Ask any partner at a professional-services firm how much a proposal costs and you'll get a confident answer. Ask them to show their math and you'll get a different one. The gap is almost always larger than people think.
The number most firms quote
The number is usually some version of "we spend about 40 hours on a proposal." Multiply by a blended hourly rate, you get $6,000-$10,000. That's what gets logged in the system if anyone logs it.
That number is wrong in three directions, and they all stack.
What's missing: the opportunity cost
The 40 hours don't come from anywhere. They come out of billable time, biz-dev relationship time, or the partner's evenings. None of those are free.
If a senior associate spends 8 hours on a proposal, that's 8 hours of $300/hr billable revenue not generated — $2,400 in lost margin, on top of the $1,600 in fully-loaded labor cost. The total cost of that associate's 8 hours is closer to $4,000, not $1,600.
Across a 40-hour proposal with the typical staff mix, your true labor cost is usually 1.8-2.4× the loaded labor cost. Most firms ignore this.
What's missing: the overtime tax
Proposal cycles concentrate. The last 72 hours before submission take twice as much human energy as the prior week combined. That energy is overtime, which means:
- Diminished output the following week — recovery time isn't logged but it's real.
- Quality drift on adjacent work — the project the proposal-writer was running gets worse during proposal week.
- Talent risk — the senior staff doing the all-nighters are also the ones who get recruited.
The overtime tax is usually 20-35% of the proposal's true cost. It doesn't show up on a P&L; it shows up in turnover spreadsheets a year later.
What's missing: the proposals you didn't write
The cost that hurts most is the cost of the 40+ opportunities your BD team sourced and never pursued. If your team's max proposal cadence is 1.5/week, and 40 qualified opportunities/quarter come in, you're declining 1.5/week — and the ones you decline aren't random.
You decline the ones that are different — different geography, different scope, different vertical. Which means you optimize for the deals you already know how to win and never expand into the deals that would grow the firm.
The simple model
Here's the math any partner can run in an afternoon. Run it for your last 12 months.
- Hours per proposal × loaded rate × proposal volume — the obvious cost.
- + Hours × (blended billable rate − loaded rate) — the opportunity cost.
- + 25% overtime multiplier on the last 16 hours of each cycle — the overtime tax.
- + Declined-opportunity value × declined volume × historical win rate × average deal size × 0.5 (haircut for fit) — the proposals you didn't write.
The number that comes out is uncomfortable. For a 25-person firm doing 80 proposals/year, it's commonly $1.4M-$2.2M. The "we spend $500K on proposals" answer is missing 60-75% of the actual cost.
Knowing the real number doesn't mean you should spend that much on automation. It means you finally know how much you can afford to.
What this changes
Two things, usually. First, the conversation about AI investment becomes much shorter — when you know proposals are a $1.8M problem, spending $150K to compress the cycle is obviously the right move. Second, BD strategy changes — once the implicit "decline most opportunities" filter is visible, partners start asking why it exists.
Run the model. Print it out. Bring it to your next partner meeting. The conversation that follows is the most useful one your firm will have this quarter.