When clients ask AI who to instruct, will your firm be on the list?

A managing partner said something to me recently that stuck. A new client had told them, quite casually, that they chose the firm because ChatGPT mentioned its name. It felt like a win, but it’s also a glimpse of what’s already changing.

More clients are starting their search by asking AI the questions they used to type into Google or ask friends. “Who’s the best conveyancing solicitor near me?” “Which employment lawyer is good for settlement agreements?” “How do I challenge a will?” The difference is that AI often gives an answer that feels final. A shortlist, a summary, sometimes even a recommendation. If your firm isn’t included, you can lose work without ever seeing the enquiry, and you may never know why.

For years, firms have understood how to win online. You build a good reputation, you keep directory profiles strong, you gather reviews, and you run a website that turns interest into contact. Search engines helped clients find you, then clients compared options and made a choice.

AI has changed the order of events. The client may never reach the comparison stage. Instead of ten links, they get a tidy response and move on. That means you’re no longer only competing to rank well, you’re competing to be included at the point where the client is forming a view of who is credible.

Why it matters commercially and reputationally

That’s a commercial issue, but it’s also a reputational one. You see, you are in a sector that digital marketers call ‘Your Money or Your Life.’ Mistakes by consumers/clients choosing suppliers (such as law, finance, health) in this category can literally cost them their lives, as they know them. So, as a result, people are very cautious when selecting a law firm, a wealth advisor, or a surgeon for example. Trust is everything, and if things don’t line up, well, you get the picture. 

If trust signals are not aligned, I  see three problems crop up again and again.

First, pipeline can shrink quietly. If AI keeps naming two or three competitors for your core service in your core location, your firm becomes invisible at the decision point. You can have a solid website and a strong team and still lose out because you’re not being surfaced in the first place.

Second, firms can be described inaccurately. AI can pick up old information and present it very confidently. An office that moved, a practice area you no longer offer, a partner who has left, the wrong fee earner linked to the wrong service. Most firms aren’t checking what these tools say about them, so errors can sit there, repeat, and spread and basically this means the false or outdated information is baked in. If left unchecked it will become a problem.

Third, trust is being assembled from fragments. AI pulls from whatever it can find, often across your website, directory profiles, articles, reviews, and third party mentions. When those sources don’t line up, confidence drops. When there are gaps, the model can fill them with guesswork and you really do not want that.  That’s why this can’t be written off as “just marketing”. It touches risk, reputation, and business development. All crucial for YMOYL sector firms.

What is AI actually looking for?

The useful thing to understand is that AI is trying to do what a cautious client would do. It’s looking for signals that reduce uncertainty. Firms tend to show up more often when they are very clear about who they are, specific about what they do, and backed by evidence. Consistency matters too, because AI is stitching together a picture from multiple places, and it finds contradictions quickly.

So what does a sensible response look like? You don’t need a giant project or a shiny “AI strategy” document. The simplest way to think about it is this. Treat AI like a new referrer.

Firms already manage referrers. You nurture introducers, you keep directory listings accurate, you maintain your Google Business Profile, and you pay attention to reviews. AI needs the same level of care, because it’s now sitting upstream of the enquiry.

A real-world example of how firms are being filtered out

Here’s a simple example I now use with partners because it makes the shift very clear.

A client types into ChatGPT:

“Who are good employment solicitors in Manchester for settlement agreements?”

The response isn’t a list of links in AI search.  It’s a short paragraph naming two or three firms, explaining briefly why they’re suitable, and often mentioning things like employer-side experience, seniority of the team, or volume of similar matters handled. One firm is described as “specialist in employer settlement agreements”, another as “well-known for advising SMEs”. A third firm might be mentioned as an alternative.

What’s striking is not just who appears, but who doesn’t. In one recent check, a mid-sized firm with a strong local reputation and a busy employment team did not appear at all. When we looked at why, nothing was “wrong” in the traditional sense. The firm had good rankings, solid reviews, and a well-designed website.

But their employment pages were broad rather than specific. Settlement agreements were mentioned only in passing and fee earners weren’t clearly linked to that work. Older directory profiles still referenced different locations. Meanwhile, a smaller competitor had a clearly written settlement agreement page, named specialists, published short explanatory articles, and consistent profiles across directories.

AI didn’t judge quality, it judged clarity and confidence.

Step one: assign ownership for AI representation

Start by making one person responsible for your firm’s AI representation. Not AI adoption in general, and not experimentation with tools, but one clear job: what does AI say about us, and is it accurate? This usually sits best between marketing and risk, with enough authority to coordinate updates across the website and profiles or to give sign off to outside GEO agencies.

Step two: run a monthly check using real client questions

Then run a simple monthly check using real client questions. Pick ten to fifteen prompts a client would genuinely ask, and include your top practice areas and locations. You’re looking for whether the firm is mentioned, which competitors show up, and whether anything said about your firm is wrong or misleading. Keep a record. Over a few weeks, patterns appear, and those patterns tell you where to focus.

Step three: improve service pages so they reduce uncertainty

Next, look hard at your service pages. Many law firm websites are written for peers or referrers. Clients and AI respond better when the page reduces uncertainty. That usually means being clear on who the service is for, what typical scenarios look like, what your process is, and what makes your team credible. It also means addressing the practical points clients care about, like timeframes, what information you need upfront, and what tends to affect cost. This isn’t about writing more. It’s about writing in a way that helps someone choose you with confidence.

Step four: strengthen evidence, not adjectives

At the same time, strengthen evidence rather than adjectives. “Leading” and “trusted” don’t carry much weight on their own. AI tends to respond to verifiable signals, such as accreditations, memberships, recognised rankings, speaking and training, published commentary, and case studies where you have consent. Even anonymised examples can be powerful if they show the type of work you do and the outcomes you achieve. Clear authorship helps too. When a piece is signed by a solicitor who genuinely works in that area, it’s easier for both clients and machines to trust it.

Step five: make your firm details consistent across the web

Finally, get your basics consistent across the web. Make sure addresses, phone numbers, regulated entity names, staff roles, and practice area descriptions match across your website and key profiles. If there are old pages that conflict with your current positioning, tidy them up. If there are neglected directory entries, either refresh them or remove them where possible. When those details don’t agree, AI confidence drops, and it’s less likely to recommend you.

What I think firms should do next

My view is that every firm should add one question to its quarterly commercial review. That question is :

“If a prospective client asks AI who to instruct for our core matters in our core locations, what answer do they get, and is it accurate?”

If you can’t answer that today, you’re operating with a blind spot. In a competitive market, blind spots cost money.

The good news is that this is fixable, and it’s not mysterious. It’s disciplined work. Make your expertise clear, make your evidence easy to find, and make your presence consistent. AI search is already influencing who gets considered. The choice is whether your firm is introduced properly, or whether a competitor gets the introduction instead. It’s all about trust, but then, you already know that.