How to

Winning combinations: the facts your wins had together

The short answer

A winning combination is a set of facts that your won deals had together and your lost deals mostly did not, such as a first meeting within 14 days, a follow-up within a day of it, and an operations leader on the deal. Kaypo reads every closed deal as it stood about 30 days before it closed, learns separately for New, Expansion and Renewal, and shows a combination only after it holds on the newest deals, is found again on reshuffled deals, points the same way across the team's history, and moves the win rate enough to matter. Open deals that carry a combination, or are one step from it, are named on the deal.

Why combinations

A score weighs each signal on its own. Many deals are decided by facts that matter only together: pricing read by the right person after a fast first meeting, or a champion leaving while the customer researches a competitor.

Searching every combination of a team's facts is fast enough to do exhaustively on closed deals, so the work is in checking what it finds.

What a combination has to pass

It has to hold on the newest closed deals, which the search never saw. It has to be found again when the deals are reshuffled. It has to point the same way across most of the team's history. And it has to move the win rate by at least 10 points, or cut the losses by a third, to be worth a rep's time.

Each fact in it has to earn its place: among the deals with every other fact, the ones that also have this one have to win, or lose, measurably more.

How Kaypo answers this for your pipeline

Every open deal, scored on what the buyer did, with the move that worked in your own wins.

Kaypo reads your CRM, email and calls, checks every signal against the deals you have already won and lost, and puts the answer on each deal: the score, why it moved, and the next step with the email written.

Nothing written at the close

Every closed deal is read as it stood about 30 days before it closed. Fields filled in at or after the close, such as the loss reason, the final amount or a probability, are set aside with the reason, so the combinations describe the deal and never the outcome.

On the deal itself

An open deal that carries a losing combination says so, with the count behind it. One that is a single step from a winning combination names the step, when the step is the rep's to take. When a deal carries a winning and a losing combination at once, Kaypo shows how the closed deals that had both actually ended.

Questions people ask

How many closed deals do winning combinations need?

Winning combinations need enough closed deals of each type to hold a newest share back for checking. A few hundred is a floor for strong patterns, and the subtler ones usually appear between 1,000 and 3,000 closed deals of that type.

How does Kaypo keep combinations from finding noise?

Kaypo checks every combination on the newest closed deals, again on reshuffled deals, and across periods of the team's history, and drops any fact that does not earn its place. On made-up deals with nothing planted in them, this shows nothing at all.

Are combinations different for renewals and expansions?

Combinations are learned separately for New, Expansion and Renewal, because the facts that decide a renewal (tenure, usage, the champion, earlier contracts) are mostly about the customer instead of the opportunity.

Reviewed October 27, 2026 against the product as it behaves today.

Your reps never log in. Kaypo comes to them.

Kaypo checks every buyer signal against the deals your team has won and lost. Every morning in Slack, each rep gets the deals worth their time, why each one moved, and the next step that worked in wins like it, with the email already written.

$1,500 a month, or $15,000 a year. One admin and up to 15 users. No per seat charge. Connect a CRM and it runs. No implementation fee. 14 day trial. No card.

See all of Kaypo Back test my own deals, free Try it Pricing How Kaypo works

Related reading

Which sales signals predict together

Deal scoring weights signals one at a time. Pairs that predict together are a separate question, and most scoring models never ask it.

What predicts whether a customer renews

Renewals turn on the customer more than the renewal opportunity. The CRM facts that tell renewals apart, and how to check them on your own.

What predicts whether a customer expands

Expansion follows room to grow and a reason to buy more. The facts that point to it in your CRM and your usage data, and the ones that point the other way.

Back testing a deal score against your own closed deals

A back test scores deals you have already closed as they stood before they ended, then checks the scores against what happened. The steps.

What actually predicts whether a deal closes

Every scoring tool asserts which signals matter. The only way to know is to check them against the deals you already closed.