Turned a 400-lead Pipedrive account into a forecast the founders trust for a B2B corporate training company

Client snapshot

A corporate training company with two founders and three account managers, selling leadership and compliance workshops to mid-sized companies in Germany and Austria. Deals range from a single workshop to year-long programmes. Pipedrive had been in use for two years, set up by whoever needed it first.

The problem

The account held about 400 leads and deals. Some sat in the Leads Inbox, some in a pipeline with eleven stages that meant different things to different people. "Proposal sent" was used both for a rough price by email and for a signed-off offer. Fewer than half the open deals had a value, and about a third had an expected close date, so Pipedrive's forecast showed whatever happened to be filled in. Each quarter the founders rebuilt the forecast in a spreadsheet by asking every account manager what they thought would close.

What we built

We started on paper, not in Pipedrive.

1. Stages with exit criteria. Eleven stages became six: Qualified, Needs confirmed, Offer sent, Negotiation, Verbal yes, Won. Each stage has a one-line rule for when a deal may move into it, for example "Offer sent" means a written offer with a price has gone to the decision maker.
2. A field model. We agreed which fields matter for the forecast: deal value, expected close date, decision maker, programme type and lead source. Everything else was archived or merged. Status and lost reasons became fixed dropdowns instead of free text.
3. Required fields per stage. Pipedrive asks for the value and decision maker before a deal can enter "Offer sent", and for the expected close date before "Negotiation". Nobody can move a deal forward with the forecast fields empty.
4. Stage probabilities. Each stage got a win probability based on the last 18 months of won and lost deals, so the weighted value in Pipedrive means something.
5. Lead cleanup. All 400 records were exported to a Google Sheet, sorted into keep, merge and archive with the founders, missing fields filled where the history allowed, and re-imported with Pipedrive's merge on ID. Duplicate organisations were merged.
6. The forecast view. An Insights report shows weighted revenue by expected close month, per account manager, next to won revenue. A small Pipedrive automation creates a follow-up activity when a deal's expected close date passes without a stage change, so stale dates get fixed instead of sitting in the forecast.

We agreed in the first session what "done" means: every open deal in the new stages with all forecast fields filled, and one weekly review run from the Insights report instead of the spreadsheet.

Results, one quarter in

• Open deals with both a value and an expected close date: about 35% → all of them.
• Quarterly forecast against actual revenue: off by about 40% → within about 10%.
• The weekly pipeline review: around 90 minutes → around 30, run straight from Pipedrive.
• The quarterly forecast spreadsheet is gone.

Quote

"For the first time the number in Pipedrive is the number we plan with." Co-founder

Build snapshot

• Timeline: 2 weeks, about 16 hours
• Stack: Pipedrive, Pipedrive Insights, Pipedrive Automations, Google Sheets
• Investment: $720
• Who owned it: one founder owns the stage rules and reviews the stage probabilities each quarter.

The pattern

A forecast is only as good as the stage definitions and the fields behind it. Most CRM forecasts fail because stages mean different things to different people and the close date is optional. If your pipeline grew by itself, the fix is the same: agree the rules first, make Pipedrive enforce them, then clean the data once.

---

This is an illustrative composite drawn from operator engagements. Names, identifying details, and exact metrics are anonymised. The workflow, stack, and approach are real.