Most consulting and advisory firms in the $1M to $25M range don’t have a revenue problem. They have a leverage problem. The partners are good at winning work and good at delivering it. What’s eating them alive is everything in between, the proposal that takes three days to write, the research phase that starts from zero on every single engagement, and the knowledge that walks out the door with every consultant who leaves.

We’ve looked at enough firms in this band to put a number on it. The typical range we see is $80,000 to $300,000 a year in leakage, not from bad work, but from good people doing work a machine should be doing instead. This article walks through where that money actually goes and what it looks like to claw it back.

The proposal problem nobody budgets for

Ask any partner how long a major proposal takes and you’ll get a number between 20 and 40 hours. That’s not the junior associate’s time either. That’s senior time, the people whose billable rate is the highest in the building, spending days pulling together a deck that looks a lot like the last five decks they wrote.

The pattern is always the same. Someone digs through old proposals looking for a case study that fits. Someone else rewrites the methodology section because nobody can find the last version. Pricing gets rebuilt from a spreadsheet that lives on one person’s laptop. By the time the proposal goes out, the firm has spent a meaningful chunk of the deal’s margin just trying to win it.

Win rates in this range tend to be fine, often 30 to 50 percent depending on the firm and the market. The issue isn’t quality. It’s cost of sale. A firm winning six-figure engagements can’t keep spending partner-level hours assembling documents that are 70 percent recycled content and 30 percent new thinking. That ratio should be closer to reversed.

This is exactly what a Proposal Generation Agent is built to fix. It pulls from every past proposal, every case study, every pricing model the firm has ever used, and produces a tailored first draft against the new opportunity in minutes, not days. The partner still shapes the strategy and the narrative. The agent just removes the eight hours of assembly work that used to come before that.

Research that gets paid for twice

Every new engagement starts the same way. Someone on the team spends one to three weeks doing secondary research, understanding the client’s industry, mapping the competitive set, reading through analyst reports, building a point of view before the real work even starts.

Here’s the part that should bother any partner running the numbers. A firm serving multiple clients in the same industry, which is most firms, ends up paying for that same research over and over. The healthcare vertical gets re-researched every time a new healthcare client signs. The manufacturing landscape gets rebuilt from scratch for every manufacturing engagement. Nobody is stealing this time. It’s just genuinely repeated work, and it compounds across the firm every quarter.

A Research Agent changes the shape of that cost. It runs structured industry and company research automatically at the start of every engagement, pulling from live sources, producing summaries, and delivering a one-page brief the team can act on inside a day instead of inside three weeks. It doesn’t replace the analyst’s judgment. It replaces the grinding, repetitive first pass that judgment used to depend on.

If your firm runs four or five engagements a year in overlapping sectors, this alone tends to be the single biggest line item in the leakage number. It’s also the easiest one to see once someone actually adds up the hours.

The knowledge management debt that never gets paid down

This is the one most firms don’t talk about because it’s hard to see on a P&L. Every engagement your firm runs produces intellectual property. Frameworks, client insights, market data, deliverable templates, lessons on what worked and what didn’t. Almost none of it gets reused in any structured way.

It sits in a folder. It sits in someone’s inbox. It sits in a deck that only the person who wrote it remembers exists. When a similar question comes up on a different engagement eighteen months later, the firm doesn’t retrieve that IP. It rebuilds it. That’s the debt. The firm already paid for the insight once, in billable hours, and now it’s paying for it again.

This debt gets worse as firms grow, not better. More consultants means more decks, more transcripts, more one-off Slack threads with a genuinely good idea buried in message forty of ninety. Without a system to hold onto it, growth actually makes the knowledge problem harder, not easier.

A Knowledge Agent is the fix for this specific issue. It reads every deck, every document, and every meeting transcript the firm produces and lets anyone on the team ask questions directly across that entire corpus. “What have we told clients about pricing model transitions in the last two years?” becomes a two-minute answer instead of a hallway conversation with the one partner who might remember.

What this actually looks like end to end

Picture a mid-size advisory firm that just got a lead for a $400,000 engagement in a sector they’ve served twice before. Here’s the difference between the old way and the agent-run way.

