I get asked this question every week now: “How do I know if AEO is actually working?” The honest answer is harder than it should be. AI search platforms don’t give you a dashboard. ChatGPT doesn’t send you a report showing how many times it cited your plumbing company in Nampa. Perplexity won’t tell you how often it mentioned your law firm when someone asked about Idaho estate planning. You’re flying half-blind, measuring downstream signals instead of direct visibility. But those signals exist, and I’ve been tracking them for Idaho clients since mid-2024. Here’s what actually works when traditional analytics fall short.
Why doesn’t Google Analytics show AEO performance?
GA4 wasn’t built for a world where conversational AI answers questions without sending clicks. When someone asks ChatGPT “best HVAC contractor in Boise” and gets your name, no referral hit registers unless they click through to your site. Most users don’t click. They screenshot the answer, copy your phone number, or just remember your name for later.
GA4 will show you referral traffic from chat.openai.com or perplexity.ai when clicks do happen, but that’s a trailing indicator of citation success, not a measure of how often you’re mentioned. I’ve seen Boise clients get cited dozens of times in testing before a single referral click showed up in analytics.
The other problem: AI platforms strip most referrer data. You might see a visit from “perplexity.ai” but you won’t see the original query that triggered your citation. You’re left guessing which content earned the mention. Traditional conversion tracking still works once someone lands on your site, but the top of your funnel is now partially invisible.
What is direct prompting and how do I use it?
Direct prompting means you manually ask AI platforms to answer queries related to your business, then check if your brand appears. It’s tedious. It’s not statistically valid. It works anyway.
Here’s my weekly routine for a Meridian commercial real estate client:
– Open ChatGPT, Perplexity, and Gemini in separate tabs
– Ask 8-12 variations of queries my client should rank for: “commercial property management Treasure Valley”, “best CRE broker for industrial space Idaho”, “who handles multifamily sales in Boise”
– Screenshot every result
– Track mentions in a simple spreadsheet: date, platform, query, cited yes/no, position if cited
After six weeks, patterns emerge. You see which platforms favor your content, which query types trigger citations, and whether you’re moving from position 4 to position 2 in Perplexity’s answers. One Caldwell client went from zero mentions in October 2024 to appearing in 60% of tested queries by February 2025 after we restructured their service pages with AEO patterns.
The big limitation: you’re only testing queries you think of. Real users ask thousands of variations you’ll never anticipate. Direct prompting measures a sample, not the universe.
Why is branded search volume a leading indicator?
When AI platforms cite your business, people search for your brand name later. It’s the most reliable signal I’ve found.
In Google Search Console, filter for queries containing your exact business name. Look at the trend line for impressions, not clicks. If AEO is working, you should see 15-30% lifts in branded impressions within 60-90 days of implementing citation-focused content.
I tracked this for a Boise litigation attorney. Before AEO work: 340 monthly branded searches average. After restructuring practice area pages and adding case outcome specifics: 520 monthly branded searches four months later. The attorney didn’t change his advertising. He didn’t get a big case covered in the Idaho Statesman. The increase came from AI citations introducing his name to people who’d never heard of him.
Google Search Console also shows “related queries”, phrases people search right before or after your brand name. When AEO works, you see more informational queries paired with your brand: “[Business Name] + reviews”, “[Business Name] + pricing”, “[Business Name] + hours”. That tells you people are using AI for discovery, then switching to Google for verification.
Branded search is a lagging indicator by 30-60 days, but it’s measurable and it correlates directly with revenue for every client I’ve tracked.
How do I track referral traffic from AI platforms?
GA4 will show referral traffic from AI platforms if you know where to look. Go to Reports > Acquisition > Traffic Acquisition, then add a filter for source containing “openai”, “perplexity”, “claude”, or “you.com”.
What you’ll actually see:
– chat.openai.com (ChatGPT web interface clicks)
– perplexity.ai (Perplexity answer citations)
– you.com (You.com search results)
Volumes are low. A successful month for a small Boise business might be 40-80 visits from these sources combined. But the quality is high. Conversion rates from AI referral traffic run 2-3x higher than organic search in my client data because the user already got a pre-qualified recommendation.
One thing I watch closely: new vs. returning users from AI platforms. If 90%+ are new users, your citations are working as discovery tools. If you see high returning user rates, it means people are using AI as a bookmark system for businesses they already know, less valuable.
