Why Traditional ROI Math Breaks for AI Search Visibility
Standard SEO reporting runs on clicks. AI visibility reporting doesn’t have that column yet. Google Search Console’s generative AI performance report shows impressions only, with no click data attached, and it combines AI Overviews and AI Mode into a single number you can’t separate. Rollout has also been uneven; not every property has access to the report yet.
That gap changes the conversation before you set a single KPI. If the platform measuring your visibility can’t confirm someone clicked through, “ROI” needs a different definition than the one your partners are used to from paid search.
What Google’s Free Tools Can (and Can’t) Tell You About AI Visibility
GA4 and Google Search Console are the two tools most firms already have access to, and each covers half the picture.
GA4 can show referral traffic from ChatGPT and Perplexity through its Acquisition filters, but it won’t show you citations or mentions, only the sessions that resulted from someone actually clicking a link. Google Search Console has no way to filter AI Overview impressions out from regular organic impressions; a spike in impressions on a given page is the only proxy available, and even that spike is a blend of AI Overviews and AI Mode with no way to split them apart.
Between the two, GA4 acquisition filters are currently the best free way to see AI referral traffic actually landing on the site. Search Console tells you visibility is happening somewhere in the AI layer; it just can’t tell you where or how much of it converts.
The Five Leading Indicators to Track Monthly
Without clicks, the measurement set shifts to leading indicators:
- Generative AI impressions over time, tracked in Search Console as a trend line, not a single snapshot.
- Branded search volume, the most common conversion path when someone reads an AI answer, then searches the firm’s name directly.
- Direct traffic, with the caveat that direct traffic has always been a mixed bag and AI-driven visits are one contributor among several.
- Intake data, asking new clients where they heard the firm’s name. This remains the only human-verified attribution channel available right now.
- A monthly prompt-tracking scoreboard, covering a fixed list of the questions your clients actually ask.
Track these five together and you get a directional read on whether visibility is moving, even without a clicks column to anchor it.
The Scale Reality: AI Referral Traffic in 2026
This is the number that should anchor every AI visibility conversation with partners, because it’s the most defensible data point available and it isn’t the one most vendors lead with.
Amsive studied 54 sites and found organic search converts at 4.60% against 4.87% for LLM referrals, a gap that is not statistically significant (p = 0.794). On roughly 90% of the sites in that study, LLM referral traffic made up less than 0.6% of total site traffic.
Read plainly: AI referrals convert about as well as organic search does, but there’s barely any AI referral traffic to convert yet. That makes AEO/GEO in 2026 a positioning investment, not an acquisition channel. Firms that treat it as the latter are going to have an uncomfortable quarterly review.
Dedicated AI Visibility Tracking Tools Compared
Several vendors now build dedicated AI visibility tracking, sitting alongside or on top of what GA4 and Search Console give you for free.
Semrush AI Toolkit tracks citations across ChatGPT, Claude, Perplexity, and AI Overviews, priced from $129.95 to $499.95 per month depending on tier.
Ahrefs Brand Radar monitors five separate AI indexes against a database of more than 100 million prompts, starting at $129/month on its Lite plan.
Profound processes more than 5 million citations daily across more than 4 million crawler visits and over 1 million prompts, spanning more than 20 languages; pricing is enterprise and quoted directly.
Conductor offers a unified SEO and AI tracking dashboard under custom enterprise pricing, aimed at teams that want AI visibility sitting next to their existing organic reporting rather than in a separate tool.
Pick based on budget and where you already live. A firm already running Semrush for SEO gains the least friction adding its AI Toolkit; a firm with enterprise budget and a dedicated GEO program gets more depth from Profound.
Connecting AI Visibility to Signed Cases
This is the hardest link in the chain, and also the one partners care about most.
Intake scripts asking “how did you hear about us” remain the only human-verified way to trace an AI-influenced lead through to a signed case; no platform reports this automatically yet. The workaround several agencies have built is a unified dashboard pulling GA4, Search Console, CRM data, call tracking, and ad platforms into one view, so an AI-attributed lead can be followed manually from first touch through intake.
Vendor case studies exist here, and they’re worth citing with the caveat attached. Consultwebs reports a client average of a 266% case increase and 6–8X ROI in the first year using its LawEval tracking layer; treat that as a vendor-reported figure rather than an independently verified benchmark. Separately, a Legora and Ari Kaplan legal AI ROI study found 42% of firms said AI helped them win new work and 94% said it helped them respond to clients faster; that study measures internal legal AI adoption, not AI search visibility, and shouldn’t be conflated with the GEO numbers above.
How to Present AEO/GEO ROI for Law Firms to Partners
Lead with the limitation before you lead with the metric. Tell partners upfront that clicks aren’t in the report yet, then walk through what is: impressions, branded search, intake attribution, and the prompt scoreboard.
Frame the whole program as a positioning bet with a review date attached, not a channel with a guaranteed acquisition number. And bring one real result from your own firm’s year-over-year data rather than an industry projection; a single concrete data point from your own site carries more weight in that room than any vendor benchmark.
FAQ
Can I see how many clicks I’m getting from Google AI Overviews or AI Mode? No. Search Console’s generative AI report shows impressions only.
How do I track ChatGPT and Perplexity referral traffic to my law firm site? GA4 acquisition filters currently show this traffic best among the free tools.
What’s the best free tool for measuring AI visibility? GA4 combined with Search Console impressions, understanding both have real limits.
What are the best paid tools to track AI citations for law firms? Semrush AI Toolkit, Ahrefs Brand Radar, Profound, and Conductor cover the space at different price points and depths.
Does AI referral traffic convert better than organic? Statistically the same: 4.87% versus 4.60%, per Amsive’s study, with p = 0.794.
How much of my website traffic will come from AI right now? Less than 0.6% on roughly 90% of the sites Amsive studied.
How do I prove AEO/GEO ROI to partners? Frame it as a positioning investment, lead with the limitations of current reporting, and track the leading indicators rather than a projected acquisition number.
What should I ask new clients at intake to track AI visibility? Where they heard the firm’s name; it’s still the only human-verified attribution channel.
How often should I check which AI models are citing my firm? Monthly, run against a fixed list of the questions your clients actually ask.
Helpful resources
Discover the power of effective digital marketing.
Sign up to receive 9Sail’s exclusive content and tactical tips, focused on helping law firms grow.
9Sail takes your privacy seriously and will only use your personal information to deliver communications you have requested of us. You can change your preferences at any time.