Forecast from CRM history

Two weeks into every quarter, Tallowtree's reps commit to a number. Leadership asked for a second read from the CRM's own history to sit next to it.

A commit is the number reps are sure of, so it's built to come in low. Over the last 7 quarters it landed under what was won every time, by 24% on average. The model aims for the middle instead. Built only from what the CRM showed on the same day, it missed by 11% on average. For Q4 2026 the model says $642K, likely between $454K and $850K. The commit is $446K.
11%model's average miss
24%commit's average gap to what was won
$642KQ4 2026 model call

How the model works

For every open deal on day 14 of the quarter, the model reads two things from the CRM: the furthest stage the deal has reached, and how many of three buyer-evidence fields are filled (budget confirmed, economic buyer met, mutual close plan). It looks up how often deals in that same position won within the quarter, in earlier quarters only, and multiplies by the deal's annual value. It then adds the average value of deals that were created after day 14 and still won in the same quarter, over the last four quarters.

Furthest stage, because deals get moved backward when a rep tidies up, and the furthest point tells you more than where a deal sits today. Buyer evidence, because a deal at Proposal with a confirmed budget and a met economic buyer is a different deal from one with neither.

The rates it uses for Q4 2026, learned from every quarter before it:

Furthest stageNo evidence1 field2 fieldsAll 3
Solution27%
59 deals
16%
91 deals
36%
33 deals
23%
0 deals
Proposal33%
18 deals
46%
37 deals
71%
41 deals
67%
15 deals
Negotiation59%
13 deals
53%
40 deals
65%
34 deals
59%
7 deals

Gray cells have fewer than 15 deals behind them, so the model uses that stage's overall rate instead. Some neighboring cells don't rise in order. That's real noise from small groups. I leave it in, because smoothing it would hide how thin the data is.

Seven quarters, tested blind

Each quarter's model only saw what was in the CRM on its own day 14, with rates learned from earlier quarters. Nothing from the quarter being called leaks in.

Each quarter: what was won, against the two calls made on day 14
$0$300K$600K$900K$1.2MQ1 2025: won $606KQ1 2025: commit $425KQ1 2025: model $614K Q12025Q2 2025: won $860KQ2 2025: commit $541KQ2 2025: model $700K Q22025Q3 2025: won $945KQ3 2025: commit $621KQ3 2025: model $1.01M Q32025Q4 2025: won $960KQ4 2025: commit $688KQ4 2025: model $789K Q42025Q1 2026: won $768KQ1 2026: commit $766KQ1 2026: model $776K Q12026Q2 2026: won $896KQ2 2026: commit $589KQ2 2026: model $666K Q22026Q3 2026: won $752KQ3 2026: commit $727KQ3 2026: model $787K Q32026Q4 2026: 80% range $454K to $850KQ4 2026: commit $446KQ4 2026: model $642K Q42026

Won Model, day 14 Reps' commit, day 14 Model range, 80%

QuarterWonModelMissCommitMiss
Q1 2025$606K$614K1%$425K-30%
Q2 2025$860K$700K-19%$541K-37%
Q3 2025$945K$1.01M7%$621K-34%
Q4 2025$960K$789K-18%$688K-28%
Q1 2026$768K$776K1%$766K0%
Q2 2026$896K$666K-26%$589K-34%
Q3 2026$752K$787K5%$727K-3%

Does a 40% deal win 40% of the time?

A forecast can be right in total and wrong deal by deal. This check groups every back-tested deal by the chance the model gave it and counts how many actually won.

Model's chanceDealsAverage chance givenShare that won
0% to 5%3154%3%
5% to 20%24811%13%
20% to 40%5530%35%
40% to 60%5651%45%
60% to 100%7870%56%

The low bands are close. The top band is too generous: deals the model gave 60% or more won less often than that. On a client's data I'd find out why before trusting the top of the range. Deal size is the first thing I'd check.

Q4 2026, as of October 15

Furthest stageOpen dealsAnnual valueModel expects
Solution20$564K$134K
Proposal7$235K$145K
Negotiation8$212K$116K
Discovery and Qualified68$184K
Not created yet$62K
Total103$642K

The range comes from playing the quarter 4,000 times: each open deal wins or loses at its own rate, plus one of the last four quarters' late wins. In 80% of those runs the quarter landed between $454K and $850K.

What it can't know

The model is a second opinion beside the reps' call, and the useful part is where the two disagree. Those are the deals to talk about in the forecast meeting.

What you get

A forecast from your CRM history, for your company

I build and back-test this on your CRM data. Fixed price $2,500, scoped in writing before I start. Hire me on Upwork.