Four costs most law firms miss, including one that's growing fast
If you have never considered the underlying costs of software investments, then you might not be considering items you don't see on the pricing sheet like human costs and future-proofing your decision. In this first of two parts, we will look at those other costs and in part two, what it actually costs to stay where you are.
The line-item blind spot
Many law firms run a familiar calculation each year. HR and talent leaders look at the cost of their current performance review platform: the per-attorney annual fee, the implementation amortization, the support contract, and conclude the number is reasonable enough to keep running. The reviews get done. The contract auto-renews.
That calculation isn’t wrong, exactly. It’s incomplete. In our experience, the line item usually captures somewhere in the range of 15-20% of what a performance review platform actually costs the firm every year. The rest tends to live elsewhere: in HR calendars, attorney inboxes, IT tickets, partner office hours, and the four-to-six weeks each review cycle slips past its planned close date.
That’s worth naming as its own cost: the cost of standing still. It doesn’t show up on a procurement spreadsheet, and it rarely comes up at the budget meeting. For many firms, it can be five to ten times the size of the line item that does.
None of this is a knock on any one firm's judgment. Every firm is weighing the same kind of tradeoff, and figuring out where these costs actually live is something we work through with clients regularly. It's also what part two of this series digs into: what closing that gap can look like in practice.
Below, we’ll walk through four cost categories many firms miss, three common patterns behind why those costs tend to stick around, and the category that’s growing fastest: one that didn't exist three years ago and, three years from now, will likely be the largest of the four.
Four costs and only one of them lands on the invoice
Performance review platforms cost firms across four distinct categories. The first is visible, and the other three usually aren’t.
Category 1: The visible line item
This is what the contract says: the per-attorney annual fee, plus implementation, plus any premium support tier. For a 350-attorney AmLaw 200 firm running a legacy review platform, this typically falls somewhere between $40,000-$80,000 per year. This is the number procurement looks at first, and it's also the smallest number in this piece.
Category 2: The human hours
This is what the cycle actually asks from the firm. For a typical 350-attorney firm running an attorney-only review cycle, here's a rough picture of the unbilled labor behind it:
Multiply each row by a realistic fully-loaded hourly rate, roughly $75/hr for HR, $400-$500/hr for attorneys, $50/hr for admin staff, and $100/hr for IT. This category often runs three to five times the size of the visible line item for every cycle, every year. For a 350-attorney firm, this category alone typically lands in the $150,000-$220,000 range per year.
Category 3: The hidden decisions impact: outcomes you don't see
This category is harder to calculate, but it may matter most for a firm’s long-term outcomes. A few examples of what it can look like in practice:
Cycle-time delay. Most firms aim for a six-week review cycle and, in practice, often land closer to eight or nine weeks. Those extra two to three weeks are partner attention, feedback delivery, and promotion decisions. It’s hard to put a precise number on what a decision made two weeks late cost a class-year cohort, but it’s rarely nothing.
Reviewer fatigue. When a review tool feels clunky, reviewing partners tend to write shorter, vaguer, more boilerplate feedback. The associate gets a less useful signal and performance trends that should have surfaced can get buried. It’s hard to spot in a single cycle, but across two or three, in can compound into a retention problem.
Costly promotion decisions. This isn't a cost we hear framed often in law firm conversations, but it's worth naming: a promotion decision that doesn't hold up can cost a firm one to two times the associate's annual compensation in attrition and lost productivity. A platform that surfaces clear, calibrated performance signals can help reduce that risk. A platform that leaves those signals scattered across PDFs and Word docs makes it harder to catch early. This one rarely appears on a procurement spreadsheet, but it may be the cost most worth a firm's attention.
Category 4: The AI-capability gap, and it’s widening
Keeping a current platform also means keeping its current capabilities, and that's worth watching closely in 2026, because AI-supported review workflows are moving faster than almost any other part of legal tech right now.
Flo surveyed 96 professional development leaders at law firms between April and May 2026. About 71% said they’re using AI in at least one part of their PD work today. The most common use cases: drafting training material (53%), administrative workflows (47%), performance review support (41%), and coaching and feedback prep (41%). AI has clearly found its way into PD work already.
Narrow the question to performance evaluations specifically, though, and a gap opens up. The NALP Foundation's 2025 “Measuring What Matters” study found that 69% of law firms aren't using AI for performance evaluations at all. The gap tends to track with firm size: at firms with 250 or fewer attorneys, 83% don't use AI in evaluations, and at firms with 251 to 500 attorneys, that number drops to 57%. The largest firms, more than 1,000 attorneys, come closer to parity at 59%, likely because they have the budget and in-house teams to push their platforms past the defaults.
This is not a story about AI replacing reviewers. The PD leaders Flo surveyed were clear about that: the role they see for AI in attorney evaluations is "drafting assistance with human review," by far the largest group of respondents, followed by "data analysis support.” Only seven of 96 respondents thought AI should take a significant role in drafting themselves. Put simply, firms seem comfortable using AI to support evaluations, but not to make them.
The concerns firms raised are real and worth taking seriously: AI-generated language that feels too generic, the privacy of employee data, and the risk of over-reliance becoming, as one respondent put it, "a crutch for human thinking and genuine feedback." A platform that handles AI thoughtfully addresses those concerns directly, through narrow use cases, clear data boundaries, and reviewer judgment that stays in the loop at every step. Skipping AI altogether doesn't solve those same concerns, it just removes the capability entirely, so reviewers never get the option to use it carefully in the first place.
Next, in part two, “Three Common Performance Review Setups, and What They’re Costing Law Firms,” we will dive into the actual costs of the status quo.