Benchmark 6 min readUpdated August 4, 2026

Bid-hit ratio: how to measure and actually improve your win rate

Almost every subcontractor quotes a win rate. Very few can show the arithmetic behind it, and the ones who can usually discover the number is not what they thought. Here is how to calculate it so it means something, and what to do with it once it does.

The formula, and the three ways it goes wrong

The arithmetic is trivial: bids won ÷ bids submitted, over a defined period. Everything hard about it is in the definitions.

Mistake 1: counting invitations instead of submissions

If your denominator includes invitations you declined, your win rate is measuring how selective you are, not how competitive you are. Two subs with identical pricing and identical results will report wildly different numbers if one of them declines half their invites and the other logs them all.

Track both, separately. Bid/no-bid rate (submitted ÷ invited) tells you about selectivity. Win rate (won ÷ submitted) tells you about competitiveness. Collapsing them into one number destroys both signals.

Mistake 2: leaving the unresolved bids out

The tempting move is to compute win rate over bids with a known outcome, quietly dropping the ones that went silent. This inflates the number, sometimes dramatically — and it inflates it most for the desk with the worst follow-up discipline, which is exactly backwards.

Better: age unresolved bids out on a fixed rule. Anything with no response 90 days past the bid date counts as a loss. It will be wrong occasionally. It will be far less wrong than excluding it.

Mistake 3: treating one regret letter as one lost job

On a project several GCs are bidding, a regret from one of them means you lost with that GC — not that the project is gone. If you record it as a total loss you have both corrupted your win rate and hidden the fact that you are still live with someone else on the same job.

Count at the GC level, not the project level

A bid is a you-to-one-GC event. If you bid the same project to three GCs, that is three bids in the denominator and up to one win in the numerator. Any other convention makes the number impossible to compare across periods.

The overall number is nearly useless

A single blended win rate averages together situations with nothing in common: the GC you have worked with for nine years, the national who invites twelve subs per package to satisfy a procurement policy, the developer three hours away whose schedule you cannot service, and the one job type where you have a genuine cost advantage.

The blended number hides all of it. Cut it four ways and it starts telling you what to do:

Cut by	The question it answers	What you do about it
General contractor	Who actually awards us work?	Prioritize invites; have a direct conversation with the zero-win accounts
Project size	Where is our sweet spot?	Stop bidding the band where you never win
Geography / distance	Where does our cost advantage end?	Set a travel radius and hold it
Job / system type	Which scopes do we price well?	Chase the strong ones; add margin or decline the weak ones

The per-GC cut is usually the most revealing. A desk with an overall 15% rate frequently turns out to be running near 40% with four contractors and near zero with eleven others — and the eleven are consuming most of the estimating hours.

The four levers that actually move it

Bid less
The fastest lever, and the least popular. Declining the invitations you have historically never won frees estimating hours for the ones you do. Win rate rises because the denominator falls — and revenue usually holds, because you were not winning those anyway. This only works if you have per-GC history; without it you are guessing.
Respond faster
On packages where numbers arrive over a window rather than at a hard deadline, being early matters more than most subs think. An estimator assembling a bid tab anchors on the first credible numbers in. Being late means being compared to an anchor rather than setting one.
Follow up on every submitted bid
The cheapest points on the board. Some meaningful share of quiet bids are not losses — they are jobs where the estimator has an unresolved question about your scope. A single clarifying email converts some of those. See the bid follow-up playbook.
Find out why you lost
The winning number, if the GC will share it, is the single most valuable piece of feedback in the business. Consistently 3-4% high with one GC is a fixable pricing problem. Consistently 30% high is a scope-interpretation problem — you are pricing something they are not asking for. Those are completely different fixes and you cannot tell them apart without asking.

What about the industry average?

Published averages for subcontractor hit rates vary so widely across trades, regions, project types, and — crucially — across how each survey defines a "bid" that comparing yourself to one is close to meaningless. A trade with few qualified bidders in a tight market and a trade where every GC solicits ten numbers will produce figures with nothing in common.

The comparison that matters is against your own number last quarter, cut the four ways above. That one is measured consistently, reflects your actual market, and is the only version you can act on.

A reasonable first goal

Before optimizing the ratio, get to the point where you can produce it on demand, correctly, for any GC, for any period, without a night of digging. Most desks cannot. Getting there tends to surface two or three obvious decisions on its own.

Instrumenting it

This measurement is only sustainable if outcomes are captured as a by-product of work you already do, rather than as a monthly data-entry chore that gets skipped in busy season. That means: invitations logged automatically on arrival, the real GC resolved rather than the relay platform, submissions timestamped, and awards and regrets recorded against the specific contractor when the reply comes in.

That is the design goal behind IntelBid — the pipeline records outcomes as they happen so win rate is a query, not a project. If you are starting from a spreadsheet, the history can come with you: see switching from spreadsheets.

How do you calculate a subcontractor bid win rate?

Divide bids won by bids submitted over a defined period. Exclude invitations you declined from the denominator — those belong to a separate bid/no-bid rate. Count each bid as one you-to-one-GC event, so a project bid to three general contractors is three entries in the denominator.

What is a good bid hit ratio for a subcontractor?

There is no single credible benchmark. Published figures vary enormously by trade, region, project type, and how each survey defines a bid. The useful comparison is against your own prior periods, segmented by general contractor, project size, distance, and scope type.

Should bids with no response count as losses?

Yes, on a fixed aging rule — 90 days past the bid date with no reply is a common threshold. Excluding unresolved bids inflates the number most for the desks with the weakest follow-up, which makes the metric actively misleading.

Why is per-GC win rate more useful than an overall win rate?

A blended rate averages together relationships with completely different dynamics. Segmented by contractor, most desks find a small group of GCs producing most of their wins and a long tail producing almost none while consuming most of the estimating hours — which is a directly actionable finding.

Know your win rate by GC, not by feel

IntelBid records outcomes as replies come in, so per-contractor win rate is a filter rather than a research project. Import your existing bid history to start with real numbers.

Start your free trial