By Steve Merrill, Founder of WRKNG Digital | July 4, 2026
How Do You Actually Know If AI Is Recommending Your Store?
You can't see it in Google Analytics. There's no AI referral tag in your traffic sources. When ChatGPT or Perplexity sends someone to your Shopify store, that visit shows up as direct traffic — completely invisible in the data you're already watching. The measurement problem is real, and most stores are flying blind.
This isn't a minor gap. It's the core reason brands are underinvesting in AEO — answer engine optimization. They can't prove it's working. So they don't try.
Why Traditional Analytics Won't Show You AI Traffic
AI assistants don't pass referrer data the way Google does. A user asks ChatGPT "what's the best sustainable yoga mat brand?" — ChatGPT cites your store — the user clicks the link. Your analytics records a direct session. You see nothing that ties that conversion to an AI recommendation.
This changed significantly after ChatGPT launched its Ads platform in early 2026. Now there's a paid channel where AI citations are tracked and billed. But organic AI mentions — the recommendations happening right now across Perplexity, Google AI Mode, Gemini, and standard ChatGPT — still produce no clean referral signal.
So you need different methods. Not one. Several.
4 Methods That Actually Work
1. Manual Prompt Testing
The bluntest instrument. Also the most revealing.
Go to ChatGPT, Perplexity, and Google AI Mode. Ask the prompts your customers would actually use. "Best running shoes for flat feet under $150." "Most sustainable protein powder for athletes." "Top-rated skincare for sensitive skin that ships fast." Whatever your category is — ask it the way a real person would.
Track who gets cited. Not your brand specifically — track who gets cited at all. Those are your competitors setting the benchmark. I've run this across 40+ store audits, and the pattern is consistent: brands with detailed product descriptions, genuine customer reviews, and structured data show up. Brands without those things don't, regardless of how good the product actually is.
Run this weekly. Log the results in a simple spreadsheet. Date, platform, prompt, brands cited, your brand cited (yes/no). Over time, you'll see movement — or you'll see that you're stuck.
2. Brand Mention Tracking Tools
Tools like Otterly.ai and Semrush's AI brand monitoring feature are built specifically for this. They run automated prompts across AI platforms on a schedule and track when your brand name appears in the responses.
Otterly is the more specialized option for AEO tracking. You set up your brand, your competitors, and a library of relevant prompts. It runs them repeatedly across different AI models and shows you your citation rate — what percentage of AI responses that mention your category include your brand name.
A 10% citation rate in a competitive category is meaningful. Going from 4% to 9% over 60 days is a real signal that something's working. These tools give you numbers you can actually report on, which matters when you're trying to make the case internally that AEO investment is paying off.
3. Direct Traffic Anomaly Analysis
This one requires more work but costs nothing extra.
Pull your direct traffic in Google Analytics 4. Segment it by device type, landing page, and time of day. AI-driven visits tend to show certain patterns: higher mobile rates (people using AI apps on their phones), landing on specific product pages rather than the homepage, and slightly higher average session duration because AI-referred visitors often have higher purchase intent — they've already been pre-sold by the recommendation.
You're looking for unexplained spikes in direct traffic that correlate with content you've published or AEO changes you've made. It's not clean data. But it's a signal worth watching, especially if you're seeing direct traffic grow while paid and organic search stays flat.
Set up a custom segment in GA4 for direct traffic to product pages and collection pages specifically. That's where AI referrals are most likely landing.
4. Brand Search Volume
When AI recommends your store, some percentage of users don't click the link — they open a new tab and search for you directly. That behavior shows up in branded search volume.
Use Google Search Console to track branded queries — searches that include your store or brand name. If your branded impressions and clicks are trending up while your organic keyword rankings haven't changed, that's a signal that something external is driving awareness. AI mentions are one plausible cause.
This isn't a perfect measurement. Brand awareness from social, PR, and word-of-mouth all affect branded search too. But in the absence of clean referral data, a rising branded search trend alongside increased manual AI citation visibility is a reasonable inference that your AEO efforts are doing something.