Set up a GA4 segment for “AI Platform Referrals” so you can track this group’s behavior separately. Do they convert faster? Do they view pricing pages more often? For a Nampa SaaS company, AI referrals had a 4-day shorter sales cycle than organic search, probably because the AI pre-framed the value proposition.
What citation tracking tools actually measure?
Third-party citation tracking tools emerged in 2024-2025. Most work by running automated queries through AI platforms and parsing responses for brand mentions. The ones I’ve tested or seen demos for: SEO.ai, Profound, BrightEdge, and a couple of startups not worth naming yet.
What they do well:
– Run hundreds of queries you’d never test manually
– Track your citation rate over time (what % of relevant queries mention your brand)
– Compare your citation frequency to competitors
– Identify which content pages get cited most often
What they don’t do:
– Measure actual impression volume (the AI platforms don’t share that data)
– Show query volume for the searches they test (you don’t know if you’re ranking for high-volume or obscure queries)
– Guarantee correlation with traffic or revenue
I’ve used SEO.ai for three Idaho clients. It’s helpful for competitive benchmarking. If you’re a personal injury attorney in Idaho Falls and the tool shows you’re cited in 12% of relevant queries while your main competitor hits 34%, you know you’ve got work to do. The directionality is useful even if the absolute numbers are fuzzy.
Pricing for these tools runs $200-$800/month depending on query volume and features. Worth it if you’re investing heavily in AEO and need to justify the spend to a CFO. Probably overkill if you’re a solo operator testing the waters.
When should I expect to see AEO results?
AI platforms crawl differently than Google, but they still need time to ingest and weight your content. Here’s the realistic timeline I give clients:
Weeks 1-4: Nothing measurable. You’re publishing citation-worthy content but AI models haven’t re-indexed your site yet. Direct prompting might show sporadic mentions if you test obsessively.
Weeks 5-8: First signals appear. Branded search in GSC ticks up 10-20%. You might see 5-10 referral visits from chat.openai.com. Citation tracking tools start showing your brand in answers, usually at position 3-5.
Weeks 9-16: Compounding effects. Branded search lifts 20-40%. AI referral traffic becomes consistent. You move from occasional mentions to regular inclusion in answers. One Boise ecommerce client saw their first Perplexity citation in week 6, then saw 11 more citations in the following month as the platform’s model updated.
Month 5+: Mature performance. You should have clear baseline data for branded search, referral traffic, and citation frequency. You can A/B test content changes and see impact within 3-4 weeks instead of 3-4 months.
The big variable: how much quality content you publish. A client who adds two citation-optimized pages per month sees slower gains than one who restructures 20 existing pages in month one. I’ve seen Treasure Valley B2B companies get traction in 6 weeks with aggressive content overhauls. I’ve also seen cautious businesses take five months to see meaningful lift because they published slowly.
How do I connect AEO metrics to actual revenue?
This is the question that matters. None of this works if you can’t draw a line from citations to closed deals.
Start with source/medium tracking in GA4. When someone converts (form fill, phone call, purchase), GA4 attributes it to a source. Filter your conversion reports for AI platform referrals and calculate revenue per session. For three Idaho clients I track closely, AI referral traffic converts at $180-$340 per session vs. $60-$90 for organic search.
The harder measurement: assisted conversions. Someone gets your name from ChatGPT, searches your brand in Google a week later, clicks an ad, and converts. GA4’s default attribution gives credit to the paid ad. The Advertising > Attribution reports show assisted conversions, but AI platforms usually don’t appear because the original interaction happened outside the browser.
I use a workaround: add a “How did you hear about us?” field to lead forms with “AI search (ChatGPT, Perplexity, etc.)” as an option. About 40% of prospects who came from AI citations will self-report it. Not perfect, but better than nothing.
For phone call tracking, ask your intake team to note when someone says “I asked ChatGPT” or “Perplexity recommended you.” Train them to listen for it. A Boise HVAC company logged 23 AI-sourced calls in January 2025 after training their call center to ask. Zero of those showed up in GA4.
The cleanest measurement: track branded search volume against new customer volume month over month. When branded search goes up 30% and new customers increase 18% with no other marketing changes, AEO is contributing.