Why Hourly Rate and RFP Benchmarks Miss the Point
Just when I thought clients were starting to get it, a new group of market entrants has launched a rate benchmarking tool and sells it like manna from heaven.
For the past decade, I've told clients the same thing: the hourly rate is table-stakes information. Yes, it's one half of the hours-times-rate equation, so it matters in an arithmetic sense. But it is the least useful half of that equation, and a client who optimizes for it is optimizing for the wrong thing.
Look at What the Math Says
Assume a firm's rate is $1,000 an hour. You negotiate a 10 percent discount and bring it down to $900. Congratulations. Now assume the task takes one hour.
• Option One: 1.0 hour at $1,000 = $1,000.
• Option Two: 1.1 hours at $900 = $990.
Your 10 percent discount just bought you $10, or 1 percent. One extra tenth of an hour, the standard six-minute billing increment, wiped out 90 percent of the savings. Stretch the task to 1.11 hours and the discount is gone entirely. You won the negotiation and lost the invoice.
Here's the problem: rates are a fixed, visible number, and time is a variable, largely invisible one. The rate is what the client sees on the engagement letter. The hours are what the client pays for. A discount on the first without any control over the second is a rounding error with good optics.
Time is the Manipulable Half
I'm not suggesting anyone is acting in bad faith. In my experience, inadvertence explains far more than greed. But the incentives are what they are. Hours depend on staffing, how a task is broken up, how many people attend the call, and whether the first draft comes from a senior partner or a second-year. Every one of those choices moves the total far more than a five or ten percent swing in the rate.
That's why the question clients should be asking is not "What does this firm charge per hour?" but "How long does this kind of task take, who does it, and what should it cost?" Benchmark time, tasks, and staffing, and you learn something about value. Benchmark the rate, and you learn what the firm puts on its price list.
The AI Era Makes it Worse
Now add generative AI (GenAI) to the picture. If a machine can produce a competent first cut of a document in seconds, why on earth would anyone care whether the human hour behind it is priced at $900 or $1,000? The relevant question is whether a human hour should be on the invoice at all, and if so, how many. Rates assume the hour is the unit of value. GenAI is exposing, faster than any consultant ever could, that it never was. Clients buy knowledge, skill, and judgment. They don't buy time.
A rate benchmark in this environment is a very precise measurement of a unit that is rapidly losing its meaning.
The RFP numbers are worse.
If the rate benchmark is shaky, the pricing data pulled from RFP tools is shakier. Think about what those numbers represent. They aren't what firms normally charge, and they aren't what the work is worth. They're what a firm was willing to quote to win the work.
There are a dozen reasons a firm might bid low: it wants the logo on its client list, it's filling a lull in a practice group, it's trying to get in the door for the bigger matter, or it's responding to a procurement process that rewards the lowest number. None of those reasons says anything about the value of the work.
What an RFP rate tells you is the price of a firm's appetite on a given day.
So when a tool tells a general counsel that the "market rate" for a mid-level partner is a particular figure, ask what's inside that figure. It's a collection of bids, shaped by desperation, strategy, and negotiating leverage, and it's presented as if it were a price discovered in a functioning market.
Two Weak Signals Don’t Make a Strong One
Here's where it gets truly unhelpful. Combine the two. Take a rate that says little about value, built from bids that say little about the market, and multiply it by hours that nobody is measuring against the task. What comes out the other end?
Not insight. A bigger error bar.
Each factor is marginally useful on its own. Multiply them, and you don't get a more reliable answer; you get a number that looks authoritative and rests on nothing firm. Garbage in, garbage out, now with a dashboard.
What To Do Instead
I'm not saying clients should ignore rates. I'm saying rates should be the last thing they look at, not the first. Here's the order of operations I'd suggest:
1. Benchmark time before price. Ask how long comparable tasks take on comparable matters, using actual time entries, not estimates in a pitch deck.
2. Benchmark staffing. Who did the work, at what level, and was that the lowest efficient biller for the task?
3. Budget at the task level. A matter-level number hides everything. A budget by phase and task lets you see variance as it happens.
4. Ask what GenAI changes. Any firm that can't tell you which of its tasks are now faster, and what it plans to do with the time, hasn't started thinking about this.
5. Treat RFP pricing as a data point about the bidder. Use it to understand how a firm wants to be perceived, never to define the market.
6. Pay for outputs. Where work is routine and repeatable, price it as a flat fee. Where it's bespoke and high-stakes, pay a premium for the judgment. In both cases, the price should track what the client receives, not what the clock recorded.
None of this means lawyers should earn less. Strip out the low-value work and the high-value work that's left can justify strong rates. That's the point. Rates and value aren't the same thing, and the data that matters is the data that connects them.
The Bottom Line
A rate tells you what an hour costs. It tells you nothing about what the hour is worth, or whether you needed the hour in the first place. In the age of AI, that gap is only going to get wider, and the clients who figure that out first will be the ones who stop buying time and start buying results.
Joe Tiano is the Co-Founder and President of Legal Decoder. After nearly 20 years as a partner at an Am Law 100 firm building a cross-border finance practice, Joe co-founded Legal Decoder to close the data gap that kept legal billing from being understood or improved. He writes and speaks widely on legal analytics, AI, and legal operations. As President, he drives the company’s market strategy and its mission to bring real intelligence to legal spend.