// GuidesHow-to
Pipeline hygiene: the 8 metrics, a weekly checklist, and what to automate (2026)
Short answer
Pipeline hygiene is how closely the deals in your CRM match what buyers are actually doing. A clean pipeline has a real close date, an amount, a specific next step, recent buyer activity and more than one contact on every open deal, and dead deals are closed out. Measure it with eight metrics (pipeline coverage, stale deals, close-date pushes, deals with no next step, single-threaded deals, stage-time breaches, slip rate and missing required fields), run a 20-minute weekly pass on the exceptions per rep, and have reps fix them async before the pipeline review so the review is about deals rather than data entry. Automate detection and the Monday send; keep the buyer-reason judgment and close-lost decisions human.
What is pipeline hygiene?
Pipeline hygiene is the gap between what your CRM says and what your buyers are doing. Every open deal makes claims (this close date, this amount, this stage, someone working it), and hygiene is how many of those claims are still true. A pipeline is clean when a stranger could read it and reach the same forecast you would.
It is a manager’s problem before it is a rep’s problem. Nobody logs the deal that quietly died, because nothing happened, and “nothing happened” is what a CRM cannot show you. So hygiene is measured as exceptions (deals that break a rule), per rep, weekly. As Sybill puts it, “The single biggest pipeline review failure is reviewing deals based on data that does not exist.”
- Not activity volume. Ten logged calls on a dead deal are still a dead deal.
- Not a quarterly clean-up. Cleaned once a quarter means dirty for eleven weeks of it.
Watch out
The mistake most teams make
Treating hygiene as a discipline problem. A team-wide “please update your CRM” updates the deals reps remember; the dead ones stay untouched because nobody is thinking about them. Hygiene moves when exceptions are listed by deal, per rep, with the rule each broke.
Which 8 metrics measure pipeline hygiene?
Eight numbers cover it. Six are counts rather than percentages, because a percentage hides which rep and which deal, while three stale deals with names attached is a Monday conversation; all six come from CRM reports you can pull this afternoon. Coverage and slip rate are ratios, computed weekly for the team and once a period per rep.
| Metric | Definition / formula | Healthy range or rule | Why it matters |
|---|---|---|---|
| Pipeline coverage | Open pipeline ÷ quota for the period. Required coverage = 1 ÷ historical win rate (landbase). | 60% win rate: 1.7×. 25–40% (mid-market): 2.5–4×. 15–25% (enterprise): 4–7×. Far above your number usually means dead deals. | Dead deals inflate coverage and hide a real shortfall. |
| Stale deals | Open deals with no logged buyer-facing activity for N days. | A rule of thumb, adjusted to your data: 7 days for a sales cycle under 30 days, 14 days for 30 to 90 days, 21 days for over 90. Committed deals get a tighter clock; orm-tech drops one from commit after fourteen days without meaningful activity unless the rep shows buyer engagement. | The commonest way a forecast lies. Nothing happened, so nothing was logged, so it still counts. |
| Close-date pushes | Times the close date moved later this period, per deal, from field history. | A workable flag is two pushes in a period, or any push without a buyer-side reason. orm-tech requires a documented buyer-side event before an in-period close date. | The clearest leading indicator of a slip; forecastio lists repeated close-date changes among common risk signals. |
| Deals with no next step | Next-step field empty, undated, or a placeholder such as “follow up”. | Zero past first qualification. nbh’s test: is it defined, or just a vague “follow up”? | No next step, no plan. Also the fastest field to fix, so start here. |
| Single-threaded deals | One associated contact, or one buyer-side person replying in the last 30 days. | For example, two engaged contacts by proposal stage, plus someone who can sign on deals above your median size. | One contact going quiet stalls the deal while the CRM still shows it healthy. |
| Stage-time breaches | Days in the current stage versus the median for deals your team has won. | A workable rule is 1.5× the won-deal median for that stage; refresh the medians quarterly. | Deals well past the norm mostly lose, and the breach says where. |
| Slip rate | Value pushed past the period ÷ value dated to close in it at period start, as a percentage (orm-tech). | No universal number; track yours quarter over quarter and make it fall. orm-tech’s worked example is 50%, and across its customers only about 20% of day-one value closes inside the quarter. | The scoreboard for the other seven. If it is not falling, the rules are not being followed. |
| Missing required fields | Share of open deals past qualification with any required field empty (close date, amount, next step, key contacts). | Zero. Enforce by stage in the CRM; forecastio’s rule is to define what must exist at every stage and enforce it. | An empty amount is a zero in the forecast; a guessed one is a wrong number nobody questions. |
Which fields must always be filled?
Four fields you enforce, two your CRM fills in. forecastio’s list for every stage is close date, deal amount, next step and key contacts; the CRM adds stage-entry date and last activity date as long as reps move stages and activity is synced. Make each required at the stage below, in the CRM itself. That costs the rep ten seconds at the moment they know the answer; a reminder costs you a week.
