Why legal tech onboarding fails in week two

If you want to know whether a legal tech rollout will succeed, don’t look at the launch. Launches nearly always go well. There’s a training session with decent attendance, a few enthusiastic early users, a vendor who answers emails within the hour and an internal announcement with a slightly too cheerful subject line. Everyone leaves week one feeling good. Then week two arrives, and quietly, without anyone deciding anything, usage starts to fall away. Having been on the vendor side of more rollouts than I can count, I’ve come to think week two is where adoption is actually decided. It’s also the week almost nobody plans for.

Week one is a honeymoon you paid for

Week one is artificial, and that’s not a criticism. It’s designed to be. People are using the tool because they’ve just been shown it, because their calendar told them to and because someone is watching. The vendor’s customer success team is at full attention. The innovation lead is walking the floor. Questions get answered in minutes.

None of that is how the tool will be used for the next three years. Week one measures curiosity and goodwill, which are real but cheap. It tells you almost nothing about whether the tool has earned a place in the work. The mistake is reading week-one numbers as a signal of adoption, when they’re mostly a signal that the launch was well organised.

Week two is when the real work comes back

By week two, the normal pressures of practice have reasserted themselves. The matters that were put on hold for training are now urgent. A completion has moved forward. A client wants a call. The associate who was happy to experiment on Tuesday of week one is now trying to get a report out by six, and under that kind of pressure people reach for whatever they already know works.

This is the real test, and it’s a much harder one. The tool is no longer competing against nothing. It’s competing against the existing way of doing the work, which has the enormous advantage of being familiar, and it’s being judged by someone who is tired and short of time. A tool that was delightful in a training room can feel like an obstacle at 5.30pm on a deadline day, not because it’s got worse, but because the circumstances have.

The first bad experience is decisive

Here’s the pattern I’ve seen most often. Somewhere in week two, a user tries the tool on a live matter and something goes slightly wrong. The document won’t upload because it’s a scanned PDF. The output misreads a defined term. The login has timed out and the single sign-on loops. Nothing catastrophic, just friction at the wrong moment.

In week one, they’d have asked someone. In week two, there’s nobody obvious to ask and no time to ask them, so they do the work the old way and make a quiet mental note: not reliable. That note is extraordinarily hard to erase. Lawyers are trained to be risk-averse, and one bad experience under pressure is enough evidence for most of them to stop trying. You rarely lose users to a bad tool. You lose them to a single bad afternoon that nobody was there to rescue.

Nobody owns week two

The reason week two goes unplanned is structural. The vendor’s onboarding programme usually ends with the training. The innovation team has moved on to the next pilot. The supervising partner, who never really used the tool, isn’t asking about it. The users themselves have no one whose job it is to notice they’ve stopped.

So the most important stretch of the whole rollout sits in a gap between owners. Everyone did their part: the vendor trained, the firm launched, the users attended. And the tool still dies, because the part that mattered was nobody’s part. I’ve watched firms run thorough, well-resourced launches and then lose the adoption entirely in the fortnight after, with no single person able to say what went wrong.

Design for week two, not launch day

The fix isn’t complicated. It just requires treating week two as the main event rather than the aftermath.

Start with live matters, not training exercises. If people’s first real use of the tool happens on something they actually need to finish, the tool earns its place in the work from day one instead of having to break in later. Keep support visible after the launch: a named person, a short daily window when someone is reachable, a channel where a question gets answered before the deadline passes. Schedule a check-in around day eight with every user, short and specific, asking what they tried and what got in the way. That conversation will surface the small frictions while they’re still fixable.

Measure the right thing. Training attendance and week-one logins are vanity numbers. The number that predicts the future is how many people used the tool, unprompted, on a real matter in week two and week three. And get supervisors using it themselves, because a junior will keep going with a tool their partner understands and abandon one their partner has never opened.

Most legal tech doesn’t fail because it doesn’t work. It fails because it was introduced as an event when it needed to be introduced as a habit, and habits are formed in the boring weeks after the excitement wears off. If you’re planning a rollout, spend less energy on launch day and far more on the fortnight that follows. That’s where your users decide, usually without telling you, whether the tool is part of how they work or just something they were shown once.