Most sales pipelines have a leak somewhere. Leads enter at the top, but somewhere between first contact and closed deal, a significant portion disappear. The challenge isn’t that deals drop off — that’s expected. The challenge is not knowing where the leaks are or why they exist.

Funnel analysis in your CRM gives you the visibility to find those leaks and fix them systematically. When you can see exactly how many deals move from each stage to the next — and how many don’t — you can stop guessing about where your pipeline is broken and start making targeted improvements.

This guide walks you through how to build a funnel analysis in your CRM, how to interpret what you find, how to diagnose the root causes of drop-off, and how to measure whether your fixes are actually working.

What Funnel Analysis Is

Funnel analysis is a stage-by-stage conversion analysis of your sales pipeline. It answers: of all the deals that entered Stage X, what percentage moved forward to Stage X+1?

Done across every stage in your pipeline, it produces a conversion waterfall — a clear picture of where deals are progressing, where they’re stalling, and where they’re dropping out entirely.

The Difference Between Funnel Analysis and Pipeline Reports

A standard pipeline report shows you what’s in each stage right now — how many deals, total value, weighted value. Funnel analysis is different. It looks backward at a cohort of deals — all deals that were created or entered your funnel in a given period — and tracks what happened to each of them over time.

This cohort-based view is what makes funnel analysis actionable. Instead of asking “how much is in my pipeline today?” you’re asking “of the deals that entered my pipeline last quarter, how many made it to each stage, and how many are still active?”

Building Your Stage-by-Stage Conversion Analysis

Step 1: Define Your Stages Clearly

Before you can analyze stage-by-stage conversion, each stage needs a clear, consistent definition. If different reps use the same stage labels for different situations, your funnel data will be meaningless.

For each pipeline stage, document:

• What must be true for a deal to be in this stage
• What specific action or event moves a deal from this stage to the next
• Who is responsible for that progression

Stage	Entry Criteria	Exit Criteria (to next stage)
Lead	Contact created, source recorded	Meeting booked
Discovery	Meeting completed	Qualification criteria confirmed
Proposal	Proposal sent	Verbal agreement to move forward
Negotiation	Commercial terms in discussion	Decision-maker commitment obtained
Closed Won	Contract signed	Deal recorded as revenue

The more precise your stage definitions, the more reliable your funnel data.

Step 2: Choose a Cohort and Time Window

Funnel analysis requires you to pick a cohort — a group of deals defined by when they entered the funnel — and a time window long enough for most of them to have reached an outcome.

For a typical B2B sales cycle of 60 to 90 days, analyzing deals that entered the pipeline three to six months ago gives most of them enough time to close or disqualify. Using a too-recent cohort will show artificially high “still in progress” numbers that distort your conversion rates.

Step 3: Run the Stage-by-Stage Report

In your CRM, pull a report that shows, for your chosen cohort:

• How many deals entered each stage
• How many progressed to the next stage
• How many stalled or were lost at each stage
• The average time spent in each stage

Stage	Deals Entered	Progressed	Dropped	Stage Conversion Rate
Lead	500	200	300	40%
Discovery	200	120	80	60%
Proposal	120	72	48	60%
Negotiation	72	52	20	72%
Closed Won	52	—	—	—

Your overall pipeline conversion rate (Lead to Close) in this example is 10.4% — roughly 52 out of 500 deals that entered the funnel closed. But the stage-by-stage view tells you far more: the biggest drop-off is happening at Lead-to-Discovery (60% drop), which means the earliest stage of qualification is where most deals are dying.

Identifying Where Deals Stall

Drop-off can take two forms: exits (deals marked as lost or disqualified) and stalls (deals that remain in a stage far longer than average without progressing).

Stalled deals are harder to see in funnel analysis if you’re only counting exits. Include average stage duration in your analysis to surface them.

Spotting Stalls with Age Data

Calculate the average number of days deals spend in each stage for deals that successfully progressed. Then compare that to the current age of deals stuck in each stage. Deals significantly older than the average are stalled, not progressing.

Stage	Average Days (Progressing Deals)	Deals in Stage >2x Average
Discovery	14 days	12 deals
Proposal	21 days	8 deals
Negotiation	18 days	3 deals

The 12 deals stalled in Discovery beyond 28 days are high-priority: either they need an intervention to move forward or they should be disqualified and removed from the pipeline.

Diagnosing Root Causes of Drop-Off

Finding the drop-off point is only half the work. You need to understand why deals are dropping off at that stage if you want to fix it. Here are the most common root causes by stage type.

Drop-Off at Early Stages (Lead, Discovery)

Early-stage drop-off usually signals a qualification problem. Either you’re letting too many unqualified leads into the pipeline, or your qualification process isn’t effective at sorting fit from no-fit quickly.