Old way: a senior partner spends a weekend assembling the proposal from three old documents. A manager spends two weeks building a research brief that mostly repeats work done eighteen months ago on a different client in the same sector. Six months into delivery, someone asks “didn’t we solve a version of this problem for another client last year?” and nobody can find the answer fast enough for it to matter.

Agent-run way: the Proposal Generation Agent produces a full first draft within the hour, built from the firm’s actual win history and pricing logic. The Research Agent delivers the one-page brief on day one instead of day fourteen. Three weeks into delivery, a consultant asks the Knowledge Agent whether the firm has handled this exact problem before, and gets a direct answer with the source document attached, in under a minute.

Nothing about the quality of the thinking changes. What changes is how much of the team’s time goes toward the parts of the job that actually require a person, the judgment, the client relationship, the point of view. That’s the entire case for building this the right way, and it’s worth reading more on how the mechanics work in our broader guides on deploying AI inside professional services firms.

Why most firms don’t fix this on their own

Every partner we talk to already knows these three problems exist. Almost none of them have fixed it, and it’s not because they don’t care. It’s because building this internally means someone has to own it, and the people capable of owning it are the same people billing $400 an hour on client work. The opportunity cost of building it yourself is often higher than the problem it solves.

That’s the gap Omni is built to close. Instead of asking a partner to become a part-time AI project manager, we design and deploy the agents directly into the firm’s existing workflow. You can see the shape of that approach across Omni’s ops-focused agent work and in how we think about advisory-specific deployments over on our advisory practice page. None of it requires the firm to hire a data team or stand up new infrastructure.

The Omni Audit, and why it’s 60 minutes and not a sales pitch

We don’t open with a deck and we don’t ask you to sit through a pitch. The Omni Audit is 60 minutes, and it produces three concrete outputs: a map of where your firm’s hours are actually going across proposals, research, and knowledge work, a dollar estimate of what that’s costing you annually, and a short list of which agent would move the needle first if you deployed it in the next 90 days.

No proposal, no retainer conversation, no obligation. Just a clear picture of the number and where it’s coming from. Firms that go through it usually tell us the research and proposal categories are bigger than they expected, and the knowledge management piece is the one they hadn’t thought to measure at all.

If you want to see how this works specifically for firms like yours, the AI audit for consulting firms walks through the exact framework we use, and you can also see Omni for consulting firms in more detail before you book anything.

When you’re ready to see your own numbers, Book a 60-min Omni Audit and we’ll walk through it live.

A practical starting point if you’re not ready to book yet

Not every firm is ready to jump straight to an audit call, and that’s fine. If you want a lower-commitment way to understand what deploying an agent actually involves, we put together a worksheet called Deploy Your First Business Agent. It walks through how to pick the right first use case, what data the agent needs access to, and how to measure whether it’s actually saving hours once it’s live. You can grab it from the download page or go straight to the Deploy Your First Business Agent worksheet if you want to start today.

It won’t replace a proper audit of your firm’s specific numbers, but it will give you a realistic sense of what the first 90 days looks like before you commit to anything.

The math that matters

Run the numbers on your own firm for a minute. If a senior partner spends even 15 hours a month on proposal assembly that a Proposal Generation Agent could draft in the first pass, that’s roughly 180 hours a year recovered at whatever your top billable rate is. Add in a Research Agent cutting a three-week research phase down to two or three days across four to six engagements a year, and the recovered capacity starts to look like a full extra engagement’s worth of time. Layer in a Knowledge Agent reducing the rebuild-from-scratch problem across the firm’s existing IP, and you’re looking at the same $80,000 to $300,000 range we see across firms this size, industry ranges being what they are.

None of this requires replacing your team or changing how you sell. It requires removing the repeated, structural work that was never really strategy in the first place, and letting your best people spend their time on the parts of the job that actually justify their rate.

If that number is bothering you the way it should, the next step doesn’t need to be complicated. Read a bit more in our resources on AI deployment for services firms, and when you’re ready to see the real figure for your own firm, Book my Omni Audit and we’ll go through it together, no deck required.