What a sales funnel audit for B2B actually checks (and what it should hand you)

What a sales funnel audit is actually checking

A sales funnel audit is not a review of your CRM software choice, your ad creative, or your sales team's talk track. It is a structural check of five things: whether every stage in your pipeline has a written qualification criterion, whether the conversion rate between each pair of adjacent stages is actually measured, where the steepest drop-off sits relative to a reasonable benchmark for your sales motion, whether the marketing-to-sales handoff has a defined speed and information standard, and whether lost deals are tagged with a specific stage and a specific reason rather than left blank. Most companies that believe they already know where their funnel is broken are working from instinct or the loudest complaint in the last sales meeting, not from this kind of stage-by-stage data. The audit exists to replace instinct with a number.

Step one: map the funnel as it actually runs, not as it is configured

Every CRM has a list of deal stages. Very few of those lists reflect what genuinely has to be true for a deal to be in that stage. A common finding in this first step is a stage like "Proposal Sent" that a deal enters the moment a document goes out, regardless of whether the buyer has confirmed budget, timeline, or decision authority. That gap between the stage name and the real qualification bar is where forecasts get inflated and where a close rate under 10 percent with healthy lead volume usually traces back to. Mapping the funnel as it actually runs means interviewing whoever moves deals between stages and writing down the real trigger for each move, then comparing that against what the CRM stage name implies. The distance between those two things is usually the first real finding of the audit, before a single number gets pulled.

Step two: pull the actual stage-by-stage conversion rates

Once the real stages are mapped, the audit pulls the conversion rate between every adjacent pair over a trailing period, typically the last two to four quarters depending on deal volume and cycle length. This is different from looking at how many deals sit in each stage today, which is a snapshot, not a rate. The rate answers a different question: of the deals that entered this stage, what percentage advanced to the next one, and how long did that take on average. Most teams have never seen this number broken out stage by stage; they have a top-of-funnel lead count and a bottom-line close rate, with no visibility into which of the stages in between is actually responsible for the gap. The steepest drop in this data, not the stage everyone assumes is the problem, is where the audit should focus the rest of its time.

Step three: check the handoff, because that is where the data usually breaks

The marketing-to-sales handoff, the moment a lead crosses from marketing-qualified to sales-accepted, is disproportionately where funnel data quality collapses. Marketing reports one lead count delivered; sales reports a much smaller number of opportunities accepted, and the gap between those two figures is rarely explained anywhere in the CRM. The audit checks three specific things here: whether there is a written, shared definition of what qualifies a lead for handoff, whether an SDR task or notification fires automatically within a defined window of that qualification event, and whether the leads that get silently dropped between the two counts are visible anywhere as data rather than disappearing. A funnel that looks fine at the top and fine at the bottom but has an unexplained gap in the middle almost always traces back to this handoff being undefined rather than to a genuine lead quality problem.

Step four: check whether lost deals actually explain themselves

The last diagnostic check, and the one most CRMs fail outright, is whether a lost deal record tells you anything useful. In a properly instrumented funnel, every closed-lost deal carries the specific stage it died at and a specific, categorized reason: price, timing, competitor, no budget, went dark. In most CRMs, the "reason" field is either empty, filled with a generic note, or not required at all, which means the business has no aggregate view of why deals are actually being lost. This is a cheap fix, a required field and a short, enforced list of reason categories, but it is the single highest-leverage change most audits recommend, because six months of that data turns "deals are stalling" into a specific, actionable pattern.

Sales funnel audit checklist: the steps in order

Use this as a working checklist for a funnel audit, whether you run it internally or brief someone else to. The steps are in the order they depend on each other: there is no point pulling conversion rates before the stages mean something.

1. Agree the scope and the question. Pick the funnel (one product, one market, inbound or outbound) and write down the question the audit must answer, for example "why did pipeline grow but bookings did not?"
2. Export the raw data. Every lead, contact, opportunity and closed deal for the trailing two to four quarters, with created date, stage history, source, owner, amount and close reason.
3. Write down every stage and its entry criterion. For each CRM stage, record what must be true for a deal to enter it, as the team actually works, and note where the name and the real trigger differ.
4. Check data completeness. Count records missing source, amount, owner or stage dates. Fields that are mostly empty cannot support a conclusion, and the gaps are a finding in themselves.
5. Calculate stage-to-stage conversion rates on cohorts. Of the leads or deals that entered each stage in the period, what share reached the next stage, and how long did it take? Split by source and segment.
6. Compare against your own history and a fair benchmark. Use your previous periods first, then a published benchmark with a matching definition, such as the B2B funnel conversion rate benchmarks.
7. Measure the handoff. Time from MQL to first sales touch, the share of MQLs sales accepts, and how many leads disappear between the two counts.
8. Review stalled and aged deals. List open deals that have sat in one stage far longer than your median time in that stage, and check whether each still has a next step and a date.
9. Categorise lost deals. Group closed-lost deals by stage and reason, and flag the share with no reason recorded.
10. Size each leak. For the two or three weakest stages, estimate the revenue a realistic improvement would add, using your own average deal value and win rate.
11. Prioritise and assign owners. Rank the fixes by revenue impact and effort, pick the first one, give it an owner and a review date.
12. Set the re-measure date. Agree when the same conversion table will be pulled again so the fix is judged on data, not opinion.

If the audit points to CRM structure, routing or reporting as the root cause, a RevOps audit goes deeper into the systems layer. If the leak is in how deals are qualified and worked, that is the work of a sales funnel consultant.

What a completed audit should hand you

A finished sales funnel audit should produce four concrete outputs: a stage-by-stage conversion rate table with the steepest drop identified and quantified, a written comparison between what your CRM stages are named and what actually triggers movement between them, a specific finding on the marketing-to-sales handoff with a measured or estimated speed-to-lead number, and a categorized breakdown of lost-deal reasons from whatever historical data exists, however incomplete. If an audit hands you a generic list of best practices instead of these four things tied to your actual data, it has not done the diagnostic work, it has done a template exercise. The value of an audit is entirely in how specific its findings are to your numbers, not in how comprehensive its advice sounds in the abstract.

Where this fits relative to fixing it

An audit is a diagnosis, not a rebuild. It should tell you precisely which stage to fix first and why, with the data to back that prioritization, but running the audit does not itself install lead scoring, rewrite your CRM's stage architecture, or build the handoff automation. Those are the next phase of work, and they are worth doing in the order the audit surfaces, fixing the steepest, most specific leak first rather than rebuilding the whole funnel at once. A good diagnostic makes that next phase faster and cheaper because it removes the guessing about where to start, which is usually the most expensive part of a funnel rebuild done without one.