Pipeline Hygiene: The Monthly Practice That Prevents Forecast Chaos

Most B2B pipeline forecasts are wrong by meaningful margins. A team forecasts £800k in closable pipeline for the quarter; actual close comes in at £520k or £1.1M. The variance isn't due to sales team quality or bad luck. It's due to underlying pipeline data that wasn't maintained honestly in the months before the forecast was run. Deals were sitting at stages they shouldn't have been at. Probabilities were optimistic rather than calibrated. Deals that had effectively died were still showing as active. By the time the forecast runs, the data is so noisy that the forecast is essentially guesswork with a spreadsheet.

The fix isn't better forecasting tooling. It's maintaining pipeline honestly through the months leading up to the forecast. This takes one 90-minute meeting per month, run with specific discipline. Teams that install it produce forecasts that are reliable within 10-15%; teams that skip it produce forecasts that swing by 30-50% from what actually happens. This article is about the specific practice that produces the first outcome.

Why Pipeline Decays Without Maintenance

Three specific ways pipeline data drifts between month-end reviews.

Deals stick at stages longer than they should. A deal moved to "Proposal" stage six weeks ago and hasn't progressed. The salesperson hasn't formally closed it because they believe it might still come back. The deal sits there, still counted in pipeline, still weighted at the stage's probability. Multiply this pattern across a sales team and pipeline fills with deals that aren't really active but haven't been formally recognised as dead.

Close dates get pushed quietly. A deal's expected close date was end of month; the salesperson quietly updates it to end of next month. Then to end of the quarter after. Each individual push seems small. Cumulatively, pipeline accumulates deals whose close dates have been pushed three or four times without anyone explicitly acknowledging that the original forecast assumption is broken.

Deal probabilities don't reflect reality. A deal's stage implies 40% probability. The salesperson privately knows this specific deal has bigger problems than the stage probability implies. They don't adjust the probability down because doing so visibly reduces their forecasted pipeline. The private knowledge stays private; the forecast uses stage probabilities that don't match reality.

Each of these is a small individual distortion. Collectively they produce pipeline data that's unreliable by the time the forecast runs.

The 90-Minute Monthly Practice

The maintenance ritual that keeps pipeline honest. Runs once a month, involves the Head of Sales (or founder running sales) and usually the Head of Growth or RevOps person. 90 minutes, structured around three specific reviews.

Review 1: Stale Deals (30 minutes)

Pull every deal that's been in the same stage for more than 30 days. For each one, force a binary decision: confirm still active with specific next step, or mark closed-lost.

"Still active with specific next step" means a concrete action with a date, "prospect is reviewing proposal with CFO, call scheduled 12th." Not "following up" or "keeping warm." An actual next step that would produce movement.

"Closed-lost" means the deal is out of pipeline. It can come back later if circumstances change; it's not dead forever. But it's not counted in current pipeline any more.

The third option, "keep in pipeline at current stage, no specific next step", isn't available in this review. That's the option that creates the pipeline pollution problem. Forcing every stale deal into one of the two binary outcomes cleanses pipeline of the accumulated drift.

Review 2: Pushed Close Dates (30 minutes)

Pull every deal whose expected close date has been pushed twice or more. For each one, honest diagnosis: is this deal actually going to close or has the motion genuinely stalled?

A deal pushed once from month-end to the following month is usually fine, small timing slippage that happens normally. A deal pushed twice or more has usually lost momentum that isn't coming back without specific intervention. The second push is the signal to investigate rather than accept.

The investigation produces one of three outcomes. Confirm the deal is still closing with a specific reason the original dates slipped (champion went on holiday, legal review took longer than expected, specific reasonable cause) and a concrete new close date. Downgrade the deal to an earlier pipeline stage reflecting its actual progress. Or mark it closed-lost because the motion has clearly stalled and pretending otherwise is noise in the forecast.

Review 3: Probability Calibration (30 minutes)

Pull the top 10-15 deals by size, weighted by current stage probability. For each one, the sales rep running it describes the deal's actual current state, what's working, what's at risk, what the prospect is signalling.

The question at the end: "given what you just described, does the stage probability match reality?" Often the answer is no. A deal nominally at 60% probability might realistically be 30% because of a specific concern the salesperson hasn't formally surfaced. A deal at 25% probability might realistically be 50% because a specific signal has shifted.

The output is a manual adjustment to the probability of these top deals. Smaller deals run on stage defaults because the effort doesn't justify individual calibration. Top deals get explicit calibration because they disproportionately drive the forecast.

Why This Doesn't Take Longer

90 minutes sounds constrained. It's deliberately so.

The practice isn't meant to surface every pipeline issue in depth. It's meant to remove the obvious distortions that otherwise accumulate. Depth-review on individual deals happens in regular sales meetings and deal strategy sessions. The monthly hygiene practice is a lighter touch that specifically addresses the three drift patterns above.

Longer meetings usually produce less effective hygiene because they lose focus and get into discussion of deal tactics rather than structured review. 30 minutes per review category, sharp binary decisions, move on. The output is clean pipeline data, not a comprehensive strategy session.

What This Produces

Teams that run this monthly consistently produce three observable outcomes over 3-6 months.

Forecast accuracy improves to within 10-20% of actual results. The noise from stale deals, pushed dates, and miscalibrated probabilities has been systematically removed.

Sales team discipline improves. Knowing that stale deals will be forced into a binary decision monthly, salespeople don't let deals drift as passively. They either progress deals or close them lost earlier, which produces cleaner pipeline naturally.

Leadership trust in pipeline reporting increases. When the forecast has been reliable for three consecutive months, conversations about pipeline shift from "is this number real?" to "what should we do about this number?" The productive question.

What Not To Do

Three specific mistakes in implementing this practice.

Running it as a group meeting with the whole sales team. The hygiene review is for pipeline quality, not team development. Group format adds hours and reduces the directness needed for binary decisions. Run it as a small leadership review (2-3 people) and communicate outcomes back to the team.

Skipping it when the month is busy. The practice matters most when the team is busiest, because busy months produce the most drift. Skipping a cycle because "we don't have time this month" produces the noise it was designed to prevent.

Using it to challenge individual salespeople's judgement in detail. The practice focuses on systemic patterns (stale deals, pushed dates, probability miscalibration) rather than interrogating specific deal decisions. Salespeople should feel the practice as protective of forecast quality, not as surveillance of their individual deal management.

Where This Sits in the NBT Growth Model

Pipeline hygiene is a CRM & Revenue Ops capability in Stage 3 Navigation, with strong connections to Stage 4 Optimisation because reliable pipeline data is prerequisite for effective quarterly business reviews and forecasting.

The connection to CRM data hygiene is direct. General CRM hygiene covers the broader data quality discipline; pipeline hygiene is the specific high-leverage ritual for the pipeline-data subset. The connection to full-funnel management is also close; reliable pipeline data is what makes the marketing-to-sales handoff visibly accountable rather than theoretical.

If your pipeline forecasts swing more than you'd like from actual results, or you're setting up sales operations and want to install the hygiene practice from the start, get in touch for a strategy call. We'll walk through the specific implementation for your stage and team.