How To Diagnose a Leaky SaaS Funnel

The VP of Sales pulled up the CRM and pointed at the pipeline. “We have 400 open opportunities,” he said. “We should be closing 80 of those this quarter.” By end of quarter, they closed 31. Not because the sales team was bad. Because the other 369 opportunities were not real.

Some were leads who had filled out a form but had no buying authority. Some were contacts from accounts where only one person was engaged and the rest of the buying group had never heard of the product. Some had been sitting in “Negotiation” for four months because nobody had the courage to mark them dead.

A leaky funnel is not a funnel with too few leads. It is a funnel where value disappears between stages, and the team cannot tell where it goes or why. Diagnosing it requires looking at account-level progression, not lead counts, because lead counts will always tell you the pipeline is healthy right up until the quarter ends and the number does not show up.

Define the funnel you are actually trying to measure

Most companies have a funnel in their CRM. Few have a funnel that means the same thing to every team.

Start by naming the stages and making sure the definitions are shared across marketing, sales, and finance. Traffic. Engagement. Qualification. Pipeline. Revenue. Each stage should have a clear entry criterion and a clear exit criterion. “Qualified opportunity” should mean the same thing in the CRM as it does in the board deck.

6sense’s pipeline measurement research pushes companies toward buying-group coverage as a qualification criterion, not just individual contact activity. An account with one engaged contact is not at the same stage as an account with four engaged contacts across different roles. If the CRM treats them the same, the pipeline forecast is fiction.

The second step is to stop reporting by lead count. A funnel report that shows 10,000 leads, 2,000 MQLs, 500 SQLs, and 100 opportunities looks healthy. But if the conversion from MQL to SQL is 25 percent and the conversion from SQL to opportunity is 20 percent, and those rates vary wildly by segment, the aggregate view is hiding the problem. The aggregate says “we have a pipeline.” The segment view says “we have a pipeline in one segment and a desert in three others.”

Find the leak by segment, not in aggregate

The single most useful diagnostic move is to break every funnel metric by segment.

Segments that matter: ICP tier (are tier-one accounts converting differently from tier-three?), channel (does paid search produce different pipeline quality from events?), product line (do enterprise accounts convert differently from mid-market?), ACV band (do larger deals stall at a different stage?), and geography (does one region outperform another for reasons the team has not examined?).

This analysis almost always reveals that the funnel is not universally leaky. It is leaky in specific places, for specific segments, at specific stages. One channel might produce excellent top-of-funnel volume but terrible qualification rates. One ICP tier might convert well through demo but stall at proposal. One product line might close fast in one region and slowly in another.

The segmented view turns “our funnel is broken” into “enterprise accounts from paid search stall between demo and proposal at twice the rate of enterprise accounts from events.” The second statement is diagnosable. The first is not.

A practical tool here is a segmented funnel audit sheet: a table with rows for each segment and columns for each stage conversion. Colour-code the cells (green for above-benchmark, amber for at-benchmark, red for below). The red cells are where the investigation starts.

Measure stage speed and stage quality

Conversion rate tells you how many accounts move between stages. Stage speed tells you how long it takes. Both matter, and they tell different stories.

An account that moves from demo to opportunity in five days is different from an account that moves from demo to opportunity in forty-five days. The conversion rate might be the same, but the speed difference usually indicates a difference in buyer readiness, deal quality, or sales process effectiveness.

Four speed metrics to track: time from engagement to first conversation (how quickly the team responds to buying signals), time from MQL to SQL (how long qualification takes), time from meeting to qualified opportunity (how efficiently discovery happens), and time from opportunity to close (how well the sales process moves deals through evaluation and negotiation).

6sense and Gartner both emphasise the importance of timely and relevant engagement. When buyers are in-market, they move through their evaluation quickly. If the company’s response process is slow (days instead of hours between signal and contact), the buyer has already progressed past the stage the sales team is trying to engage them at.

Speed-to-lead analysis often reveals the single biggest leak. If the median time from form-fill to first outreach is 48 hours, and the win rate for leads contacted within 2 hours is three times higher than leads contacted after 24 hours, the leak is not in the funnel. It is in the response process.

Check buying-group depth

Single-threaded deals are the quiet killer of B2B pipeline.

