A few days ago, a lawyer on our firm’s internal lawyer chat channel posted this comment: “I thought you'd enjoy knowing how my 11 a.m. appointment found me: He did a deep search with the latest version of ChatGPT. It searched for everything about our firm over the past five years. He said it gave him a five-page readout and even included a summary of our firm's reviews that was glowing. He said everything was so awesome that he scheduled a full hour appointment with me to talk about his immigration issues. Pretty cool.”
Obviously, this was nice to hear. But is this just a novelty, or does it reflect real change in how clients are shopping for legal services? Are artificial intelligence (AI) models altering how consumers shop for legal services? Should lawyers be rethinking how they approach online marketing?
The short answer to all three of these questions is “yes.” And honestly, it’s not hard to see the appeal of the models. Search engines just return links while AI models return judgments with rankings and summaries.
Before answering why many people prefer AI for the task of finding a lawyer, it helps to know which products consumers are likely using. There are six major AI models typically being used for tasks like this. They are OpenAI’s ChatGPT, Anthropic’s Claude, Gemini from Google, CoPilot from Microsoft (included with the Microsoft Edge browser), Grok from xAI and Perplexity.ai. All the products now draw from the web which is different than when generative AI products started launching 2022 and these tools could only answer based on a version of what was on the Internet many months before.
Unlike a search engine, someone using an AI model can get a lot more specific. They might describe their legal difficulty, educate themselves on whether they might have legal options and then ask the model to recommend someone in that specialty in their geographic area. They then might do a deeper dive on a listed firm and get the AI’s in-depth feedback, much like the client I described above did regarding my firm.
You might notice that firms that show up high in Google with a simple query (e.g., “personal injury law firms in Jackson, Tennessee”) also tend to show up near the top when you ask AI “What are personal injury firms in Jackson, Tennessee?” A key difference will likely be that firms that are paying to be listed at the top of a Google search page are nowhere to be found. This is the big question people are asking about how Google can straddle staying on top in AI while seemingly killing the goose that laid the golden egg, leaving many to wonder whether AI results will eventually be shaped by companies willing to pay for being mentioned. Companies giving into that temptation risk their products’ reputations for quality being diminished.
A user might also use superlatives like “who are the best” or “who do you recommend” and hope the AI can filter out the firms with the top reputations. And the models may oblige because they access resources like Super Lawyers, Best Lawyers in America, Chambers, Lawdragon, etc. Right now, ratings on sites like Google and Avvo have a lot of influence, but the AI tools generally don’t consider consumer-facing review platforms in their answers to these questions. They’re typically using the ranking sites, firm websites, legal industry media and legal directories to answer those types of queries. That’s not to say that the reviews never matter. A user may want to know about a lawyer’s client reviews and get an answer that synthesizes them.
Deep research and reasoning skills in the new AI tools may also change how consumers evaluate a firm. The real-world example I noted above where a person simply asked ChatGPT’s deep research model to review our firm for the last five years, synthesize reviews and media coverage, summarize bios and discuss our capabilities relating to their type of legal case is not some far off future. It’s rapidly becoming the norm for how people educate themselves as consumers. The five-page report written precisely for the user without the puffery of a website is, not surprisingly, going to be appealing to a lot of people.
Is this good news for lawyers? Well, as a lot of lawyers say, it depends. Lawyers who have benefited from diligent search engine optimization (SEO) work may find that they get much less return on their time and consultant spending. And what’s old is new again. In the past, lawyers who published, spoke regularly, had niche expertise and were recognized by their peers in rankings, generally had the edge. And then a lot changed when lawyers seemingly could jump to the top by getting lots of ratings, deploying sophisticated SEO, making viral social media posts and paying for expensive Google AdWords placement, even if they were not actually highly regarded for subject matter expertise or as a leader in the field. But now we’re back where a lawyer’s reputation over the course of a career may matter a lot after all.
Of course, there are still questions about the reliability of AI. Most lawyers are familiar with examples of AI tools hallucinating––making up case citations in submitted legal filings. Do these tools do the same thing when someone asks an AI to write a biography of a lawyer or synthesize their ratings into a summary? It is certainly true that no AI tool has eliminated hallucinations. But they’re way better than they were in 2023. In 2022 and early 2023, I used to post on LinkedIn about various tests I ran on the competing AI models to see how accurate the answers were. One test I ran was to ask the AI to write a biography of me. And the results were not great. There was a lot of mixing up facts with other lawyers and some outright inventing of biographical details, like quotes about why I went into immigration law. But it’s a very different story today. The same query posted in the various AI models returned excellent biographies that were completely accurate.
What can lawyers do to optimize for generative AI discovery? For one, focus on publishing high-quality content. Blogs are going to again be a premiere place for lawyers to focus their writing efforts. Websites were viewed by many as not so important because of social media’s dominance, but that may change again as bios, case studies, client testimonials, practice area descriptions and posted news about the firm become content easily parsed by AI models. Posting video content can still help (Google’s Gemini is ingesting YouTube’s content, for example) and posting a transcript along with the video content helps. And lawyers will need to work on visibility in peer review sites and maintaining positive client online reviews as AI models will consider them in their answers too.
The future of client acquisition for a lawyer is going to depend on whether clients’ new, trusted advisor tells them to consider that lawyer. That advisor is now an AI and once again the whole ballgame has changed for online legal marketing.
