Stop giving away price that wins you nothing.
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 you walk away with
The diagnostic is four artifacts, not a dashboard you have to interpret
Each one is built so a Head of Pricing can forward it to the CFO unchanged. No screenshots-of-screenshots, no copy-paste cleanup.
A one-page written narrative your CFO can read in two minutes — booked value, price realization, win point, the headline upside, the top three deals to investigate. Plain English, no jargon.
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.
For each open opportunity you send: 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.
Won deals split into 'discount earned the win' (defend in sales review) vs 'discount worth a retrospective look' (route to deal-desk debrief). Sales-friendly framing of the same data.
Back-of-envelope estimator
What might your pricing upside look like?
Move the sliders. The math is the same logic our diagnostic uses, applied to round numbers. Your real number comes from your CSV.
What Pricekeel surfaced on the bundled sample
2,000 closed deals → three numbers a CFO can act on
Live figures from the bundled sample. Your numbers come from your CSV.
What it does
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.
Find the win point
The discount level where win rate stops improving. Everything given beyond it is a list to investigate, not a refund.
Guide the next discount
A win-probability model recommends the discount that maximizes expected value on a given deal, with a plain-language why.
Defend the decision
Every Pricekeel recommendation is logged with the math behind it. The Copilot answers CFO questions with citations to the source signal — no LLM-invented numbers.
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 land with the Phase 3 margin layer.
Margin Enhancement
Next, Pricekeel connects to your contracts and CRM to find margin across the whole book: special pricing agreements, fixed discounts, renewal uplift left on the table, and price-floor breaches. From a retrospective diagnostic to a live margin engine.
M.S. Marketing Analytics with a Pricing specialization from Simon Business School (University of Rochester). Previously led a $626M margin-recovery analysis across 1,200+ opportunities. Building Pricekeel to make that kind of analysis routine for any pricing team — and defensible to finance.
I’m taking three design partners this quarter. Free diagnostic under NDA, 30-minute readout with me, and if the upside is real we keep going on a quarterly cadence. If it’s not, you keep the analysis and we part as friends.
The cloud LLM sees only aggregate figures, column headers, document chunks you upload, and your questions — never row-level data. Under zero-retention provider terms.
See it on your own deals
Export a CSV of your closed opportunities and get your own diagnostic. Your data is processed to produce the analysis and is not stored. If you need an NDA first, we will sign one.