condition |
| Retention cohorts | "Net retention 110%" | Cohorts recomputed from revenue by customer by month | Reported versus recomputed gap; the base case moves |
Three deserve a note. Win rate by segment is where the stated ideal customer meets the record: on the ten-area scorecard, 70% to 85% of new revenue from the stated profile supports a 4, and below 50% does not reach a 3. Attainment spread tells you whether the number is a team result or one person's; top three sellers above 70% of bookings means retention terms before any hiring wave. Retention is one of the memo's most quoted numbers: recomputed from revenue by customer by month, on one cohort, it becomes a figure you can test instead of one you were told.
What does this look like on one deal?
Illustrative, round figures. A target plans CHF 20 million of new business next year and presents 3.0 times pipeline coverage, a 40% win rate and net revenue retention of 110%. The extract says something else.
- Coverage: a fifth of open pipeline is past its close date and a further fifth has been idle for 60 days. Clean coverage is 1.8 times: below the 2.5 times a 40% win rate needs, far below the 4.5 times a 22% win rate needs.
- Win rate: 40% holds inside the stated segment, which produced a third of last year's new revenue. Outside it the win rate is 22%, and two thirds of new logos came from there.
- Attainment: twelve quota carriers, four at quota. The top three carried 72% of bookings. Reps hired in the last 18 months took ten months to reach 80% of quota; the plan assumed six.
- Pipeline created: even at the memo's 40% win rate, the plan needs about CHF 4 million of new pipeline a month. The trailing six months averaged CHF 2.6 million.
- Retention: on one cohort, net retention recomputes to 103%, not 110%, because the memo left churned customers out of the cohort.
Each line changes the deal. Clean coverage and the created-pipeline gap mean next year's number does not yet exist in the records: a price conversation, and a 100-day plan led by pipeline generation rather than hiring. Attainment concentration means retention terms for three people. The retention gap moves the base case. The verdict is Adjust: proceed only with answers, or protection in price and terms, on these points.
What should you ask for?
Four files, as they are: opportunities with full stage history for four quarters (eight for a Deal DD); forecast snapshots as they stood each week or month; quota and attainment per rep per quarter, including leavers; and revenue by customer by month. The data-room request list gives the fields, formats and fallbacks.
Ask for the records uncleaned, with native IDs kept so the files join, and a half-page note on what the company already knows is wrong with its data. A target that cleans before sending removes exactly what you are looking for. A target that will not send an extract at all has also told you something.
What does a clean extract look like versus a spreadsheet export?
A clean extract is a copy of the history, not a report of today's view. Stage changes come from the history table, one row per change with its date and user. Close-date and amount changes come from field history. Forecast snapshots are the pipeline as it stood on each date. Every file carries the same opportunity and account IDs.
"We export to a spreadsheet" usually means a report of open deals as they look today, and that view has no memory. A deal pushed four times shows one close date. The spreadsheet answers how large the pipeline is; only the history answers whether it is real.
| Question | Spreadsheet export | History extract |
|---|---|---|
| How often did this deal slip? | One close date, no answer | Every change, with its date |
| How long has it sat in this stage? | The stage today | An entry date for every stage |
| What did the forecast say in week six? | Nothing | The snapshot as it stood |
| Who changed the amount, and when? | Nothing | Field history, with the user |
If history is switched off, or was lost in a migration, the affected areas are capped and listed as not tested. That is a finding in itself, and a cheap 100-day action.
Why does this need no benchmark?
Because the test is internal consistency. The memo's claim and the company's records describe the same deals; when they disagree, that is a finding, whatever the industry median says. A 3.0 times coverage claim that becomes 1.8 times needs no peer group to be a problem. A 110% retention claim that recomputes to 103% is a finding about the memo, not about the market.
Benchmarks come later, for the ratios: whether a 22% win rate or a ten-month ramp is normal for a company of this size, stated against named public sources. The evidence that moves price is the gap between claim and record, and the company supplied both.
This is the method behind Deal DD for GTM: the extract, the interviews, and ten areas scored from 1 to 5, with every figure traced back to the rows that produced it. If you are inside an exclusivity window and nobody has opened the CRM yet, book 30 minutes with me and bring the memo's strongest number.
Key takeaways
- The memo is an argument; the CRM is a record. Recompute the claim instead of accepting it.
- One extract supports nine tests: stale pipeline, pushed dates, win rate by segment, discounting, attainment spread, ramp, forecast against actual, pipeline created and retention cohorts.
- A finding maps to price, a condition or the 100-day plan, or it is not a finding.
- Ask for history tables and snapshots, uncleaned, with IDs kept. A spreadsheet of today's pipeline has no memory.
- The gap between claim and record needs no benchmark. Benchmarks come later, for the ratios.
Frequently asked questions
What CRM data is needed for due diligence?
Opportunities with full stage history, forecast snapshots as they stood, quota and attainment per rep per quarter, and revenue by customer by month. Four quarters for a Health Check, eight for a Deal DD. The data-room request list on the toolkit page sets out fields, formats and fallbacks.
How long does the CRM analysis take?
Days one to three of a Deal DD, inside the fifteen working days. The target needs one to two hours of RevOps time to pull the files. Every headline figure is traced back to the records before it is used in an interview or a score.
What if the target has no stage history?
Then the forecast and pipeline areas are capped and listed as not tested. Monthly pipeline exports or board-pack pipeline pages are an acceptable fallback. The absence is itself a finding: management cannot audit its own forecast either, and switching history on is a first-week action.
Does the CRM replace interviews?
No. The records show what happened; interviews show why, and whether the people running the engine see it the same way. When a sales leader describes a reliable forecast and the snapshots show otherwise, that disagreement is the finding, and it goes into the report.
Can a CEO run this on their own company?
Yes. The GTM Health Check applies the same tests to a four-quarter extract in five working days, for CHF 8,000, with four to eight hours of the team's time in total. The verdict for your own company reads On track, Needs work or At risk.