A partner at a 12-person advisory firm told me his team spends a full day and a half building the pricing model for every major proposal. Not writing the proposal. Just the pricing. Scoping the hours, guessing at complexity, sanity-checking against what they charged last time, and hoping nobody undersells it in the client meeting.

That’s 10 to 15 hours of senior time, every time, before a single word of the actual deck gets written. If your firm does even 20 proposals a year, that’s 200 to 300 hours of partner or principal time spent on spreadsheets and gut checks. At a fully loaded rate of $150-$250 an hour for the people usually doing this work, you’re looking at $30K to $75K a year just in the pricing exercise. That’s before you count the deals priced wrong.

The real question isn’t time, it’s accuracy

Time is the easy problem to see. The expensive problem is harder to see because it hides inside your win rate.

Here’s what actually happens in most consulting firms doing $1M-$25M in revenue. Pricing gets built from memory and instinct. Someone recalls “the last engagement like this was around $80K” and adjusts up or down based on gut feel about the client, the timeline, or how badly the firm needs the work that quarter. There’s no systematic pull of what similar engagements actually cost to deliver, how many hours they really took versus what was scoped, or which project types quietly ran over margin every single time.

We see this constantly in our network. Firms with strong win rates that are still underpricing by 10-20% on a meaningful chunk of their engagements, because the pricing model has no memory. Every proposal starts from a blank page and a partner’s recollection, which is a bad system for anything you do more than a few times a year.

The fix isn’t a better spreadsheet template. It’s giving the pricing process access to what the firm actually knows, which is buried in old proposals, closed-project actuals, and scope documents that nobody has time to go back and mine.

What 20-40 hours of proposal work actually costs you

Pricing is only one piece. The proposal itself, the deck, the case studies, the tailored narrative, typically eats another 20-40 hours of senior time on anything that matters. Multiply that across a dozen or two major proposals a year and you’ve got a meaningful chunk of your best people’s time going into documents that get thrown away the moment the deal closes or doesn’t.

This is the pattern we walk through in the AI audit for consulting firms, and it usually surprises owners how much of it is repeatable. Most of a proposal, the structure, the relevant case studies, the standard scope language, the pricing logic, has already been written before. Somewhere. For a different client. It just isn’t organized in a way anyone can pull from quickly, so it gets rebuilt from scratch every time.

That’s the gap a Proposal Generation Agent is built to close.

What the Proposal Generation Agent actually does

This isn’t a template generator. It’s an agent that has read every past proposal your firm has sent, knows which ones won and which didn’t, and has access to the actual pricing and scope data from delivered engagements, not just the numbers you quoted going in.

When a new opportunity comes in, the workflow looks like this:

A partner or business development lead feeds the agent the basic shape of the opportunity, industry, engagement type, rough scope, timeline. The agent pulls the closest matching past proposals and, more importantly, the closest matching delivered projects, and cross-references what was quoted against what it actually took to deliver. That’s the piece most firms never do manually because nobody has time to go back and reconcile 40 old projects against their original scopes.

From there it drafts a tailored proposal, pricing included, grounded in what similar work has actually cost your firm to deliver, not what someone remembers charging two years ago. It flags where the new opportunity looks riskier than past comparables, complexity, timeline pressure, unclear scope, so the team can price in a margin cushion instead of finding out six weeks into delivery that they underbid.

A partner still reviews it. This isn’t about removing judgment from pricing. It’s about making sure judgment starts from real data instead of memory, and that the 10-15 hours of model-building gets compressed to something closer to 2-3 hours of review and adjustment.

The Research Agent problem sitting right next to it

Pricing accuracy solves half the leakage. The other half sits earlier in the process, in the research every engagement kicks off with.

Every new client relationship starts the same way. Someone on the team spends one to two weeks pulling industry data, competitor positioning, and company background before the real work can start. Then the next engagement in a similar space starts, and a different team member does a version of the same research again, because nobody wrote it down anywhere searchable.