The must-fill fields, and the stage that makes each one mandatory
- Close date, from a buyer-side event (signature date, budget cycle, go-live). Required from qualification. A date with no event behind it is hope, and hope slips.
- Amount, a specific number from a quote or scoped proposal. Required from qualification. nbh’s tell for a guess is a round number.
- Next step, dated, with a named buyer-side person. Required on every open deal, at every stage.
- Key contacts, at least the champion and the signer, associated to the deal (a name in a note cannot be reported on). Required by proposal.
- Stage and stage-entry date, automatic if reps move the stage when the buyer moves, and never to dress a report.
- Last activity date, automatic if calls, emails and meetings are synced. If they are not, every stale-deal count is a guess.
The manager’s weekly hygiene checklist, by rep
Twenty minutes on Monday morning. This pass counts exceptions per rep and decides who gets which list. The fixing happens later, by the reps.
Weekly hygiene pass (20 minutes)
- Pull the six exception reports by owner: stale, pushed twice or more, no next step, single-threaded, stage breach, missing fields. (5 min)
- Write each rep’s count into the scorecard, next to last week’s. (3 min)
- Open the three largest deals on the list. Any in commit tops the review agenda whether or not the rep fixes it. (5 min)
- Mark pushes that already carry a buyer-side reason; those are deal problems and go to the review as deals. (2 min)
- Send each rep their own list, with the rule each deal broke and a deadline (next section). (4 min)
- Log the team total and the trend. A falling total goes to your VP; a rising one on a single rep goes to the 1:1. (1 min)
| Rep | Stale | Pushed ≥2 | No next step | Single-threaded | Stage breach | Missing fields | Total (last week) |
|---|---|---|---|---|---|---|---|
| Rep A | 3 | 1 | 4 | 2 | 1 | 0 | 11 (14) |
| Rep B | 0 | 2 | 0 | 1 | 2 | 1 | 6 (5) |
| Rep C | 5 | 3 | 2 | 3 | 2 | 2 | 17 (9) |
| Team | 8 | 6 | 6 | 6 | 5 | 3 | 34 (28) |
Read it across, then down. Across, Rep C’s row says the pipeline has stopped being maintained, a workload conversation before it is a performance one. Down, “no next step” as the team’s biggest column is a process gap you fix once, with a required field.
How do I run hygiene without nagging reps? Fix it async before the review
Give each rep their own list, the rule each deal broke, and 48 hours. Nagging is what happens when the ask is vague (“update your CRM”) and public (the whole team). A named list with a deadline is neither. Sybill puts combined prep for a 30-minute review at 45 to 90 minutes, much of it on data reps could have fixed on Monday.
- Monday, 9:00. Run the weekly pass and export each rep’s exceptions with the rule next to each deal.
- Monday, 9:30. Send each rep their list privately, in one message, due Wednesday noon, with three allowed answers per deal: fixed, closed out, or a buyer-side reason in the note.
- Tuesday and Wednesday. Reps fix in the CRM. A push needs the buyer-side event in the note. Closed-out deals come off the pipeline and coverage drops to its true number. Let it.
- Wednesday, 13:00. Re-run the six reports. What survives, plus every deal that got a buyer-side reason, is the review agenda.
- Thursday, the review. No field gets edited in the meeting. Each surviving deal leaves with a decision (fix now, close lost, or the manager takes an action). Sybill’s 30 minutes and 3 to 5 priority deals only work if this step already happened.
- Friday. Log this week’s counts next to last week’s. Say so when a count fell. Take a rising count to the 1:1 with the deals attached, and ask what got in the way first.
Tip
The Monday message, in full
“Three deals on your list this week. Acme (no buyer activity in 24 days), Birch (close date pushed twice, no reason in the note), Cedar (no next step). By Wednesday noon each is fixed, closed out, or has a buyer-side reason in the note. Thursday’s review covers what is left.” Same wording every week, so it reads as process rather than mood.
Watch out
Never send the team’s list to the team
A public list turns hygiene into a scoreboard, and reps game the fields (a fake next step, a close date parked on the last day of the quarter). Private lists, public trend. The team total can go on the wall; the names stay in the 1:1.
What should I automate, and what stays human?
