Short answer: the deck and the export usually disagree for definitional reasons, not arithmetic ones. They were pulled at different times, with different filters, from different stage and date definitions, and sometimes from different systems. Six reconciliations find almost every gap: timing, scope, stage, amount, dates and duplicates.
For CFOs, finance leaders and fractional CFOs who have to sign off on a pipeline number they did not build.
Why the two numbers drift
The pipeline figure in a board deck is a report someone built once, with filters they chose then. The CRM export is raw records as they stand today. Between the two sit a dozen small decisions nobody wrote down: which stages count as pipeline, which currency, which close-date window, whether renewals are included, whether a deal pushed to next quarter is still “this quarter’s pipeline” in the deck.
None of that is fraud or incompetence. It is what happens when a number is defined in a report instead of in a document. The fix is to reconcile the two line by line and write the definition down.
The six reconciliations
1. Timing: when was each number pulled?
A CRM changes every day. A deck built on the 3rd and an export pulled on the 20th will differ by every deal created, moved or closed in between. Ask for: the report’s run date and the export date. If the deal-level history is available, re-run the report as of the deck date.
2. Scope: which deals are in?
Report filters silently exclude things: a pipeline, a business unit, a record type, a deal owner who left, deals below a minimum amount. Ask for: the report’s filters, written out. Then filter the export the same way and compare counts before comparing values.
3. Stage: what counts as pipeline?
One team counts everything from “qualified” onwards; another starts at “proposal”. Some decks weight by stage probability; exports show unweighted amounts. Ask for: the list of stages included and whether the figure is weighted. If it is weighted, get the probability per stage and check that it matches the CRM’s settings.
4. Amount: which number is “the amount”?
Deals can carry several money fields: amount, annual contract value, total contract value, amounts in different currencies with conversion rates set at different times. Line items may not add up to the deal amount. Ask for: which field the deck uses, the currency conversion date, and a count of open deals with no amount at all. Deals with a blank amount inflate count-based metrics and deflate value-based ones.
5. Dates: close dates and created dates
A pipeline window depends on close date, and close dates in most CRMs are edited by hand. Deals whose close date is in the past but are still open are the classic source of a “phantom” pipeline. If a CRM sync or a migration touched the records, created dates may reflect the import, not when the opportunity really began. Ask for: open deals with a close date in the past, and a check on whether created dates cluster on a migration or import date.
6. Duplicates and the second system
When marketing automation and the CRM sync, or when opportunities exist in two systems, the same deal can appear twice, or appear in one system with different values. Ask for: the system of record for pipeline, and a count of deals that share a company and a close month with a near-identical amount.
A worked reconciliation, step by step
Take the deck number and the export. Apply, in order: the deck’s date, then its filters, then its stage list, then its amount field and currency, then remove open deals with past close dates, then remove duplicates. After each step, record the new total and the difference. The table you end with is the bridge between the two numbers, and each line of it is either a definition to write down or a data problem to fix.
What to do with the result
Write a one-page pipeline definition: stages included, weighting, amount field, currency rule, date window, system of record. Get sales, marketing and finance to sign it. Then ask for the board report to be rebuilt from that definition, not the other way round.
If the reconciliation turns up data problems (duplicates, missing amounts, stage skips, a sync overwriting values), those sit in the CRM and the systems feeding it. That is what our read-only GTM Systems Audit looks at: three weeks, from USD 1,500, every finding with the export behind it. Finance leaders can start with our page for finance leaders, the five numbers to pull before trusting the forecast, or send one deals export for a free 48-hour pipeline teardown.
FAQ
Why doesn’t the pipeline in our board deck match the CRM?
Usually because the two were built with different definitions: different run dates, report filters, stages counted, amount fields, close-date windows and duplicate handling. Reconciling those six items in order explains almost every gap and shows which differences are definitions and which are data problems.
What should a CFO check before trusting the CRM pipeline?
Open deals with close dates in the past, deals with no amount, stage definitions and weighting, the amount field and currency used, and duplicates between systems. Each can move the pipeline number materially without anyone changing a deal.
Should pipeline be weighted by stage probability in board reporting?
Either can work, but the deck must say which. If weighted, the probabilities per stage should be written down and match the CRM settings. Mixing weighted and unweighted figures across quarters makes trends meaningless.
Who should own the pipeline definition?
Finance, sales and marketing jointly, with one written definition signed by all three and one named system of record. Reports should be rebuilt from the definition, not the definition inferred from an old report.
How long does a pipeline reconciliation take?
For one quarter and one CRM, a finance or ops analyst can usually bridge the two numbers in a day or two using the six steps. Fixing the underlying data problems takes longer and depends on what the reconciliation finds.