A CRM with defined stages doesn’t mean deals are moving through them. In most B2B pipelines, two or three stage transitions hold most of the latency — lead-to-meeting conversion is lower than it should be, proposals sit unanswered longer than the cycle time allows, and forecast accuracy stays low because time-in-stage varies too much to model. The bottleneck is usually identifiable with data, but the data is rarely normalized or examined stage by stage.
We pull CRM data, normalize it, and run a stage-by-stage throughput analysis alongside qualitative interviews with reps and ops. We’re looking for where deals stall, why (automation gaps vs. process gaps vs. enforcement gaps), and what a realistic time-in-stage target looks like for each transition. The output is a bottleneck heatmap, a fix list mapped to specific stages, and KPI targets designed to feed directly into implementation sprints. This diagnosis is only useful if there’s existing pipeline volume to analyze — a thin pipeline produces inconclusive data.
The diagnostic typically finds two categories of problem that operators can’t easily see from inside the system. The first is structural: stage definitions that made sense at 20 deals a month stop working at 80, because the implicit rules that sales reps use to move a deal forward were never codified into the CRM — so “moved to proposal” means different things to different reps, and the heatmap reflects that variance as noise rather than a real bottleneck. The second is automation coverage: stages where there’s no triggered next step, so momentum depends entirely on rep discipline. The combination — ambiguous definitions plus no enforcement — is where the most time-in-stage latency accumulates.
We need a CRM data export or API access, stage history for at least 90 days of deals, and time with three groups: the ops lead who owns the CRM configuration, two or three reps who work deals through the stages daily, and the sales leader who reviews the forecast. The rep sessions are the most important qualitative input — they surface the workarounds, the stages reps skip, and the reasons deals get parked that don’t show up anywhere in the CRM data. A diagnosis based on data alone, without those conversations, produces fixes that address the symptom rather than the cause.
Occasionally the data pull reveals that the CRM isn’t being used consistently enough to produce reliable throughput numbers — stages are too coarse, timestamps are missing, or the deal population in the export is too small. When that happens, we say so and reframe the engagement: the fix plan shifts toward establishing the measurement foundation first, so a re-run of the diagnosis in 60–90 days can produce actionable numbers. The bottleneck heatmap and KPI target sheet in the deliverable are designed for the ops lead and sales manager — they’re operational documents, not exec summaries. The most common next step is directly into a conversion recovery or deal desk automation sprint, targeting the two or three transitions the diagnosis identified as highest leverage. Teams that enforce the time-in-stage targets from the deliverable typically see forecast accuracy improve meaningfully within two quarters, simply because the definitions become consistent.