Competition hub
GAWhy deals are won and lost · per-deal AI reasoning on every battlecard, interview-driven decision factors, and review-mined switching.
What it is
Win/Loss Analysis turns closed deals into a repeatable competitive learning loop, showing why deals are won and lost across outcomes, reasons, competitors, segments, and buyer interviews. It connects to the rest of the platform: buyer interviews from Engagement feed your decision factors and head-to-head competitor ratings; the feature gaps that recur in lost deals surface in Gap Analysis; the most-cited loss reasons become Roadmaps priorities; and lost revenue by competitor informs Portfolio Management investment decisions.
See it in the product
Win/Loss Analysis
A live capture from the production workspace lands here once Win/Loss Analysis ships its public preview cohort.
Top 4 capabilities
Each capability is the parent of dozens of typed features in the production taxonomy. Hover any feature in Studio to drill into the underlying nodes.
Every rate is empirical-Bayes shrunk, and slices under ten decided deals are flagged as low sample. A guard test fails the build if new code computes a raw won-over-total rate without going through the shared function or being explicitly allowlisted with a reason. The statistics are enforced, not merely intended.
The write path refuses a loss without a governed code, so analytics never run over free text, and codes map onto decision drivers separately from the outcome class — keeping what happened and why it happened as two axes rather than one muddled field.
The interview runs without a login, and on conclusion the win/loss analysis is written and the deal is stamped as responded so outreach stops chasing it. Prioritisation goes never-asked first, then reachable, then newest — which fixed a starvation where unreachable recent deals consumed every capped run.
Themes are evidence-gated: they may cite only deals actually fed in, with no invented frequency. First-party win/loss is never bridged into the cross-tenant signal corpus — it reaches competitive surfaces tenant-scoped, so your losses sharpen your battlecards without ever informing someone else's.
More from the Competition hub
Live competitive cards assembled from the graph on demand. No 60-90 day staleness.
Competitor hiring intelligence. Job-posting trends, skills-gap analysis, position-level drilldowns.
Upload an RFP/RFI, auto-extract every requirement, draft cited responses in under 5 hours.
Score vendor products against extracted requirements and taxonomy features on the same 10-dimension rubric as battlecards · review or AI signals.
Every buying opportunity as one record — the pipeline row, the vendor matrix and the RFX workspace are three views of it, with participants stored as graph nodes rather than names.
Public-data 360 profile of a competitor across products, reviews, talent, customers, and signals.
30-minute walkthrough. We'll filter the workspace to your category and walk through the top capabilities live.
