Written by: Kamil Ponicki, Director of Talent Acquisition, Digital Colliers
If you run a mid-market firm, you already know the number in your gut. Clio's Legal Trends Report puts it in writing: only about 3 of a lawyer's 8 hours get billed on average. That's not a productivity problem in the usual sense. It's a data problem wearing a productivity costume.
The 3-of-8 figure is a lagging indicator. By the time it shows up in your realisation report, the decisions that caused it happened weeks ago, in dozens of small moments nobody wrote down. The interesting question isn't why the ratio is 3-of-8. It's where the other 5 hours actually went, and why your systems can't tell you.
Where the other five hours actually go
When you pull apart a typical fee-earner's day, the missing time tends to fall into four buckets. None of them are new. What's new is how easy they are to measure if you can get your data in one place.
- Genuine admin: file setup, conflicts checks, matter opening, KYC follow-ups, expense entry.
- Non-billable client contact: quick calls, reassurance emails, scope creep absorbed rather than raised.
- Internal coordination: partner check-ins, supervision of juniors, precedent hunting, knowledge management.
- Capture gaps: work that was billable but never made it onto a timesheet, or got written down so late the narrative was too thin to survive review.
The last bucket is the one most firms underestimate. In the pattern I keep seeing, capture gaps alone account for a meaningful slice of the 5 missing hours. The work happened. The client would have paid for it. Nobody wrote it down in time.
The join nobody has run
Here's the exercise worth doing. Take your time-capture data, your billing data, your matter management data, and your calendar and email metadata. Join them on fee-earner and matter. Then look at the gaps.
When you ship that join, three things usually fall out:
- A ranked list of matters where recorded time is materially lower than the calendar and email activity suggests.
- A per-fee-earner curve showing how many hours between an activity and its time entry. The longer the delay, the thinner the narrative, the more likely it gets written down.
- A cluster of admin tasks that repeat across matters and could be templated, delegated, or automated without anyone losing sleep over quality.
None of this requires AI. It's SQL and patience. But most firms have never done it because the data lives in four systems that were bought in four different years by four different people.
The cost of not shipping the join
The cost of inaction here is quiet, which is what makes it dangerous. If your realisation is sitting at industry average, you're leaving real money on the floor every month. In a 40-lawyer firm, even a modest recovery of capture-gap hours is a seven-figure line at the end of the year.
And the ground is shifting under you. ABA Formal Opinion 512, issued in 2024, is explicit that lawyers cannot bill hours that AI actually saved. The SRA issued its own AI guidance back in November 2023. Both point in the same direction: as more work gets AI-assisted, the honest answer to "how long did this take" gets harder to reconstruct from memory. If your capture is already weak, AI adoption makes it weaker, not stronger.
Meanwhile, around 95% of enterprise AI projects never reach production or ROI. Firms that skip the join and jump straight to an AI drafting tool tend to end up in that 95%. The tool works fine. The data underneath it doesn't.
What the operators shipping this in 2026 do differently
The firms getting realisation up aren't buying a new time-recording product. They're doing the boring work first.
- They consolidate time, billing, matter, and activity data into one queryable place, even if it's just a warehouse and a BI tool.
- They measure the delay between activity and time entry, and treat that delay as the primary metric, not hours recorded.
- They rebuild narratives from activity data when a fee-earner forgets, rather than losing the entry entirely.
- They only add AI assistance once the join runs cleanly, so they can actually see what the AI changed.
The 3-of-8 ratio isn't destiny. It's a symptom of a join you haven't run yet. Your practice data already knows where the other five hours went. Someone just has to ask it.