Mark Orttung is CEO at Projectworks, delivering Project Intelligence for consulting, drawing from extensive services and tech expertise.getty​Open the dashboard at almost any consulting firm today, and you'll see the same KPIs you saw 10 years ago: billable utilization, revenue per FTE and margin by project. Those metrics aren't wrong. But the benchmarks attached to them were set in a different era, and most firms haven't touched them since.Those numbers were calibrated for a world where output scaled directly with headcount. AI has changed the underlying math. The targets firms have been working toward no longer reflect what good performance actually looks like.What The Old Benchmarks Got Wrong​When I joined Nexient, the team was almost entirely engineers, and we ran at roughly 95% utilization. That number was a direct reflection of the business we were running: staff augmentation, billed by the hour. Over several years, we differentiated the offering. We moved from selling individual engineers to selling startup-like squads of product management, UX and engineering. That shift roughly doubled our average bill rate over seven years. As the work became more strategic, the utilization targets shifted, too. PM and UX teams ran closer to 80%. Engineering stayed around 90%. The number that mattered changed because the business changed.KPI targets encode assumptions about what the business actually is, and those assumptions can go stale.A project manager using AI-assisted scheduling can carry a significantly larger project load than one working manually. A salesperson supported by automation can run more pipelines, generate proposals faster and close more deals in the same time frame. At that point, a 75% utilization rate stops being a reliable signal. It might reflect a high-performing team operating well below its actual capacity. Or it could be masking a serious productivity problem. Without updating the benchmarks, you can't tell the difference.The metrics most consulting firms track are still relevant, but the expected values need to shift. A PM who used to manage two to three concurrent projects can handle four to six, with AI handling schedule conflicts, flagging risks and surfacing resourcing gaps. Gross margins should be climbing by multiple points as AI strips out administrative overhead. Revenue per FTE should be moving from the old $150,000 to $200,000 range toward $300,000 or more. And proposal velocity per salesperson should be doubling.You can't update all of those at once. The better approach is to pick one function, run a pilot from end to end and let the results set the new benchmark.Proposals: A Pilot That Proves The Model​There's a reason proposals make a good starting point. The market has forced the issue.Across Projectworks' customer base, the average number of signed projects per firm increased 266% between 2023 and 2025, from 18 to 66 per year. Meanwhile, the average budget of those signed projects dropped 56%. Firms need to win significantly more work just to sustain the same revenue.In our research into how consulting firms build proposals, we surveyed 45 professional services firms and paired our findings with third-party research. We found that SMB firms submit an average of 91 proposals per year, each taking roughly 24 hours to produce. And much of that time comes from the people you can least afford to pull off client work. Senior staff contribute 47% of proposal development time. That works out to roughly 11 hours of senior time per proposal.Here's how to run the pilot:1. Map The Process​Walk through how a proposal actually gets built in your firm today:• Where does the brief come in?• Who scopes it?• Who writes it?• Who prices it?Most firms have never documented this.2. Measure The BaselinePut numbers against it:• Hours per proposal• Proposals per salesperson per quarter• Win rate• Cost per proposal• Senior staff hours as a percentage of totalThese are the numbers you'll measure against after the pilot.3. Pick The Tools And Run It​In our research, when we asked professionals what they want to improve in their proposal process, the top three answers were content generation (46%), data integration and accuracy (41%) and speed and efficiency (32%). AI can assist directly with all three. A searchable repository of past proposals and case studies can solve the content problem. First-draft generation trained on your firm's winning proposals can address speed. Automated pricing that pulls from your actual rate cards and project history can fix the data accuracy gap. You don't need all of this on day one. Pick the heaviest bottleneck from your process map and start there.4. Check The Metrics​Run the pilot for a quarter, then compare:• Did hours per proposal drop?• Did proposals per person go up?• Did win rate hold or improve?The math is simple. Take your current hours per proposal, subtract what the pilot achieved and multiply by the number of proposals your firm submits per year. That's your recovered capacity in hours. Multiply by your blended billing rate, and you have a dollar figure.Using our research averages as an example, if your firm went from 24 hours to 14 across 91 proposals a year at $200 an hour, that's over $180,000 in recoverable capacity.Why One Pilot Can Change EverythingProposals sit at the intersection of sales, delivery and operations. A successful pilot there creates proof that translates across the business. If your salesperson went from eight proposals a quarter to 16, with the same or better win rate and half the senior staff time, that's a result every other function leader will pay attention to.Once you've proven the methodology in proposals, apply it to project delivery, then resourcing, then reporting.The consulting firms that want to pull ahead don't need to do something radically different. They need to update their definition of a good result. Don't try to reset everything at once. Pick one function. Map it. Measure it. Pilot the tools. Check the numbers. The KPIs your firm has been tracking for years are probably the right ones. The targets attached to them are overdue for a reset.​​​Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?