Ask a simple question for every tier-one opportunity: how many relevant stakeholders are engaged? If the answer is one, the deal is fragile. When that one person goes on leave, changes their mind, gets overruled, or leaves the company, the opportunity dies because no other relationship exists to sustain it.

6sense’s buying-group research shows that B2B purchases involve multiple decision-makers, and that most of them never fill out forms. A pipeline built on single-contact deals will always have a leakage problem that looks like “deals go dark” or “opportunities that seemed strong suddenly went cold.”

The fix is to measure buying-group coverage as a pipeline quality indicator. For each target account, identify the key roles (economic buyer, technical evaluator, champion, influencer) and track engagement across all of them. Set a minimum coverage threshold for advancement to the next stage. An opportunity with one engaged contact stays in early-stage until additional stakeholders are engaged.

Check messaging consistency and value clarity

Gartner’s buyer research found that 69 percent of buyers reported inconsistency between a company’s website messaging and what the sales team communicated. That inconsistency creates friction. The buyer arrives at a conversation expecting to discuss the value proposition they read on the website, and the sales rep tells a different story.

The diagnostic is a value-clarity checklist across touchpoints. Compare the website promise, the ad copy, the SDR’s opening email, the AE’s discovery questions, and the demo narrative. Are they telling the same story? Are they using the same language? Are they emphasising the same benefits?

If the website promises “reduce sales cycle by 30 percent” and the AE’s discovery call focuses on “improve team collaboration,” the buyer experiences a disconnect. They came for one thing and are being sold another. That disconnect creates doubt, and doubt slows deals.

Walk through the buyer journey from the buyer’s perspective: see the ad, visit the landing page, request a demo, receive the SDR’s email, attend the demo, review the proposal. At each step, ask whether the message is consistent and whether the value is clear. Every point of inconsistency is a potential leak.

Check data cleanliness and routing

Duplicate data, incomplete fields, and poor source mapping create invisible leaks.

Salesforce’s 2026 State of Sales research found that manual errors, duplicate contacts, and data trapped in silos directly cost companies revenue. When two sales reps have the same account because the CRM has duplicate records, the buyer gets contacted twice with conflicting messages. When source mapping is inconsistent, the team cannot tell which channels produce revenue and the budget conversation becomes a guessing game.

A practical “fix first” list for the first thirty days: deduplicate contacts (most CRMs have built-in or plugin deduplication tools), standardise required fields at each stage (if a field is not required, it will not be filled in), fix routing rules to ensure leads go to the right rep within the agreed SLA, audit source mapping to ensure every channel and campaign is tagged consistently, and reconcile CRM definitions with the definitions used in board reporting.

This work is not exciting. It will not appear in a marketing case study. But it is the work that makes every other investment in the funnel perform better, because the team can finally see the truth and act on it quickly enough to matter.

FAQs

What causes a leaky SaaS funnel?

Five structural issues cause most funnel leakage: poor ICP fit (attracting the wrong accounts), weak buying-group coverage (deals built on a single contact), slow or low-quality follow-up (missing the window when the buyer is active), inconsistent messaging across touchpoints (confusing the buyer), and dirty data or broken routing (losing leads in the system). Most companies have a combination of two or three.

How do I find where my funnel is leaking?

Break every funnel metric by segment (ICP tier, channel, product line, ACV, geography) and by stage. The aggregate view hides the problem. Look for specific segments where conversion between specific stages is significantly below benchmark. Then diagnose the cause: is it ICP mismatch, weak buying-group coverage, slow response, messaging inconsistency, or data problems?

What is buying-group coverage and why does it matter for pipeline quality?

Buying-group coverage measures how many relevant stakeholders within a target account are engaged in the sales process. B2B purchases involve multiple decision-makers, and deals built on a single contact have higher failure rates. Measuring buying-group coverage as a pipeline quality indicator correlates with higher win rates, faster deal velocity, and more accurate forecasting.

The VP of Sales from the opening eventually ran the segmented audit. The leak was in two places: enterprise accounts from paid search stalled between demo and proposal because the messaging on the landing page did not match the AE’s discovery narrative, and mid-market accounts across all channels were single-threaded with no buying-group engagement beyond the initial form-fill. Two fixes, one in messaging alignment and one in multi-threading process, reduced pipeline leakage by enough to close the quarter within 10 percent of target. The funnel was never broken everywhere. It was broken in two places, and the segmented view made those places visible.