How to track quote win rate so the number means something

A general manager asks what the win rate is. The sales head says 30 percent. The estimator says it is closer to 15. Both are looking at the same quarter and both are right, because nobody agreed what goes underneath the line.

Quote win rate is the most quoted number in a quoting operation and the least defined. Getting the definition settled is worth more than any improvement to the number itself, because until it is settled nobody can tell whether it moved. It is also the number that moves when response time falls from 10 days to 1.

Why one quarter gives 3 different win rate answers

Take an illustrative quarter. 360 requests arrive, 240 get quoted, 120 never do, 72 are won and 165 are decided. Against quotes decided the win rate is 44 percent. Against quotes sent it is 30 percent. Against requests received it is 20 percent.

Every one of those is defensible and they answer different questions. Against decided measures how competitive you are when somebody actually chooses. Against sent measures the same thing but drags in quotes still pending. Against received measures how much of your market you are addressing at all.

Agree the denominator first.

The one most floors report is against sent, because it is the easiest to extract and the most flattering of the 3. It also hides the number that usually matters most, which is the 120 requests that never got quoted.

Pick one as the headline and keep the others visible. What breaks a management conversation is 2 people using different denominators without either of them knowing it.

Why unquoted requests belong in the denominator

A request you never answered is a loss with no paperwork. The customer asked, you did not reply in time, and somebody else made the part. Excluding those from the measurement makes a capacity problem invisible.

This is the specific failure that a quote never sent is about, and it is why the received denominator is the one that tracks revenue rather than tracking pricing skill.

There is a fair objection. Some unquoted requests were deliberate no-quotes, declined for good reasons, and counting a decision to decline as a loss is misleading in the other direction. The answer is to count them separately rather than to exclude them silently.

Three buckets settle it. Quoted, declined with a reason, and never answered. The third bucket is the one to drive to zero, and on most floors nobody knows its size because it was never a category.

Segmenting so the number is actionable

A single company-wide win rate is a vanity metric. It moves for reasons nobody can name and it never tells you what to do differently on Monday.

Segment by customer first. Winning 60 percent at one account and 8 percent at another is 2 completely different situations averaged into one meaningless figure, and the 8 percent is either a pricing gap or a relationship that has already ended without telling you.

Segment by part family second. If turned parts convert at 40 percent and fabrications at 12, the fabrication rates or the fabrication process need work, and no amount of general margin adjustment will find that.

Segment by annual usage third. High volume programs and one-off jobs are different businesses with different competitors, and a floor that wins short runs and loses programs is being told something specific about where its cost structure fits.

The lag between quoting and knowing

A quote sent in March may be decided in September, and that lag breaks naive reporting. Measuring this month’s wins against this month’s quotes compares 2 different populations and produces noise that people then explain.

Cohort it instead. Take every quote sent in a month, follow that cohort until it is fully decided, and report the rate against the month it was quoted rather than the month it was awarded.

That means recent months are always incomplete, which is uncomfortable and correct. A cohort that is 60 percent decided should be reported as such rather than presented as a final figure that then moves.

The lag also has a practical use. If quotes typically decide in 8 weeks and one is still open at 20, it is almost certainly lost and nobody sent the rejection. Cleaning those out of the pending pile is how the pending pile stops flattering the number.

Cohort by the month you quoted.

What to record against every quote

The measurement is only as good as the fields captured when the quote goes out. Six of them do all the work, and none of them takes more than a few seconds at the time.

Date received and date sent, because the difference between those 2 is the variable most correlated with winning. Customer and part family, so the segmentation works. Value, so a win rate can be weighted by money rather than by count. And outcome with a date and a cause.

Weighting by value is the one most often skipped and it changes the picture. Winning 30 percent of quotes that represent 12 percent of the value quoted is a different business from winning 30 percent of the value, and only the second one pays for the year.

The cause field needs to be the one you established rather than the one you were told, which is what a proper loss review produces. A cause field populated with price on every row carries no information at all.

The trap of improving the ratio by quoting less

Any win rate can be improved by declining everything difficult. Quote only the work you are certain of and the percentage climbs while the revenue falls, which is the most common way this metric gets gamed without anybody intending to game it.

That is why the received denominator has to stay visible beside the sent one. A rate that improved because the floor addressed less of its market is not an improvement, and the pair of numbers makes it obvious where a single number hides it.

A better ratio can be a smaller business.

The same caution applies to margin. Win rate rises as prices fall, so a target expressed only as a percentage points straight at discounting. Pair it with won revenue and with margin on won work, and the incentive becomes coherent.

Watch it during a price-down cycle in particular, when accepting reductions lifts the win rate on existing accounts while the business gets worse. A customer running should-cost analysis is optimising their number, and your win rate rising is not evidence that you are winning.