Stopgivingawaypricethatwinsyounothing.
Pricekeel reads your closed deals and shows where discounting buys wins, where it just gives money away, and the discount that earns the most on the next deal. No warehouse, no integration. One CSV.
What you see when you click
The diagnostic, on a live sample of 2,000 closed deals
Below is the actual structure of the page on /sample: KPIs, win-rate curve with the calculated win point, and the deal list to investigate. Your numbers come from your own CSV.
Live numbers from the bundled sample. Your numbers come from your CSV.
What it does · what you walk away with
Four answers, four artifacts you can forward unchanged
Each pillar produces a deliverable a Head of Pricing forwards to the CFO as-is. No dashboard to interpret, no copy-paste cleanup.
Find the leakage
Price realization, win rate by discount band, and three views of discount leakage, from a plain description to the strongest claim worth acting on.
You get · Executive summary
A one-page narrative your CFO reads in two minutes: booked value, price realization, win point, the headline upside, and the top three deals to investigate. Plain English, no jargon.
Find the win point
The discount level where win rate stops improving. Everything given beyond it is a list to investigate, not a refund.
You get · Win-curve report
Win rate by discount band with the calculated reference point, the band where bigger discounts stop buying wins. The chart your Head of Pricing forwards to the CRO.
Guide the next discount
A win-probability model recommends the discount that maximizes expected value on a given deal, with a plain-language why.
You get · Per-deal guidance
For each open opportunity: recommended discount, expected-value lift, win-probability change, and the top three factors the model thinks drive the deal. Decision support for the deal-desk call.
Defend the decision
Every recommendation is logged with the math behind it. The Copilot answers CFO questions with citations to the source signal, with no LLM-invented numbers.
You get · Defended-vs-investigate list
Won deals split into 'discount earned the win' (defend in sales review) and 'discount worth a retrospective look' (route to deal-desk debrief). Sales-friendly framing of the same data.
No signup. The math is shown.
How much discount are you giving back a year?
= (18% − 15%) × list. This is the deal-desk lens. It is correlational, not a refund: some of that discount was needed to win. The real diagnostic clips every deal against your own win-rate curve, so it usually finds more than a blended average shows.
Built on named methodology, not vibes
Defensible to finance because the math has authors
Every Pricekeel signal (leakage lenses, win point, packaging signal, trade-or-give, decision log) is grounded in a published framework your Head of Pricing can cite when defending a discount approval.
Built for the stack you already run
A CSV export covers every CRM and billing system today; native connectors are on the roadmap with the margin layer.
Margin Enhancement
The planned next layer: connect contracts and CRM to surface margin across the whole book, including special pricing agreements, fixed discounts, renewal uplift left on the table, and price-floor breaches. Design partners shape the priority.
The cloud LLM sees only aggregate figures, column headers, document chunks you upload, and your questions. Never row-level data, and always under zero-retention provider terms.