How to measure traffic and leads from AI search
AI assistants now send visitors and enquiries to UAE businesses. Here is how to see that traffic in GA4 and tie it to real leads.
Measure AI search traffic by creating a custom channel in GA4 that groups referrals from ChatGPT, Perplexity, Gemini, Copilot and similar tools, tracking key events such as forms, calls and WhatsApp clicks, and asking new leads how they found you. AI Overviews clicks sit inside Google organic data in Search Console.
Key takeaways
- GA4 lumps AI assistant visits into Referral by default, so build a custom 'AI assistants' channel.
- Clicks from Google AI Overviews and AI Mode are counted inside Search Console web data, not separately.
- Some AI visits arrive with no referrer and show as Direct, so traffic numbers undercount.
- A 'How did you hear about us?' field is the simplest way to capture AI-influenced leads.
- Judge AI search by enquiries and revenue, not sessions alone.
More UAE buyers now ask ChatGPT, Gemini or Perplexity for a shortlist before they ever open Google. Someone in Dubai Marina looking for a family dentist, or a Sharjah trading company comparing freight forwarders, may get a direct recommendation, click one link and land on your site already half convinced.
The problem is that most analytics setups hide this. The visits get filed under Referral or Direct, the resulting WhatsApp messages are not tracked at all, and nobody can say whether the effort you put into Generative Engine Optimization is paying off. This guide sets out a practical measurement setup you can build in an afternoon.
What counts as AI search traffic?
AI search traffic is any visit or lead that originated from an AI-generated answer rather than a classic list of blue links. It falls into three groups, and each one is measured differently.
- AI assistant referrals: clicks from chatgpt.com, perplexity.ai, gemini.google.com, copilot.microsoft.com, claude.ai and similar tools. These usually pass a referrer and can be isolated in GA4.
- Google AI features: clicks from AI Overviews and AI Mode. Google reports these as part of normal web search in Search Console, so you cannot split them out cleanly.
- Unattributed influence: people who read an AI answer, then search your brand name, type your URL or message you on WhatsApp. This never shows as an AI referral, yet it is often the largest share.
How do you see ChatGPT and Perplexity traffic in GA4?
The quickest way is a custom channel group that pulls AI referrers out of the general Referral bucket. Out of the box, GA4 treats chatgpt.com the same as any other website linking to you, which makes the trend invisible.
Step 1: Check what is already arriving
In GA4, open Reports, then Acquisition, then Traffic acquisition. Change the primary dimension to Session source and search for "chatgpt", "perplexity", "gemini", "copilot" and "claude". ChatGPT also adds utm_source=chatgpt.com to many of the links it shows, so check Session source / medium as well.
Step 2: Create an "AI assistants" channel
Go to Admin, Data display, Channel groups, and copy the default group. Add a new channel called "AI assistants" with the condition Source matches regex, using a pattern along these lines:
chatgpt|openai|perplexity|gemini\.google|bard|copilot|claude\.ai|deepseek|meta\.ai
Move the new channel above Referral in the order, otherwise GA4 will classify the sessions as Referral first. Custom channel groups apply to historical data too, so you can see past months straight away.
Step 3: Build an exploration
Create a free-form exploration with your new channel group as rows, and landing page, sessions, engaged sessions and key events as columns. This shows which pages AI tools send people to, which is often a surprise: guides and FAQ pages tend to feature more than the homepage.
Can you track traffic from Google AI Overviews?
Not separately, at least not today. Google includes clicks and impressions from AI Overviews and AI Mode within the Web search type in Search Console, blended with standard organic results.
What you can do is watch for patterns. If a page's impressions rise sharply while its click-through rate falls, and a search for that query shows an AI Overview, the page is probably being cited or summarised. Track a list of your most important question-style queries and record manually, once a month, whether an AI Overview appears and whether you are cited. Our article on what AI Overviews mean for website traffic covers the wider picture.
Which metrics actually matter?
