Deal Flow Benchmark

When do sales engineers enter the deal?

Fifty nine percent of sales engineers enter a deal only after discovery is already finished, and twenty seven percent get in before it ends.

Source: the Deal Flow Benchmark, an ongoing study by Dmitri Lee, Presales Navigation System. Findings current as of October 2026. Answered anonymously, live, never attributed to an individual. The full field will be published when the study completes. Question: at what stage of a deal does the engineer get involved.
59%
enter after discovery is finished
27%
enter before discovery ends
14%
report no pattern at all

Most engineers inherit a question somebody else asked

Fifty nine percent of sales engineers enter a deal after discovery is already complete. Roughly a third arrive between discovery and the demo, and a quarter arrive only once a demo has been scheduled, which means the deal has a shape, a problem statement and often a champion before anyone technical has asked a question.

Twenty seven percent get in before discovery is finished, and a small minority are in before the first customer call. Fourteen percent report no pattern at all, which is its own answer: entry is being decided deal by deal, by whoever happens to be free.

The questions that decide whether a deal is real get asked once

Discovery is where a deal acquires its reasons. What the customer is trying to change, who pays for it, what happens if nothing happens. An engineer who arrives after that is not being asked to find out whether the deal is real. They are being asked to confirm that the product fits a problem statement written for another purpose.

That is the gap this measurement keeps exposing. It is also the one finding that leaders consistently say they already suspected and had never seen written down.

Entry point is a decision, even when nobody makes it

The spread between organisations is wider than anything else measured. In some teams nobody at all gets in before discovery ends. In others most of the team does, and a few are in before the first customer call. Same role, same market, opposite practice.

Early access did not translate into better outcomes on its own. Across the field there is no clean relationship between getting in early and spending the week well. Access is not the same as a method for using it.

Six questions, answered live, never attributed

The instrument

Six fixed questions, put to a whole team at once inside a meeting they already hold. Fifteen minutes. Everyone answers at the same time from their phone, anonymously, and the distribution appears in front of the room as it fills.

The wording never changes between organisations, because changing it would destroy the comparison. Nothing is attributed to a person and nothing is attributed to a company. Leaders see their own team’s results before anyone sees anything.

This is a small number of organisations measured in depth rather than a large survey. It is not a representative sample of the profession and it is not presented as one. What it is: the same six questions, asked the same way, of teams who answered without a manager watching.

Decide the entry point once, rather than per deal

Fourteen percent reporting no pattern is the most actionable number here, because it is the only one that describes an absence rather than a choice. Those deals are being staffed by whoever was free.

Two changes, and neither needs budget. Write the entry point down. One sentence is enough: this is the point at which an engineer joins, and this is what has to be true before the demo is booked. Teams that have one get asked for it constantly, which is how you know it was missing.

Then check it against reality for one quarter. Note the actual entry point on every deal, not the intended one. The gap between the two is the finding, and it is usually wider than anyone expects.

What this does not fix on its own is what happens after the engineer arrives. Early access turns out not to help much by itself, which is the uncomfortable part of this study and the subject of the other findings.

What people ask

When do sales engineers typically get involved in a deal?

According to the Deal Flow Benchmark, fifty nine percent of sales engineers enter a deal after discovery is already finished, twenty seven percent enter before discovery ends, and fourteen percent report no consistent pattern. The study is ongoing and these are the findings to date.

Why does it matter when a sales engineer enters a deal?

Discovery is where a deal acquires its reasons: what the customer wants to change, who pays, and what happens if nothing happens. An engineer who arrives afterwards inherits a problem statement written by someone else for another purpose, and is asked to confirm product fit rather than to establish whether the deal is real.

Run it with your own team

The benchmark is free and it stays free. Six questions, fifteen minutes, answered anonymously inside a meeting your team already has. You get a written read on your own numbers and where they sit against the field. There is nothing to install and nothing to buy.

Ask about the benchmark Or book a slot for your team