Win/Loss Synthesis
Your closed deals aggregated into the factors that move the win rate, plus the reasons your team actually recorded on won and lost deals.
Included with crocronjob . A seat without it sees an upsell here instead of the page.
Overview
Win/Loss Synthesis turns your closed deals into the small set of factors that actually move the win rate: the tailwinds to lean into and the headwinds to fix. Instead of a leader's hunch about why deals are won or lost, it is a deterministic count over the deals you closed, so it holds up in a board review and points at real levers.
What it is for, on the page. Under the headline, this page says what it produces, what makes it run, and where the output lands. Every Work page reads that line from one registry, so the three answers cannot drift from the page.
Factors and lift
For each deal attribute value, competitor, industry, segment, lead source, deal-size band, whether there was a champion, the synthesis computes the win rate at that value and its lift over your baseline win rate. A value well above baseline is a tailwind; well below is a headwind. Only values with enough supporting deals are shown, so a single fluke deal never becomes a "factor". Factors are ranked by impact, biggest and best-evidenced first.
Stated reasons
Separately, the synthesis tallies the win/loss reasons your team recorded, most common first, split into won and lost. This is the qualitative layer next to the quantitative factors: the factors tell you what correlates, the reasons tell you what your reps wrote down. Together they point at what to change in the motion.
Red-team a lost deal
Below the reasons, Lost deals, last 90 days lists each deal your CRM marked closed lost in that window. Red-team this loss asks a skeptical senior AE to read the deal's facts at close (its features, the last forecast before it closed, and the cohort patterns it matched) and write five hard questions you did not ask, two or three other explanations for the loss, the signals you ignored, and what a skeptical AE would have done at day 30. It uses only what was recorded on the deal and invents no stakeholders, dates or quotes.
Each red-team is one AI call, counted against your monthly AI budget and rate limited, and the result is saved on the deal. Opening the same loss again shows the saved version and costs nothing. A daily scan also red-teams up to five of your losses from the past week on its own, so a recent loss often opens at once. The list reads the closed deals the daily snapshot recorded, so a loss you closed today appears after the next daily run.
Where the data comes from
The synthesis reads your closed opportunities, their win/loss stage, categorical attributes, and stated reason, from your connected CRM and stored records. Only closed deals count; open deals have no outcome to learn from. If your deals are not tagged with attributes yet, the factors are sparse rather than invented, the more consistently your team tags and records reasons, the sharper the synthesis.