A Research Agent handles this at the front of every engagement. It runs structured research on the industry and the specific company, pulls sources, builds a summary, and produces a one-page brief the team can use in the kickoff meeting instead of week one. It doesn’t replace the strategic thinking your consultants bring. It removes the repeated grunt work that shouldn’t need a $200-an-hour person doing manual searches for three days.

Pair that with a Knowledge Agent that’s read every deck, report, and meeting transcript your firm has ever produced, and you’ve closed the third leak most firms don’t talk about openly: the fact that a $1M-$25M consulting firm generates enormous amounts of IP across its projects and reuses almost none of it. Every project pays for insight the firm has technically already paid for once before.

Doing the dollar math on your own firm

Take your own numbers for a minute. How many proposals over $50K did you send last year? Multiply that by the hours your senior people spent building pricing models, not writing the deck, just the numbers. Then multiply by their loaded hourly cost.

Now add the harder number. How many engagements in the last two years came in under margin because the original scope was too optimistic? If you’re not sure, that’s worth sitting with. Most firms in the $1M-$25M range aren’t sure, because nobody’s built the system to check it automatically.

This is exactly the exercise we run through in the Omni Audit. It’s 60 minutes, no slide deck, and it produces three concrete outputs, a leakage estimate specific to your proposal and delivery data, a shortlist of which agents would move the needle first, and a rough sequencing plan for what to build and when. Firms usually leave that call with a number attached to the problem instead of a hunch.

If you want to see the mechanics before that call, our overview of Omni’s operations agents walks through how the Proposal Generation, Research, and Knowledge agents connect to each other, since they all draw from the same underlying project history.

What this looks like in the first 90 days

Nobody should try to automate their entire proposal process in month one. The firms that get the most out of this start with one agent, usually the Proposal Generation Agent since it touches revenue directly, and get it trained on 18-24 months of past proposals and project actuals before touching the next piece.

We’ve put together a practical worksheet for exactly this, called Deploy Your First Business Agent. It walks through how to pick the first agent worth building, what data it needs from your firm to be useful on day one, and what a realistic 90-day rollout looks like without derailing delivery work. If you want the direct version, grab the worksheet here and use it alongside whatever comes out of your audit call.

It’s worth reading a bit more broadly too. Our insights section has a few pieces on how firms in professional services are sequencing agent rollouts against delivery capacity, and the guides library has more detail on data readiness if your project history is scattered across old drives and email threads, which is normal and fixable.

Is it actually worth it

Here’s the honest answer. If your firm sends fewer than 10 major proposals a year and pricing has never bitten you on margin, this probably isn’t your highest-priority fix right now. Go look at delivery bottlenecks instead.

But if you’re sending 15, 20, 30-plus proposals a year, if your senior people are the ones building pricing models instead of talking to clients, and if you’ve ever found out three months into a project that the scope was tighter than the price allowed, the math works. Recovering even half the time currently spent on manual pricing, at $150-$250 an hour for the people doing it, pays for the build inside a single quarter for most firms in this revenue band.

The bigger win compounds after that. Once your pricing is grounded in what you actually delivered rather than what you remember charging, your margins get more predictable, and your partners get 15-20 hours a month back to spend on the client work that actually grows the firm.

If that sounds like where your firm is right now, the next step is a conversation, not a purchase decision. Book a 60-min Omni Audit and we’ll walk through your actual proposal volume and pricing history together, no deck, no pitch, just a straight read on where the dollars are sitting.

You can also poke around our blog for more on how other consulting and advisory firms are approaching this, or start with See Omni for consulting firms if you’d rather see the specifics before booking anything. Either way, the numbers are worth knowing, even if you decide not to act on them yet.

If you’re ready to talk through what this looks like for your firm specifically, book your Omni Audit here and bring your last 10 proposals. That’s usually enough for us to tell you where the leak is.