Automate detection and delivery; keep judgment and consequences human. Every metric here can be detected by a scheduled report or a field rule. galvintech describes a Salesforce “Clean Your Room” dashboard built on exactly these lists (deals pushed after a set number of days, deals open too long, activities not followed up) to hold teams accountable for keeping deals up to date.
| Task | Automate or human? | How |
|---|---|---|
| Detecting stale deals | Automate | A scheduled report on last-activity date past your threshold, by owner. Nothing fires when nothing happens, so it runs on a schedule. |
| Pushes, stages moved backward, amount changes | Automate | Field history on close date, stage and amount, as a weekly list or an alert. |
| Required fields by stage | Automate | Make the field mandatory when a deal enters the stage. |
| Per-rep exception counts and trend | Automate | One saved dashboard by owner, with last week’s number beside this week’s. |
| Sending each rep their list | Automate the send, keep the wording yours | A fixed template, the same every Monday, delivered privately. |
| Whether a pushed date has a real buyer reason | Human | Read the note. With no event behind the date, it goes back to the rep instead of into the forecast. |
| Closing a deal lost | Human | The rep proposes; above your median deal size the manager decides, with the reason recorded. |
| A rising count on one rep | Human | Bring the deals to the 1:1. Workload, territory and one stalled big deal look identical in the count. |
| “What changed this week?” | Automate the numbers, keep the story | Weekly snapshots of open pipeline, coverage and slip rate; you say what they mean. |
How diffi helps
diffi connects to Salesforce and HubSpot read-only, so it does not write to your CRM and your required-field rules stay where they are. It derives facts from what changed (a close date pushed or slipped a quarter, a stage that went backward, an amount change, a task done late and, from Salesforce, a deal with no change for 21 days) and turns them into deal and pipeline risk signals with the evidence behind each. You can ask in plain language for open pipeline, win rate, sales cycle length or what changed this week; answers link to their sources. When a rep’s exceptions need fixing, diffi can send that rep a Slack message on your behalf once you confirm it, so the fix happens before the review rather than in it. The exception lists come from your saved CRM reports; diffi adds the change facts and, once you confirm, the message to each rep. Book a demo to see the weekly pass run on your own pipeline.
See it on your own teamHow does hygiene change forecast accuracy?
Every hygiene defect is a forecast error that has not shown up yet. A forecast sums, over deals dated to close in the period, amount times the probability the stage implies. A stale deal adds an amount that will not close. A date that should have been pushed again lands in the wrong period. An empty amount is a zero, and a stage nobody moved back carries the wrong probability. As forecastio puts it, “Without reliable data, you cannot trust reports or sales forecasting”.
orm-tech reports that across its customer base only about 20% of the pipeline value dated to close in a quarter on day one actually closes inside it. Slip rate is the one number that says whether your hygiene rules are working, because pushing is what a dirty pipeline does when the period ends.
Hygiene also fixes the inputs to coverage. If dead deals are never closed lost, your win rate is computed on too few losses and reads high, so 1 ÷ win rate understates the coverage you need, while the same dead deals inflate the coverage you appear to have. One missing habit, closing out what has died, causes both errors. Fix it weekly and the two numbers tell the truth together.
Frequently asked questions
What is pipeline hygiene?
Pipeline hygiene is how closely the deals in your CRM match what buyers are actually doing. A clean pipeline has a real close date, an amount, a specific next step, recent activity and more than one contact on every open deal, and deals that have died are closed out. It is measured as exceptions per rep, counted weekly.
Which fields must always be filled in the CRM?
Close date, amount, next step and the key contacts, enforced by stage rather than by reminder, plus the stage-entry date and last activity date, which the CRM fills in when activity is synced. Make close date, amount and next step required at qualification, and contacts required at proposal.
How many days without activity makes a deal stale?
As a rule of thumb, 7 days without logged activity for a sales cycle under 30 days, 14 days for a 30- to 90-day cycle, and 21 days for a cycle over 90 days; adjust to your data. Committed deals get a tighter clock; a common forecast-call rule is fourteen days without meaningful buyer activity.
How do I run pipeline hygiene without nagging reps?
Send each rep their own list privately on Monday, with the rule each deal broke and a deadline of Wednesday noon, and allow three answers per deal (fixed, closed out, or a buyer-side reason). Use the same wording every week. Whatever is still broken on Wednesday becomes the pipeline review agenda, so no field is ever edited in the meeting.
How does pipeline hygiene change forecast accuracy?
A forecast sums amount times probability over the deals dated to close in the period, and every hygiene defect corrupts one of those terms. A stale deal adds value that will not close, a pushed date lands in the wrong period, an empty amount is a zero and a wrong stage is a wrong probability. Clean the pipeline weekly and the error shows up as a slip rate you can measure and reduce, instead of as a surprise at quarter end.
What should a pipeline hygiene dashboard show?
Six exception counts by rep (stale, pushed twice or more, no next step, single-threaded, stage-time breach, missing required fields), each with last week’s number beside it, plus coverage and slip rate for the team. Build it as saved reports filtered by owner in Salesforce or HubSpot; the same six reports are the manager’s Monday pass.
Is pipeline hygiene the rep’s job or the manager’s?
Fixing is the rep’s job; measuring and deciding are the manager’s. Reps own the fields on their deals and the 48-hour fix. The manager owns the weekly count, the trend, the buyer-reason judgment on pushes, and the decision to close a deal lost.