Where sales reports lie (and how to make them stop)

Sales reports lie in five predictable places: pipelines stuffed with dead deals, close dates set to the last Friday of the quarter, stage definitions nobody agreed, activity counted instead of progress, and numbers compiled by hand the night before the meeting. None of these is a dishonest person at work — they are structural faults. Fix the structure and the same CRM, the same team, and the same meeting start producing numbers you can act on.

Why do sales reports lie when nobody is lying?

Because a report can only be as honest as the definitions and data underneath it, and in most 5–50-staff firms neither was ever designed. The pipeline was set up in an afternoon, stages were named by whoever installed the CRM, and the weekly numbers are assembled by a person under time pressure who knows roughly what the MD wants to see. Every small distortion is individually defensible — "that deal might still come back", "they said probably this quarter" — and collectively they produce a report that describes the pipeline as everyone wishes it were.

This matters because the report is the instrument panel. A management dashboard built on distorted inputs — the opposite of the live, self-compiling panel described in The MD Dashboard Blueprint — will steer hiring, spending and pricing decisions in the wrong direction, confidently.

What are the five most common lies?

1. The immortal deal. Deals that died months ago but were never closed as lost. They inflate pipeline value, flatter coverage ratios, and make the future look funded when it is not.
2. The hopeful close date. Close dates set to the end of the current quarter because the field was mandatory, then rolled forward each time the quarter ends. The forecast inherits the optimism.
3. The vague stage. "Qualified" or "Negotiation" defined by feel rather than by a verifiable event. Two reps with identical deals will place them two stages apart, and conversion rates between stages become noise.
4. Activity dressed as progress. Calls made and emails sent are inputs, not outcomes. A report weighted towards activity rewards motion and hides a stalled pipeline.
5. The hand-compiled number. Anything assembled in a spreadsheet the night before the meeting carries transcription errors, judgement calls and, occasionally, quiet editorial. Hand-compiled numbers also arrive too late to change anything.

How do you make the reports stop lying?

The fix is a mechanism, not a memo, and it runs in this order:

1. Define stages as verifiable events. When a discovery call has happened and is logged, then the deal is in Qualified; when a proposal document has been sent, then it is in Proposal. If you cannot point to the event, the deal does not move. Stage design of this kind is covered in pipeline stages that mean something.
2. Give every stage a time-out. When a deal exceeds the normal age for its stage with no logged activity, then it is flagged; when it exceeds a further grace period, then it is closed as lost with a reason code. Immortal deals stop accumulating.
3. Make close dates earn their keep. When a close date passes without the deal closing, then the deal owner must re-date it with a note — and the count of re-dates is itself reported. Serial optimism becomes visible.
4. Report outcomes first, activity second. The headline numbers are deals created, deals won, value won and conversion by stage. Activity appears only as a diagnostic beneath them.
5. Remove the human compiler. Reports draw live from the CRM or not at all. When a number cannot be produced automatically, then the fix is to repair the data, never to bridge the gap by hand.

When those five rules run, then the report stops being an account of hope and becomes an account of record — usually uglier, and immediately more useful.

Isn't this just a people problem?

Mostly not, and treating it as one makes it worse. Berating the team for pipeline accuracy produces a fortnight of compliance and then decay, because the effort fights a continuous flow of change with a one-off push. The durable version is automated: scheduled checks that flag stale deals, missing fields and duplicates every morning, as described in data-hygiene robots. People respond to what the system inspects, and a robot inspects daily without getting bored or diplomatic.

There is also a self-interest case worth making to the MD. When the reports lie, the correction happens in the founder's head — mentally discounting the pipeline, second-guessing the forecast — which is unscalable and exhausting. That quiet role of human error-corrector is one of the patterns examined in The Founder-as-Bottleneck Report.

What does an honest report look like?

Shorter and blunter than the current one. Pipeline value falls, often sharply, the first time the immortal deals are cleared out. Conversion rates become stable enough to compare month to month. The forecast narrows from a wish to a range. And the weekly meeting changes character: less narration of numbers, more decisions about the exceptions the system has already surfaced. Marketing numbers deserve the same treatment as sales numbers, and the pragmatic version of that is set out in attribution for small firms.

The test is simple. If the MD reads the report cold, with no verbal commentary, do they reach the same conclusions as the person who owns the pipeline? When the answer is yes, the reports have stopped lying.

---

Next step: the Growth System Audit — £450, seven days, credited against any build — maps where your growth system leaks, including where your reporting misleads you, and what to build first.