A two-and-a-half-year GTM transformation at a healthcare software platform, read through the bow tie: every stage measured as volume × conversion = output. Figures shown as rates and improvements — the shape of the change, not the book of business. Notation: S = stage, V = volume, CR = conversion rate, and V(n+1) = V(n) × CR(n).
Win count itself grew faster (roughly 4.6×, shown above as 1→4), but average deal size normalized down over the same window, so revenue per unit of demand landed at 3.8×, not 4.6×.
| Lever | Change | Verdict |
|---|---|---|
| Left wing — acquisition | ||
| Raw demand volume | +75% | up, but noisy |
| Lead validity | −38 pts | declined — routing fix identified |
| Qualified lead volume | +23% | improved |
| Lead → buyer-qualified | +11 pts | improved |
| Buyer-qualified → opportunity | +21 pts | improved |
| Win rate | +11 pts | improved |
| Sales cycle | −55% | improved |
| Deal size (new-logo, won) | −2% | roughly flat |
| Bookings (like-for-like) | +29% | improved |
| Right wing — value realization & recurring | ||
| Value realization (signed → live & using) | +6 pts | improved |
| Time to install | −77% | improved |
| Churn + contraction volume | −39% | improved |
| Gross revenue retention | +5 pts | improved |
| Expansion rate | +1 pt | the opportunity |
| Net revenue retention | +6 pts | improved |
Every conversion gate on both wings improved — demand qualification, opportunity creation, win rate, cycle time, value realization, retention. Expansion is the exception: essentially flat for three years while the installed base grew, expansion still seller-led rather than triggered by product signals. To be precise, not zero movement, the retention-versus-expansion split within NRR moved about a point, confirming this has overwhelmingly been a retention story so far, not an expansion one. That makes expansion the highest-leverage move on the board: each point of expansion rate drops straight into net revenue retention with no new demand spend and no further retention work required.
"Triggered by product signals" is the fix, made concrete: usage instrumentation at the value-realization and retention stages (login frequency, feature adoption, booking-volume trend, the same signals proposed for Nuitee's demand motions), feeding proactive nurture instead of a rep waiting for a renewal call. A voice-based account-management layer, the same conversational-agent pattern already in production elsewhere, turns a usage dip into a same-day conversation instead of a quarter-end surprise.