Winning clients via ChatGPT: the new first contact for law firms

The first contact has shifted – and most firms don't notice

The path to a lawyer used to run via a Google search, ten blue links and a comparison of firm websites. Today a potential client types their situation directly into ChatGPT: 'My employer dismissed me without notice, what can I do and who do I turn to?' The AI explains the legal situation, names deadlines – and often suggests concrete steps or points of contact. This moment is the new first contact, and it happens before the person ever lands on your website.

The decisive part: a pre-selection is already made in this conversation. If ChatGPT describes three firms in response to the question about a specialist family lawyer in Cologne and yours isn't among them, you're out of the running without noticing. There's no position-11 analysis like on Google, no impressions in the Search Console. You simply see nothing – and that's exactly what makes the shift so dangerous for firms that still rely purely on classic SEO.

How ChatGPT even decides which firm it names

AI models like ChatGPT, Gemini or Perplexity draw their answers from two sources: training knowledge and – with web search activated – from live-fetched pages. For lawyers this means concretely: named is whoever appears in the open web consistently, in a structured way and thematically unambiguously. A specialist traffic lawyer who appears on their website, in lawyer directories like anwalt.de, in local industry portals and in specialist articles always with the same focus becomes a reliable answer for the machine.

Unlike on Google, it's not a single keyword on an optimized landing page that counts. The AI assesses coherence across many sources. If your homepage says 'We advise in all areas of law' while your Google profile says 'criminal law' and a Jameda entry says 'medical law,' no clear picture emerges. The machine then prefers the firm with the unambiguous profile. Specialization is no longer a marketing disadvantage in the GEO era but the prerequisite for being recognized at all.

Important to understand: there's no ad account and no button you can buy your way in with. Visibility in AI answers arises solely through the substance and structure of your public presence. That's the good news – even a small firm with a clear focus can outmatch a large but blurrily positioned firm.

Real questions clients ask the AI – and what they reveal about you

Do the test yourself. Open ChatGPT and type the questions your clients really ask: 'I got a fine notice, do I need a traffic lawyer in Stuttgart?' or 'Which firm specializes in inheritance law and compulsory portion in Hamburg?' or 'Who helps with a warning notice for file sharing?' The answers show you mercilessly whether and how the AI categorizes your firm.

Often one of three things happens. First: your firm doesn't appear at all – then visibility is completely missing. Second: it's named but with the wrong focus ('known for tenancy law' even though your core business is construction law). Third: the AI recommends vague points of contact like 'the local bar association' or 'lawyer search services' because it can't confidently assign a concrete firm. Each of these results is a lost first contact.

These tests cost you nothing and are the most honest firm check you can do. Note down ten typical client questions from your areas of law and check them across multiple AI systems. The result is your personal GEO starting position.

Specialist lawyer title and specialization: your biggest GEO lever

Hardly any industry has such clear signals for AI visibility as the legal profession. Specialist lawyer titles are official, protected qualifications – 'specialist lawyer for employment law,' 'specialist lawyer for family law.' For an AI these are ideal anchors because they're unambiguous, verifiable and linked to an area of law. If your title appears consistently on your website, the bar register and portals, it becomes the strongest building block of your machine discoverability.

The mistake many firms make: they hide the specialization behind a jack-of-all-trades presentation. 'Lawyers for all situations in life' sounds client-friendly but is worthless to the AI. Formulate precisely instead: which two or three areas of law do you concentrate on, in which region, for which client types – private individuals, the self-employed, mid-sized businesses? The sharper this assignment, the more likely the machine names you for the matching question.

Complement this with concrete case constellations instead of abstract paragraphs. 'We represent employees in unfair dismissal claims, termination agreements and severance negotiations' is more tangible for humans and AI than 'advice in individual employment law.' Spell out the situations in which people actually seek help.

Content the AI cites: FAQ, client information, real answers

AI systems love clearly structured answers to concrete questions because they can take them directly into their own answers. For law firms this means: an advice section with real client questions is worth gold. 'How long do I have to challenge a dismissal?' – three weeks, clear answer, understandably explained. Such passages are recognized by the machine, processed and ideally output with your name.

Watch the boundaries of professional conduct rules. You may inform and educate but not make sensational promises of success and not run inadmissible advertising. That fits together well: factual, correct, well-structured legal information is exactly what both the professional conduct rules for lawyers demand and the AI prefers. Whoever writes reputably and precisely automatically optimizes for generative search.

