Most presales metrics count activity, such as demos delivered or hours logged. Those record how busy the team was, which is not the same as whether it was effective. The measures that matter track judgment: how early the team reads a deal correctly, whether that read is written down, and whether anything changes because of it.
Demos delivered, hours logged, requests handled, proofs of concept run. Every one of these can be pulled from a system without arguing with anybody, which is most of the reason they persist.
They also survive because they are defensible upwards. A presales leader asked to justify headcount can show a rising activity line and nobody in the room can dispute it. The number is real. It just answers a question nobody actually asked, which is whether the team was busy.
A team can double its activity and lose exactly the same deals it lost last year.
The usual escape from activity metrics is to borrow the seller’s. If presales supported the deal, measure presales on whether the deal won.
It is a reasonable instinct and it breaks down quickly. Presales does not own the close, the pricing, the paper process or the executive relationship, so win rate moves for reasons the engineering work never touched. A team can be marked down for a procurement decision it was never in the room for, and marked up for a renewal that was always going to land.
Forecast accuracy sits much closer to what presales actually influences, because the engineer is frequently the first person in the business with an accurate read on whether a deal is real.
Not whether the number was hit, but whether the deals called at the start of the quarter were the deals that closed. That distinction is the whole measurement, and most teams only track the first half.
Pick the deals that died after a technical win last quarter. Ask when somebody first said it was in trouble. The gap between that date and the close date is the number worth shrinking.
Sixty one percent of formal warnings change nothing and eighty nine percent travel to the account executive and no further. A concern that was never recorded cannot be scored afterwards, so nobody learns who was right.
Under two percent of sales engineers say their team runs deals the same way. If each engineer has a private method, the team’s results are thirty separate results that happen to be added together.
Forecast accuracy moved first, which is the usual order. Win rate followed, partly because fewer weak deals were being carried into a quarter where they consumed engineering time that better deals needed.
The headcount did not change. That matters for the measurement argument, because it removes the simplest alternative explanation.
Do not build a dashboard first. A dashboard with nothing behind it produces activity metrics within a month, because those are the fields that are already populated.
Start with last quarter’s losses. Take the deals that died after a technical win, and for each one find out when the first person said it was weak and what happened to that. Ten deals is enough to see the pattern.
Then measure the team against itself. A benchmark run twice, a couple of quarters apart, gives a before and after on the same questions. That is a more honest instrument than comparing your team to an industry average built from companies unlike yours.
Most presales metrics count activity, such as demos delivered or hours logged, which records how busy the team was rather than whether it was effective. The measures that matter track judgment: how early the team reads a deal correctly, whether that read is recorded, and whether anything changes because of it.
Only as context. Presales does not own the close, so win rate moves for reasons that have nothing to do with the engineering work, and a team can be measured down for a pricing decision it never saw. Forecast accuracy is closer to what presales actually influences.
Six anonymous questions, fifteen minutes, inside a meeting your team already holds. Run it now and the same six questions in two quarters tell you whether anything moved, on your own team rather than against an average.