Look at the attributes of lost deals: industry, company size, role of the main contact, lead source. If certain segments show consistently low progression rates, adjust your qualification criteria or targeting to reduce the inflow of those segments.

Drop-Off at Middle Stages (Proposal, Presentation)

Middle-stage drop-off often indicates a mismatch between what the prospect expected and what they received when they saw a proposal or presentation. Common causes:

• Proposals are generic rather than tailored to the specific problem
• Price shock — the prospect wasn’t prepped for the cost
• The wrong person is being pitched (decision-maker not involved)
• A competitor entered the conversation at this stage

Review call notes and deal close reasons in your CRM for lost deals that exited at this stage. Patterns in the notes are often the fastest way to identify the real cause.

Drop-Off at Late Stages (Negotiation, Closing)

Late-stage drop-off is the most expensive — these are deals where significant time was invested. Common causes include:

• Internal budget or approval process issues on the buyer’s side
• Competitor displacement at the last minute
• Deal champion losing influence internally
• Terms that couldn’t be bridged

For late-stage losses, conduct a brief post-mortem review. Even a few minutes of analysis per deal can reveal patterns. Log the reason for loss consistently in your CRM so you can analyze it across many deals rather than one at a time.

Testing and Measuring Fixes

Once you’ve identified a drop-off point and diagnosed a plausible root cause, you need to make a change and measure whether it worked.

The Right Way to Test Funnel Changes

Make one change at a time. If you simultaneously change your proposal template, add a new discovery call structure, and adjust your qualification criteria, you won’t know which change (if any) improved conversion.

Measure the impact over a realistic time window. For most sales cycles, you need at least one full cycle to pass after making a change before you can assess its impact — often 60 to 90 days.

Building a Funnel Improvement Baseline

Before making any change, document the current conversion rate at the stage you’re targeting. This is your baseline. After the change, run the same analysis for a comparable cohort and compare.

Stage	Baseline Conversion	Post-Change Conversion	Change
Lead to Discovery	40%	48%	+8 points

An improvement of 8 percentage points at Lead-to-Discovery, applied to 500 leads, means 40 additional deals entering Discovery — a meaningful downstream impact on pipeline.

What Counts as a Meaningful Improvement

Statistical significance matters here. A conversion rate improvement based on a small number of deals may be noise rather than a real signal. As a general rule:

• Changes based on fewer than 50 deals in each comparison group should be treated cautiously
• Changes that persist across two consecutive cohorts are more likely to reflect a real improvement
• Changes that hold across multiple reps or segments are stronger evidence of a real effect

Common Funnel Analysis Mistakes

Comparing snapshots instead of cohorts. Taking a picture of your pipeline today and comparing it to a picture from last month is not funnel analysis. You need to follow the same cohort of deals through time.

Not accounting for deal age. A deal that entered the pipeline two months ago and is still in Discovery is different from a deal that entered last week. Age data is essential for identifying stalls versus active deals.

Overreacting to short-term noise. Monthly fluctuations in conversion rates are normal. Look at quarterly or rolling averages to identify real trends rather than reacting to month-to-month variation.

Ignoring the “no decision” outcome. Many deals don’t end in won or lost — they end in “no decision” or “prospect went dark.” Track this as a distinct outcome in your CRM. High no-decision rates at specific stages often indicate a specific friction point worth addressing.

Frequently Asked Questions

How many pipeline stages should I track in funnel analysis? Track every stage that exists in your CRM pipeline, but focus your analysis attention on the transitions between stages, not the stages themselves. Most sales pipelines have 4-7 stages; funnel analysis is useful at any number. If you have too many stages (10+), consider consolidating — more stages create more complexity without necessarily more insight.

What CRM reports do I need to run funnel analysis? You need a report that shows deals grouped by their entry date (cohort), with progression status at each stage and the outcome of each deal (won, lost, still active). Most CRMs can generate this as a pipeline stage history report or a funnel report. The key requirement is that your CRM logs stage changes with timestamps — without stage history logging, you can only see current stage, not progression.

How do I handle deals that skip stages? Some deals move through multiple stages quickly (or sales reps record them all at once). This is normal. In funnel analysis, you can count a deal as having “passed through” a stage even if it didn’t spend significant time there. The more important thing is to flag deals that skipped stages and make sure it reflects real progression, not a data entry shortcut that masks a stage that was actually bypassed.

Should I segment my funnel analysis by rep, product, or deal size? Yes, if you have enough deal volume to support it. Segmented funnel analysis often reveals that a drop-off problem is concentrated in a specific rep, product line, or deal size — which makes the root cause diagnosis much more targeted. Start with aggregate analysis to find the biggest drop-offs, then segment to understand whether they’re universal or concentrated.