Bet on a question-and-answer format, short paragraphs and unambiguous statements with a date. An article 'Fine catalogue 2026: these penalties apply for speeding' with clear numbers is more likely cited than a general page 'traffic law.' Timeliness is an additional trust signal for legal topics because laws and case law change.

Consistency across all sources: the underrated foundation

The AI assembles its picture of your firm from many puzzle pieces: your own website, Google Business Profile, anwalt.de, Jameda, the bar register, local portals, press mentions. If these sources contradict each other on name, address, areas of law or specialist lawyer titles, the machine's trust sinks – and with it your chance of being named. An outdated entry with an old firm name or wrong address works here like sand in the gears.

So go through your entries systematically. Does the same firm designation, the same address, the same phone number appear everywhere? Are the areas of law named the same everywhere? Is the specialist lawyer title correct and current? This cleanup work sounds banal but is often the most effective first step, because it removes contradictions that make you invisible without you knowing.

Also think about reviews and mentions. When clients consistently connect you with one focus in reviews ('helped me with my divorce'), that reinforces your profile. Such organic signals you can't fake, but you can actively ask satisfied clients for an honest review.

What you can concretely do in the next 30 days

Start with the AI self-test: enter ten typical client questions in ChatGPT, Gemini and Perplexity and document whether and how your firm appears. That's your baseline. Then sharpen your profile on the website: reduce the presentation to your real focus areas, name specialist lawyer titles prominently, describe concrete case situations instead of abstract areas of law.

In the third step you clean up your external entries – Google profile, lawyer portals, bar register – and ensure uniform details everywhere. In parallel you build a small but clean advice section with real client questions and clear answers. Three to five good articles on your core topics are more valuable than twenty superficial ones.

Treat GEO not as a one-off project but as ongoing observation. Repeat the AI test every few weeks and see whether your mentions improve. The firms that start now secure a lead in a channel most competitors don't even have on their radar. The first contact already happens in the AI – the only question is whether your firm is part of the answer.

Data protection and attorney confidentiality: where GEO hits limits

Before you open your firm to AI visibility, clarify one thing: you make yourself visible, not your cases. Everything you publish – FAQ, specialist articles, case examples – must be anonymized so that no one can draw conclusions about concrete clients. The confidentiality obligation under Section 43a BRAO applies unchanged in the GEO era. An anonymized example like "unfair dismissal claim after operational dismissal" is permissible; real names, file numbers or recognizable details are not.

Also think about the lawyer advertising rules. Section 43b BRAO permits factual information about your activity but prohibits sensational or misleading advertising. Phrases like "guaranteed litigation success" or "best lawyer in town" harm you twice over: they're vulnerable under professional conduct rules and the AI classifies such superlatives as untrustworthy anyway. Stay factual, precise and verifiable – that's exactly the tone ChatGPT prefers to pass on.

How an AI-ready firm profile emerges: a four-week roadmap

In week one you take stock: google yourself, ask ChatGPT for "specialist lawyer family law in [your city]" and note whether and how you appear. In parallel, check your entries at lawyer search services, in the bar association's lawyer register and on your own site. Do firm name, areas of law and contact details match everywhere? Contradictions are the most common reason the AI doesn't name you in the first place.

In weeks two and three you write substance: five to eight real client questions per area of law, each with a clear, understandable answer. Not legally convoluted, but the way you'd explain it to someone seeking advice on the phone. In week four you bring it all together: a structured areas-of-law page, an FAQ section and a consistent firm profile. Then you repeat the self-query from week one and see what has moved.

Frequently asked questions from firms about AI visibility

"Do I now have to spend money on ads?" No. GEO works differently from Google Ads. The AI doesn't cite whoever pays the most, but whoever answers most clearly, most consistently and with verifiable expertise. Your most important investment is time for good content, not ad budget.

"Isn't this just hype that will soon pass?" First contact via AI assistants is growing steadily, especially among younger clients and on sensitive topics people are reluctant to raise directly – divorce, dismissal, inheritance dispute. Whoever is visible early here collects a lead that later stragglers can hardly catch up on.

"How do I measure whether it works?" Regularly ask the AI the same questions your ideal clients would ask, and document whether you're named and in what context. Complement this with the simple question on the phone: "How did you come to us?" When the answer increasingly is "ChatGPT" or "AI," you know your work is taking hold.