How to Tell If AI Is Actually Sending You Customers in 2026
Most of your AI traffic is invisible in Google Analytics, and it is some of your best. Here is how to see it, and the handful of numbers worth tracking.
- Most AI search traffic is invisible: a large share of it lands in analytics as 'Direct,' so it looks like it came from nowhere.
- A one-time Google Analytics 4 setup, a custom channel group, makes ChatGPT, Perplexity, and Gemini traffic show up by name.
- AI visitors tend to convert better than search visitors, because they already did their research inside the chat before they clicked.
- The headline metric is share of voice: how often AI assistants recommend you versus competitors, tracked against a fixed set of buyer questions.
Here is the uncomfortable truth: if you only look at Google Analytics, you are probably undercounting the customers AI sends you, badly. A large share of AI search traffic arrives with no referrer and gets filed as "Direct," so it looks like people typed your address from memory. The good news is that AI visitors are often your highest-intent traffic, and with a one-time setup you can finally see them. Here is how to measure whether ChatGPT, Perplexity, and Google AI Overviews are actually sending you business.
How do I know if AI is sending me customers?
You look in two places: who shows up on your site (traffic), and who shows up in the answers (visibility). Traffic tells you whether AI clicks are arriving and converting. Visibility tells you whether the assistants recommend you in the first place, even when no one clicks. You want both, because AI answers are increasingly zero-click: a customer can decide you are the one to call without ever visiting your site.
Step 1: Stop AI traffic from hiding in "Direct"
When someone clicks a link inside ChatGPT or another assistant, many apps strip the referrer, so analytics cannot tell where the visit came from and dumps it into Direct. The fix is a one-time setup in Google Analytics 4: a custom channel group with a rule that recognizes AI sources by their domain, ranked above Direct and Referral so it catches them first.
The rule matches the Source against the AI domains you care about:
chatgpt.com|chat.openai.com|perplexity.ai|gemini.google.com|claude.ai|copilot.microsoft.com|grok.com
For links you control, such as a link inside your own content or a profile, add UTM tags so the source is never ambiguous. Then check your numbers weekly. Many businesses are surprised to find AI is already a real, growing slice of their traffic once it stops hiding.
Catch the hidden path too
Step 2: Track your share of voice
Traffic is only half the story. The metric that matters most in AI search is share of voice: how often the assistants name or cite you versus your competitors, across the questions your customers actually ask. It is the closest thing AI search has to a ranking.
Measuring it is straightforward. Build a fixed list of 20 to 50 buyer questions ("best [your service] in [your city]," "who should I hire to [job you do]"), ask them across ChatGPT, Perplexity, Google, and Gemini on a set schedule, and record whether you appear, what the assistant says, and who it names instead. Do it by hand to start, or use a tool to automate it as you scale.
The tools worth knowing
You can begin with nothing but a spreadsheet and a weekly habit. When you want to automate, these are the categories of tracking tools, with the trade-offs.
| Tool | Best for | Trade-off |
|---|---|---|
| Spreadsheet plus weekly prompts | Getting started for free | Manual, but teaches you what to watch |
| Otterly | A quick, low-cost baseline | Lighter on deep analysis |
| Peec AI | Agencies and teams | Built for managing many brands |
| Nightwatch | SEO and AI tracking together | Rank-tracker first, AI added on |
| Profound | The most thorough tracking | Priciest, aimed at larger budgets |
The handful of numbers that actually matter
Ignore vanity metrics. For a small business, four numbers tell the whole story:
- Share of voice. How often you are mentioned or cited across your buyer questions, versus competitors. The trend matters more than any single week.
- Citation rate. When you are mentioned, does the AI actually link to your site, or just name you? Links send traffic; mentions still build trust.
- AI-referred conversions. Of the AI traffic you can now see, how many become leads or sales. This is where AI's high intent shows up.
- Sentiment and accuracy. What the AI says about you, and whether it is right. A confident, wrong answer about your hours or services costs you customers, and it is fixable.
What to do Monday
Measurement does not have to wait for a big project. This week: set up the GA4 channel group so you can see AI traffic, write your list of 20 buyer questions, and run them once across the major assistants to get a baseline. Next week, run them again and watch what moves. That simple loop is the whole discipline, and it turns AI search from a black box into something you can actually manage.
If you would rather have it set up and reported for you, that is part of what our Answer Engine Optimization work delivers: we track what every assistant says about you, fix what is wrong, and report share of voice alongside the traffic and conversions it drives. Book a free audit and we will show you your starting line.
Frequently asked questions
How do I track AI search traffic in Google Analytics?
Create a custom channel group in GA4 with a rule that matches AI sources by their Source domain (chatgpt.com, perplexity.ai, gemini.google.com, claude.ai, and others), and rank it above the Direct and Referral channels. From then on, AI traffic shows up as its own channel instead of hiding in Direct.
Why does AI search traffic show up as Direct?
Many AI apps strip the referrer when a user clicks out, so analytics cannot see where the visit came from and files it under Direct. A custom channel group catches the sources that do pass a referrer, and UTM links cover the ones you control, such as links inside your own content.
Does AI search traffic actually convert into customers?
Often better than regular search. Because people do their research inside the AI chat and click only when they are close to deciding, AI-referred visitors tend to arrive pre-qualified. Studies in 2026 show AI referrals converting at or above organic search rates, though the total volume is still smaller.
What is share of voice in AI search?
It is the percentage of relevant AI answers that mention or cite your business versus your competitors, measured across a fixed set of buyer questions. Because AI answers are often zero-click, share of voice is the closest thing to a ranking in AI search, and the metric most worth tracking over time.