Which Attorneys Should Pilot New Timekeeping Software?
A COO asked us this at a conference recently; she had budget for a pilot and a list already in her head, which was the four attorneys she chases for time at the end of every month; if the software could fix them, she figured, it could fix anyone.
We understand the logic, and we'd still start somewhere else. This is our view, from watching a lot of pilots, and other vendors might tell you differently.
Why the late filers often make a poor first group
Your late filers are usually the people least interested in trying something new, which is a big part of why their time is late in the first place. Put them in a pilot and what tends to happen is that they never really open the software, and two weeks later you're looking at a pilot that stalled.
A stalled pilot is worse than a failed one, because it tells you nothing. You can't say whether the software captured their work well, whether the entries were good, or whether review took two minutes or twenty.
There's a second cost, too: your partners will hear that the pilot went nowhere, and the next tool you bring them will start with that history attached.
Who we'd pick instead
We'd start with attorneys who already record their time and resent how long it takes. They'll use the software, since they're already doing the task it replaces; they have a baseline to compare against, because their existing records are decent; and they'll notice quality, because they know what a good entry looks like.
Within that group, we look for a mix:
A high-volume biller with good habits, who'll give you the most data and a number your partners will believe.
A newer associate, since associates tend to under-record short tasks and captured time shows it quickly.
A partner who complains about time entry, because if the pilot works for them you'll have an advocate in the partner meeting.
One skeptic, meaning someone who'll tell you exactly what's wrong with it; if they come around, the rest of the firm will listen.
Two to five people is enough for a first pilot, which is small enough to support properly and large enough that one person's odd month doesn't decide the outcome. If you can, cover more than one practice area, because a litigator's day and a transactional lawyer's day look quite different to any capture tool.
What to measure
Agree on the measures before anything's installed, and keep it to three:
Hours captured per person, against that person's own trailing average.
Minutes a day each person spends reviewing and fixing entries.
Your billing manager's read on the quality of what gets released.
Give it two weeks before you look hard at the numbers, since the first few days aren't representative for any tool that learns from edits. Our walkthrough of what a pilot looks like week by week covers the mechanics.
Then go back for the late filers
This is the step that answers the COO's real question. Once your first group has a result, you can go to the chronic late filers with something to show them, which is that the on-time people are spending a few minutes a day on this and here's what it found for them.
That's a very different conversation from "please try this new tool." What late filers object to about timekeeping is the time and tedium of reconstructing a day after the fact, and a colleague's two-week result speaks to that better than any demo we could give. We've written more about the underlying problem in how to get attorneys to submit time on time.
Add them as a second wave, with the first group around to answer questions. If your firm wants a partner vote before going wider, it's the first group's numbers and the skeptic's opinion that will carry it, and we've written up how to run that partner conversation separately.
FAQ
How many attorneys should be in a software pilot?
Two to five for a first pilot. With fewer than two, one person's habits decide everything, and with many more than five you're running a rollout without having learned anything first.
Should a pilot include attorneys who are bad at timekeeping?
In the second wave, yes. In the first wave they tend not to engage, and a pilot that wasn't really used doesn't give you much information about the software.
Should a pilot include a skeptic?
Yes, one, and ideally someone who'll be specific about what's wrong. Their criticism improves the evaluation, and their approval persuades colleagues in a way that an enthusiast's doesn't.
How long should a timekeeping pilot run?
Two weeks is the usual length; the first week is setup and adjustment, and the second is the one you measure.
