How to run a sales pipeline review. The method that catches stalled deals before forecast day.

The short answer. What a good pipeline review actually checks.

Salesforce's State of Sales report puts the scale of the problem plainly: only around a third of sales professionals trust the accuracy of the data in their own forecast. A pipeline review exists to fix that, deal by deal, before the number gets reported upward. It is not a status meeting where each rep talks through their week. It is a structured check of stage accuracy, next-step existence, close-date credibility and how many deals have gone quiet.

Done properly, a review answers four questions for every open deal: is it in the right stage, does it have a genuine next step with a date, is the close date backed by an actual buying signal, and how long has it actually sat there. Skip any one of the four and the pipeline report you present in a forecast call is a guess dressed up as data.

None of this needs to be complicated or software-heavy. A basic filtered view sorted by days-in-stage does most of the work; the discipline is in asking the same four questions of every deal, every week, rather than only the ones that happen to come up in conversation. Teams that skip the quiet, unglamorous deals in a review are usually the ones surprised when a forecast slips.

Weekly versus monthly. Two cadences with different jobs.

Most B2B organisations run a weekly or biweekly deal-level cadence between a rep and their manager, supported by a monthly strategic review that pulls in RevOps or sales operations. Confusing the two is a common reason reviews feel unproductive. The weekly review is diagnostic: it checks individual deals and fixes what it finds on the spot. The monthly review is structural: it looks at coverage ratios, win rates and stage conversion trends across the whole team, and asks why a pattern is happening rather than fixing one deal.

Running only the weekly cadence means systemic problems, a stage that consistently leaks deals, a segment with chronically low win rates, go unnoticed because no one steps back to look at the whole pipeline rather than one rep's slice of it. Running only the monthly cadence means individual stalled deals sit for weeks before anyone with the authority to act on them even sees the detail. Mature teams run both, each with a clear, separate purpose.

Step by step. A repeatable weekly method.

1. Pull the report before the meeting, not during it

Export or open the pipeline report sorted by stage and days-in-stage before the review starts. A review that begins with "let me just pull this up" burns ten minutes of a 30-minute slot on nothing, and it signals that the review is not really prepared for. Send the report to the rep beforehand too, so they arrive already aware of what is likely to come up rather than being surprised by their own pipeline.

2. Flag deals stalled past the stage norm

Every stage has a typical duration. If a discovery call usually leads to a proposal within two weeks and a deal has sat in discovery for six, that deal is stalled, not slow. Flag it before the meeting so the conversation starts with "what's happening here" rather than discovering the problem live. Build the stage norms from your own historical data over a full quarter, not from a generic benchmark, since a complex enterprise sale and a fast transactional deal have entirely different rhythms.

3. Check every deal has a genuine next step

"Follow up" with no date attached is not a next step, it is a placeholder that lets a dead deal sit in the pipeline looking active. Every open deal needs a specific action and a date. If a rep cannot name one on the spot, that is itself useful information about how live the deal really is. Deals with no answer to this question in two consecutive reviews are usually better marked lost, freeing the rep's attention for deals that are genuinely moving.

4. Interrogate close dates against buying signals

Ask what evidence supports this month's close date: a signed scope, a confirmed budget conversation, a procurement step already underway. A close date based on the rep's optimism rather than a real signal should move, and moving it now costs far less credibility than missing it in the forecast call. This single check is usually where the gap between a forecast and reality starts, because it is the easiest place for hope to substitute for evidence.

5. Calculate coverage against target

Add up open pipeline value and divide by the remaining target for the period. Most B2B teams need three to four times coverage to hit target reliably, once normal win rates and slippage are accounted for. A ratio below that most months is an early warning that needs a pipeline-generation conversation, not just a forecasting caveat. Track this number over several reviews rather than reacting to a single month, since one strong or weak week can distort a snapshot without reflecting a real trend.

6. Assign named actions with dates

Close the review with actions, not observations. "This deal needs a champion conversation by Thursday, owned by Priya" survives until the next review. "This deal looks a bit stuck" does not. This is also where clear stage definitions with real exit criteria earn their keep, because everyone in the room agrees on what "stuck" actually means. Log the action in the CRM itself, not in a separate notes document, so the next review can check it against the record rather than against someone's memory of the conversation.

A worked example. A twenty-deal pipeline, one Tuesday morning.

A rep walks into their weekly review with 20 open deals worth £480,000 against a £120,000 remaining target, a coverage ratio of exactly 4x on paper. Sorting by days-in-stage flags three deals: one has sat in "proposal sent" for five weeks against a typical two, one has a close date this month with no scheduled next step, and one is a genuinely live deal simply waiting on a client's board date.

The manager asks about the stalled proposal first. The rep admits the champion has gone quiet, no one has spoken to them in three weeks. That deal gets downgraded and its close date pushed, dropping real coverage to 3.1x, still workable but tighter than the headline number suggested. The deal with no next step gets one assigned on the call: a call booked for Thursday, owned by the rep, logged before the meeting ends. The board-date deal needs nothing except a note to check back after the date passes.

The remaining sixteen deals get a faster pass, since none are flagged and most are early stage. The manager spends thirty seconds each confirming the next step date still looks realistic, rather than working through them at the same depth as the three flagged deals. That triage, depth where it is needed, speed where it is not, is what keeps a twenty-deal review inside its thirty-minute slot instead of sprawling into an hour that neither person has.

Twenty minutes in, the pipeline number the team walks away with is smaller than the one they walked in with, and considerably more honest. That is the entire point of the exercise: a review that never changes the number it started with usually is not looking hard enough.

Common mistakes. Where reviews turn into status theatre.

Treating every deal the same regardless of size or stage. A twenty-deal pipeline does not need twenty minutes spent equally. Spend most of the time on deals that are large, at risk, or new since the last review, and cover the rest in a quick pass.

Running the review as an interrogation. A rep who feels cross-examined starts managing the review rather than the pipeline, softening bad news or hiding stalled deals until they are unavoidable. The best pipeline reviews I run feel like joint problem-solving, not an audit.

No follow-up mechanism. If last week's flagged actions are not the first thing checked this week, the review teaches the team that flags do not matter, and the same deals stay stalled for months.

Running the review off a spreadsheet because nobody trusts the CRM. This is the most common failure I see, and it is rarely a software problem. It is usually an adoption problem: the data is thin because the team was never given a working reason to keep it current. Fixing that is a sales dashboard leadership can actually trust, built on data the team enters because the review depends on it, not because a policy says they must.

Reviewing win rate and coverage in isolation from each other. A healthy coverage ratio built on a win rate that has quietly dropped from 25 per cent to 15 per cent is not actually healthy, it just has not caught up yet. Check both numbers together, not as separate line items on different reports that nobody cross-references.

None of this requires new software or a bigger CRM licence. It requires a manager who runs the same six steps every week, and a team that trusts the review is there to unblock deals rather than catch people out. Get the method right and the pipeline review becomes the single most useful half hour in a sales manager's week, the point where forecast accuracy either gets built deal by deal or quietly erodes unnoticed until the number leadership was promised does not show up.