What a Weekly AEO Measurement Routine Looks Like
Keep it simple. Thirty minutes a week, consistently, beats three hours once a quarter.
Monday: Run your manual prompts across ChatGPT, Perplexity, and Google AI Mode. Log who's cited. Note any changes from last week.
Wednesday: Check your brand mention tracking tool (Otterly or Semrush). Pull citation rate for the past seven days. Flag any prompt where a competitor is consistently appearing and you're not.
Friday: Glance at your GA4 direct traffic for the week. Check Search Console branded queries. Look for anything that looks like an unexplained spike or trend.
That's it. Seriously. The stores gaining ground in AI recommendations right now aren't running elaborate measurement programs — they're just paying consistent attention when nobody else is.
According to Semrush's 2025 AI Search Report, branded queries in AI-powered search results carry a 34% higher click-through rate than unbranded category queries. That number matters because it tells you brand recognition is worth building now, before the category gets crowded with brands that figured this out earlier.
What "Winning" Looks Like in 2026
Clear benchmarks.
A store that's performing well on AI citation right now looks like this: cited in at least 40% of relevant category prompts across two or more AI platforms, branded search volume growing month-over-month without a corresponding increase in ad spend, and direct traffic to product pages showing a consistent upward trend over 90 days.
Most Shopify stores aren't close to that. We ran brand citation tests across 200 stores in Q2 2026. Only 23% showed up in AI recommendations at all — and of those, fewer than half were consistent across multiple platforms.
The stores that are winning share a few things. Complete product data (titles, descriptions, specs, materials, use cases). Legitimate customer reviews that include specific language about product features. Structured data markup that AI crawlers can parse. And actual authority content — articles, guides, comparisons — that AI models treat as source material worth citing.
None of that is magic. It's just work most stores haven't done yet.
Shopify's own documentation on product detail optimization covers the structural basics — complete product information is the foundation. But passing that information to AI models requires more: schema markup, content authority, and consistent brand signals across the web.
FAQ
Can I see AI referral traffic in Google Analytics?
Not directly. AI assistants don't pass referrer data the way Google Search does. When someone clicks a link from ChatGPT or Perplexity, the session typically shows up in GA4 as direct traffic. The workaround is using UTM parameters on any links you control (like your own AI-optimized content), brand mention tracking tools that measure citation rate directly, and tracking branded search volume as a proxy signal for AI-driven awareness.
How often should I test whether AI is recommending my store?
Weekly, minimum. AI model updates happen frequently — a model update can shift which brands get cited in a category within days. Weekly manual testing catches those changes while they're still actionable. Monthly is too slow if you're actively trying to build AI visibility. Daily is overkill unless you're managing a large brand with a dedicated team.
What's a good citation rate to aim for?
In competitive ecommerce categories, getting cited in 25-40% of relevant prompts is a strong position. In less competitive niches, 50%+ is achievable with solid AEO fundamentals in place. The number that matters most is your trend — are you being cited more often than you were 60 days ago? That's the signal. An absolute citation rate without context doesn't tell you much.
Does paid advertising in ChatGPT affect organic AI recommendations?
These are separate systems. ChatGPT Ads (launched in early 2026) place your products in paid slots within AI responses — those are tracked and billed like standard performance advertising. Organic AI citations happen independently based on the AI model's training data, real-time web search, and the quality of your product information. Paying for ChatGPT Ads doesn't improve your organic citation rate, and vice versa.
My store is getting cited but sales aren't moving. What's wrong?
Citation without conversion usually comes down to one of two problems. Either the prompts you're being cited for don't match high-purchase-intent queries — you're showing up for informational questions rather than "best [product] to buy" type searches — or your product pages aren't converting the visitors who do arrive. AI-referred traffic often has high intent, so if they're landing and leaving, the page itself is the issue: pricing, trust signals, shipping clarity, or product detail that doesn't match what the AI said about you.
Ready to find out where your store stands? Get your free AI Commerce Readiness audit at WRKNG Digital.