Leads and revenue matter; sessions are only a leading indicator. AI referrals are often low in volume but high in intent, because the assistant has already done the comparison for the user.
| Metric | Where to find it | What it tells you |
|---|---|---|
| Sessions from AI assistants | GA4 custom channel | Whether AI tools are linking to you at all |
| Engagement rate and time on page | GA4 exploration | Whether those visitors find what the AI promised |
| Key events (forms, calls, WhatsApp clicks) | GA4 key events | Whether AI visits turn into enquiries |
| Branded search impressions | Search Console | Indirect lift from people hearing about you in AI answers |
| Self-reported source | CRM or lead form | AI influence that analytics cannot see |
| Citation share | Manual or tool-based prompt tracking | How often you are mentioned for your key questions |
How do you connect AI visits to real leads?
Make sure every way a UAE customer contacts you is tracked as a key event, then carry the source into your CRM. In this market, a large share of enquiries go through WhatsApp or a phone call rather than a form, so a form-only setup will miss most of the story.
- Track every contact action. Form submissions, clicks on tel: links, clicks on wa.me or WhatsApp buttons and email clicks should all be key events in GA4, ideally set up through Google Tag Manager.
- Capture the source in hidden fields. Store the first-touch source, medium and landing page in hidden form fields so they travel with the lead into your CRM.
- Add a "How did you hear about us?" question. Include options such as "ChatGPT or another AI assistant" and "Google search". It is imperfect, but it catches the influence that analytics misses.
- Train the team answering WhatsApp. A short scripted question at the start of a chat, recorded in the CRM, closes the biggest gap.
- Review monthly. Compare AI-sourced leads with other channels on quality and close rate, not just volume.
If your forms do not yet feed a CRM, that is usually a straightforward API integration between your website, WhatsApp tooling and CRM.
Why do AI numbers look lower than expected?
Because a good share of AI-driven visits never carry a referrer. Some apps open links in a way that strips the referrer, some users copy a URL into their browser, and many people simply search your brand name afterwards. All of these land in Direct or Organic search.
Treat your AI assistants channel as a floor, not a total. Look for supporting signals: rising Direct traffic to deep pages that nobody would type by hand, growth in branded searches, and more leads saying they found you through an AI tool. Together these give a more honest view than any single number.
A monthly AI search reporting checklist
- Sessions, engaged sessions and key events from the AI assistants channel, compared with last month.
- Top landing pages for AI traffic and whether they have clear calls to action.
- Branded query impressions and clicks in Search Console.
- Leads whose self-reported source mentions an AI tool, and how many became customers.
- Results of a fixed set of test prompts in English and Arabic across the main assistants, noting whether you are mentioned and which pages are cited.
- Any new AI referrer domains that need adding to your regex.
How Upscaleads can help
We set up GA4 channels, key events and CRM source tracking so you can see what AI search is really contributing, then use that data to decide where content effort should go. Explore our GEO services or get in touch to talk through your current tracking.
Frequently asked questions
Does Google Analytics 4 show ChatGPT traffic automatically?
GA4 records visits from chatgpt.com, but by default it places them in the Referral channel alongside every other linking website. You will see them if you filter by session source, but to track them as a trend you need a custom channel group with a regex rule for AI referrer domains, placed above Referral in the channel order.
Can I see which of my pages appear in AI Overviews?
Search Console does not flag AI Overview appearances separately. Clicks and impressions from AI Overviews and AI Mode are counted within standard web search data. To see which pages are cited, check your priority queries manually or with a monitoring tool, and watch for pages where impressions rise while click-through rate drops.
Is AI search traffic worth tracking for a small business?
Yes, because it is cheap to set up and the visitors tend to arrive with high intent. A custom GA4 channel, tracked WhatsApp and call clicks, and a 'how did you hear about us' question take a few hours to implement and give you a clear signal about whether AI-focused content is producing enquiries.
Why does my AI referral traffic seem so small?
Many AI-influenced visits lose their referrer or happen later through a branded Google search or a typed URL, so they appear as Direct or Organic. The AI assistants channel shows the minimum, not the full effect. Combine it with branded search trends and self-reported lead sources for a more accurate picture.
