Insights · Measurement
A benchmark of 87 law firms found 75% already review or measure how they appear in AI answers. The question for a managing partner is what that answer rests on.
When a managing partner asks the marketing team what the firm is doing about AI, "we're already tracking it" is now the near-universal answer. A new industry benchmark suggests that answer is common because almost every firm in the survey gives it, not because most firms can back it up. The partner who takes the answer at face value is the one who finds out later, in front of the partnership, that nobody on the team can show their work.
A benchmark of 87 law firms, published in its second annual edition by InnovAItion Partners and Gladstone Law Firm Growth Strategies, found that 75% of participating firms are already reviewing or measuring how they appear in AI-generated answers in 2026, and none of the 87 selected "no current plans." In the same benchmark, 91% of respondents held senior marketing or business development titles: CMO, CMBO, director, or senior manager.
75%
of participating firms are already reviewing or measuring how they appear in AI-generated answers in 2026.InnovAItion Partners and Gladstone · 87-firm benchmark · 2026
That is a strong-looking number. It is also, by the report's own description, drawn from "an industry benchmark of participating firms, not a probability sample of all law firms," with results that are "descriptive, with no weighting or imputation." A firm that agrees to fill out a marketing-AI benchmark survey is, almost by definition, a firm already paying attention to marketing AI. A sample built that way was never likely to produce many firms willing to admit they are not watching. Read the 75% as a floor on self-reported activity, not as a working measurement program.
That gap between "willing to say yes" and "able to show their work" is the one a managing partner has to manage. The marketing team's answer is almost always going to be some version of yes. The question worth asking next is what the yes rests on.
Watching AI and measuring it are different things, and separate research shows why the difference matters. In a 1,600-answer study of Perplexity, InterCore found that roughly half the law firms it named by their own firm name appeared in only one of three identical runs of the same question. An AI engine does not give the same answer twice on request. A marketing team that checked ChatGPT once last quarter, saw the firm named, and told the partnership "we're on it" was watching an answer that had roughly even odds of looking different the next time anyone asked.
The same 87-firm benchmark asked a second, broader question. Across everything marketing and business development teams now use AI for, drafting, research, pitch and proposal support, not visibility tracking specifically, 60% report no measurement or only anecdotal measurement of whether any of it is working. Checking an AI answer and proving a drafting tool paid off are different activities, but they draw on the same discipline. The benchmark's other numbers point the same way: 92% of participating firms have implemented an AI policy. Activity is in place almost everywhere in this survey. Proof that the activity works is rare.
60%
report no measurement, or only anecdotal measurement, of whether their AI work across marketing and business development is working.InnovAItion Partners and Gladstone · 87-firm benchmark · 2026
"Covered" is doing a lot of work in that sentence when a marketing team says AI visibility is handled, and the benchmark gives a partner a fair reason to ask what sits underneath it: which AI engines got checked, how many times, over what period, measured against what the firm looked like before. A single spot check answers a different question than the one a partner is asking, which is whether the firm's standing in AI answers is getting better, worse, or staying exactly where it was.
In the same 87-firm benchmark, only 7% of firms have a documented, active AI roadmap inside marketing, while 51% report strong leadership support for AI work. Read together, those two numbers describe a department with backing but without a plan it has written down, which is a normal place for a new discipline to sit in its second year. The question a partner should ask next follows from it directly: is anyone accountable for whether the AI work paid off.
Ask your marketing team to show, not describe, how they track the firm's AI visibility. If the answer is a screenshot of one ChatGPT conversation from a few months ago, that is a spot check, not a measurement. A real answer names which AI engines were checked, how many times each question was run, and what changed since the last time anyone looked.
Selectio measures what AI says about a firm, its practices, and its partners across five AI engines, several runs per question, benchmarked on day one and re-measured every week, so "we're on it" has a number behind it.
Run the free AI visibility check and see how the engines answer for your firm today.
See what AI says about your firm →