Blog
Five questions: what named problem it solves and what that problem costs, where client data goes and whether your obligations allow it, what happens when it is confidently wrong, whether it is a feature your existing software will add anyway, and who benefits if you say yes. The firms that do well from AI buy the few tools that solve a real, costed problem, can be used within their obligations, and still save time once the output has been checked.
A specific, named problem in your firm, and what it currently costs. What the tool can do is a demo question. If nobody can state the problem in a sentence, you are buying capability you hope to find a use for, which is how shelfware happens.
If you run or help run a law firm, you have probably been pitched AI three times this month. Some of it is useful, and a lot of it is a familiar tool with "AI" added to the name and the price. Telling the two apart is the job, and getting it wrong costs money and adds risk.
Find out before anything touches a live matter. Client confidentiality and data protection are obligations a firm cannot opt out of. You need to know where the data goes, who can see it, whether it is used to train the model, and whether your professional and regulatory obligations allow any of it. A tool you cannot use compliantly is worth nothing to a firm.
It fails fluently, which in legal work is more dangerous than an error message. Agree the checking process and who is accountable for the output, and test whether the tool still saves time once its output has been properly reviewed. If the review takes as long as the original task, there is no saving.
Often it is a feature. A lot of legal AI will arrive in software you already licence, within a year and at no extra cost, so waiting can beat buying a standalone product your existing supplier is about to make redundant.
Ask it of whoever is pitching. If the honest answer is mostly them, weigh the enthusiasm accordingly. My objection is to buying AI for the wrong reasons, which at the moment are most of the reasons.
Related evidence: How a mid size regional UK law firm grew online enquiries by 190% without rebuilding their website. Earlier in this journey: When your own staff call the website 'frustrating', listen.
Some of it. The worthwhile tools solve a real, costed problem, can be used within a firm's confidentiality and regulatory obligations, and still save time after their output has been checked. The rest is expensive noise, however good the demo.
No. The caution is about buying for the wrong reasons: capability hoping to find a use, a tool that cannot be used compliantly, or a saving that disappears once the output is checked.
Most clients begin with a conversation. No pitch, no hard sell.
Just a straightforward discussion about where you are and whether I can help.
Book a free 30-minute call