EA Analytics · Case File No. 001
From gut feel to a system the team runs itself.
Series A, agentless IT infrastructure mapping, ~50 people. Three months of GTM analytics, shown below in the client's real numbers, shared with permission.
Series A, agentless IT infrastructure mapping, ~50 people. Three months of GTM analytics, shown below in the client's real numbers, shared with permission.
The sales team was prioritizing by hand. Two years of leads went through a scoring model built on closed-won evidence. This is what came out.
Most of the leads sit in tiers that barely convert. Click through the views. That flip is the whole business case.
665 leads got two years of calls and produced zero revenue. The top tier converts at 3.8x the average, and nobody knew it existed. The score now runs live in Salesforce; the second finding, a separate 10,000+ employee "whale" segment with year-long cycles, gets its own playbook.
Bigger deals were the company's stated goal when the engagement started. The scoring gave that goal teeth: targeting refined against the tiers that close big, and every demo priced so progress was visible. Quarter by quarter, the effort showed up in the data:
Mann-Whitney p = 0.009. Not noise.
Enterprise (10k+ employee) leads: ~11% of intake in 2025, 2-3x that through 2026.
The day a demo happens, it gets a win probability and a conservative ARR estimate. The month's pipeline value exists immediately, and reality fills in behind it. No waiting a quarter to find out whether marketing delivered.
Off by one win, with revenue landing higher because the deals grew.
The team quotes the number in real time, knowing reality lands above it.
Average ICP score of each week's qualified leads. In early July the line broke below its band. Flagged the same week. Two weeks later it was back in the zone.
The gold band is the funnel's normal zone (65–72). Without the weekly read, that July dip surfaces at the quarterly review, three months of budget later. With it, the correction happened inside a fortnight.
The engagement wrapped in the summer of 2026. The scoring runs live in Salesforce. The team's own monthly metrics are defined in the model's terms: "good leads" means the top two tiers, and pipeline targets use the demo-level estimates. The forecasts stay on record for a scheduled review.
Spend, quality, demos, pipeline, wins. Operational and cohort views, so B2B lag stops distorting the story.
Deliberately conservative. Deals close above the estimate, so the number is one the team commits to.
Lead-quality drift surfaces in weeks, not quarters. Exhibit D shows it working.
Every 10,000+ employee lead tracked from lead to demo to opportunity, on milestones that respect a year-long cycle.
A free 30-minute call is enough to tell whether your data can support the same kind of system.
Book a free data audit call