Most law firm marketing conversations we’re having today involve some version of the same question: What actually matters for AI visibility? It’s the right question to be asking, but it’s hard to answer in the abstract.
So we decided to answer it concretely, in one specific market, with one specific AI tool.
We studied how ChatGPT recommends New Jersey divorce lawyers. We chose a local market deliberately, because patterns are harder to spot when you’re studying thousands of firms across dozens of practice areas in a variety of locations.
By narrowing the focus, we could identify exactly which signals ChatGPT appeared to weigh, which sources it consulted, and where the gaps in most firms’ digital presence actually lie.
Using OpenAI’s o4-mini deep research model, we ran more than a dozen searches at both the county and statewide level. Prompts included variations of “best divorce lawyer in [county],” “top divorce attorneys in New Jersey,” and “high asset divorce lawyers in [county].”
While every search surfaced different names, four patterns emerged consistently. The firms that appeared most often weren’t simply the ones with the best websites or the broadest digital footprint.
They were the ones that had given ChatGPT the most reasons, from the most independent sources, to trust them.
Here’s what we found.
Legal Directories Matter, But Not For The Reasons You Think
In every search we ran, the model turned first to legal directories, particularly Avvo, Justia, FindLaw, and Martindale-Hubbell.
These platforms served as the discovery layer: the place where ChatGPT identified which lawyers to consider before doing any deeper evaluation.
The model, however, was visibly skeptical of legal directories, and in several instances it explicitly noted that these are primarily marketing platforms, not independent credentialing sources.
And yet it used them anyway, because they’re comprehensive and structured. When a lawyer doesn’t appear in these directories, or has an incomplete profile, ChatGPT simply has less to work with.
The practical implication is straightforward: claim and fully complete your profiles on Avvo, Justia, FindLaw, and Martindale-Hubbell. Make sure your practice areas are accurate, your biography is current, and your office locations are consistent across platforms.
Add awards and recognitions where the platform supports it. They’re signals that a major AI model uses to decide whether you’re worth investigating further.
Legal directories may not establish trust, but they still appear to play a critical role in getting discovered.
Independent Credentials Are How ChatGPT Builds Confidence
Once ChatGPT identified a set of candidate attorneys through legal directories, the next step was consistent: it searched for independent evidence of their credibility. This is where the attorneys who surfaced repeatedly separated themselves from those who just appeared in a legal directory.
The attorneys who were recommended most often didn’t have one dominant credential. They had several.
The credentials ChatGPT appeared to weigh most heavily included the NJ Supreme Court Certified Matrimonial Attorney designation, AV Preeminent ratings through Martindale-Hubbell, Super Lawyers recognition, Best Lawyers inclusion, and county bar association involvement.
The New Jersey Supreme Court certification stood out in particular.
ChatGPT repeatedly referenced the certification and often described it as authoritative and state-issued. It appeared to be treated as strong independent evidence of expertise.
The larger takeaway, however, wasn’t about any one credential.
What mattered was corroboration: multiple independent sources pointing to the same conclusion about an attorney’s expertise and standing.
This has a direct implication for how New Jersey divorce lawyers should think about recognition programs. Pursuing the NJ Supreme Court Certified Matrimonial Attorney designation should be a priority for any attorney serious about AI visibility, as it is peer-reviewed, rigorous, and genuinely independent.
Super Lawyers and Best Lawyers carry similar weight because they involve third-party evaluation. AV Preeminent ratings, while tied to a commercial platform, are based on peer review and are widely cited.
County bar association participation matters in local searches for reasons we’ll address below.
The underlying principle is visibility through corroboration. ChatGPT doesn’t appear to trust any single source; it builds confidence by seeing multiple independent sources tell a consistent story about a lawyer’s expertise.
The firms that were recommended repeatedly were present on multiple platforms, not dominating one.
Statewide and County Visibility Are Different Problems
One of the clearest differences we observed was between county-level and statewide searches.
Statewide searches tended to favor attorneys with formal certifications, peer recognition, and presence in statewide directories. The signal that mattered most was the kind of credibility that travels beyond any single county.
County searches behaved differently. Local office presence mattered more.
Mentions of the county in an attorney’s digital footprint carried weight. And county bar association involvement emerged as a meaningful differentiator.
What this means in practice: if a particular county is important to your practice, “Serving Monmouth County” in the footer of your website isn’t enough. ChatGPT appears to look for independent corroboration of local presence, not self-reported coverage claims.
That means participating in county bar programs, joining local lawyer referral services where they exist, and making sure county-specific involvement appears throughout your digital presence, not just on your website.
Niche Practice Areas Are the Biggest Opportunity Right Now
Here’s where the opportunity is clearest, and where many firms are leaving the most ground uncovered.
When ChatGPT searched for attorneys that specialize in high-net-worth divorces, it frequently ran into a problem: there weren’t enough independent sources to evaluate the claimed expertise. In those cases, the model fell back on law firm websites.
Throughout our research, ChatGPT treated law firm websites with visible skepticism, explicitly noting that websites are marketing materials, not objective sources.
And yet, in the absence of anything else, it used them. The implication is that the bar for establishing niche credibility is lower than it is for a general divorce practice, simply because fewer independent authority signals exist.
Firms that invest in building genuine content depth around a niche specialization can establish themselves as the most credible option in that space by default.
For high-net-worth divorce in particular, the gap was striking. It’s a competitive and lucrative practice area, but independent, authoritative content around it is sparse.
A firm that develops educational articles, detailed FAQs, case study examples, external references, and earns mentions in third-party sources, has a meaningful opportunity to become the answer ChatGPT reaches for when the question is about high-asset divorce in New Jersey.
The same logic applies to collaborative divorce, mediation, LGBTQ+ divorce, and other defined sub-specialties. Don’t stop at a single practice area page.
Build out content that educates, answers real questions, and ideally earns citation from sources outside your own website.
What This Means for Your Firm
The consistent finding across all of our research: The attorneys who surfaced repeatedly in our searches weren’t dominating one platform; they were present across many. They were first discovered through legal directories, then evaluated against credentials that were independent or peer-reviewed.
When independent evidence wasn’t available, the model fell back on the marketing platforms it initially distrusted.
The attorneys building visibility in AI aren’t doing anything exotic. They’re earning credentials, staying active in bar associations, participating in recognition programs, and building content depth in the practice areas they actually want to be known for.
If you’re not sure where to start, the answer is probably simpler than you think: audit where you appear online outside of your own website. Count the independent, third-party sources that mention you by name and describe your expertise.
If that list is short, you have your roadmap.
The firms that will be recommended most consistently by AI over the next several years are building that evidence base now.