Updated November 19, 2026 with new 2026 click-through, buyer and referrer data, and the companion deep-dive on edge logs.
Most agencies selling AI optimization treat a data engineering problem like a creative writing exercise, and then tell you the results can't be measured. Both halves are wrong. The short version: AI answers are too random to win with hand-typed prompts, so you need deterministic pipelines to earn citations, and you need more than Google Analytics to prove they're working.
Prefer the numbers at a glance? View the AI Search Attribution Gap infographic. For the full firewall breakdown and setup guide, read Part 2: Beyond Analytics.
The Agency Trap: Prompting vs. Engineering for GEO
Most marketing agencies and consultants selling AI optimization (AIO) and generative engine optimization (GEO), the work of getting your brand named and cited in AI answers, are failing their clients, because they treat a data engineering problem like a creative writing exercise.
Right now, the standard agency playbook relies on prompters: strategists typing queries into ChatGPT or Perplexity by hand, guessing why a competitor showed up, and rewriting a few paragraphs of copy. There is no magic bullet or secret prompt that forces an LLM or Google's AI Overviews to recommend your brand. LLMs are probabilistic, so the same prompt can return different citations from run to run, and manual spot-checking produces false positives and wasted retainers.
How random is it? In SparkToro and Gumshoe.ai's test of 2,961 prompts across ChatGPT, Claude and Google's AI results, the tools returned a different brand list more than 99% of the time, and the same list in the same order less than 0.1% of the time (SparkToro research via Search Engine Land). Agencies that poll raw model APIs instead of the live interfaces buyers use make it worse: Surfer found brand overlap between API answers and live answers of just 15.5% to 32%, and page-level citation overlap of 4.8% to 19.7% (Surfer).
At Pressfit, we reject the magic-bullet approach. Showing up in AI search means connecting the dots across large, structured datasets:
- Your website's architecture, technical health and existing content.
- Your competitors' domains, authority signals and content gaps.
- The search terms and conversational intents that drive buyers to your site.
Real GEO takes time, and most agencies fail to deliver it, because automating what a multi-person agency team would spend weeks researching by hand takes deterministic compute. Instead of guessing with prompts, you build repeatable pipelines that ingest site structures, map entity relationships, cross-reference search intent and output a prioritized playbook. Most agencies don't have data engineers. They have prompters. Without deterministic engineering under the hood, they can't execute at the scale or consistency AI engines require.
You can't prompt your way into an LLM's retrieval. In March 2026, Ahrefs found that only 37.9% of URLs cited in Google AI Overviews also ranked in the top 10 results, and 31% didn't rank in the top 100 at all, down from about 76% from the top 10 in mid-2025 (Ahrefs). AI engines pick sources for entity clarity, structured topical coverage and how well a page answers the follow-up questions behind a query, not for keyword stuffing.
| Workflow | Traditional agency (prompters) | Pressfit Playbook (deterministic compute) |
|---|---|---|
| Competitor and site audit | Clicking through a few competitor blogs by hand and guessing why they rank. | Crawling and parsing the page structure and schema of your site and your competitors' sites to map gaps at scale. |
| Intent and query mapping | Spot-checking a handful of broad prompts in a single chat window. | Working from real search demand and mapping query fan-out, the sub-questions an AI engine searches behind one prompt. About 95% of those sub-questions show zero monthly search volume in keyword tools (Seer Interactive). |
| Entity and authority signals | Rewriting copy with "AI-friendly" buzzwords and hoping a model picks it up. | Extracting named entities, co-occurrence patterns and third-party brand mentions. Across 75,000 brands, Ahrefs found branded web mentions correlate with AI Overview visibility at 0.664, against 0.218 for backlinks (Ahrefs). |
| Execution and output | Subjective advice ("write more authoritative content") that varies by strategist. | A deterministic, prioritized playbook with repeatable scoring rules, so the deliverable is consistent and auditable every time. |
| Why it scales (or fails) | Breaks under client load; human spot-checks can't keep pace with model updates or quarterly citation decay. | Takes engineering time up front, then delivers predictable, compounding authority across Google AI Overviews and LLMs. |
Why Everyone Claims AI Search Can't Be Tracked
Once you do the engineering work to show up in AI Overviews and LLMs, you hit the second wall: attribution.
A vocal part of the SEO industry says AIO and GEO can't be measured, and uses that as a reason to avoid accountability. Traditional rank tracking is broken, but throwing up our hands isn't an option. Buyers search differently than they used to, and measurement has to adapt.
The 2026 data shows both sides. Trackable AI referral volume looks small because of attribution blind spots, yet the traffic that does come through is changing pipeline quality.
Zero-Click Search Is the Default
From January to April 2026, 68% of U.S. Google searches ended without a click, according to Similarweb data analyzed by SparkToro. That panel excludes the Google mobile app, so the real share is likely higher (SparkToro). First Page Sage's panel of 1,184 domains puts the U.S. zero-click rate at 74.2%, with an AI answer on 81.6% of queries (First Page Sage). When an AI summary appears, zero-click rates climb to roughly 83%.
AI Overview Citations Win the Clicks That Remain
First Page Sage found that ranking first organically is worth 84.1% less when an AI answer sits above it: click-through drops from 22.6% to 3.6%. On those same pages, citations inside the AI answer take 51.0% of all remaining clicks, and the top three cited sources take 77.9% of AI answer clicks (First Page Sage). Seer Interactive found brands cited in an AI Overview averaged a 2.07% organic click-through rate on informational queries, against 0.94% for brands that weren't cited.
The AI Referrer Blind Spot
Across 10 industries, AI referrals made up 1.08% of website traffic in Conductor's benchmark, and 87.4% of that came from ChatGPT (Conductor). Industry estimates put 60% to 70% of AI-referred sessions as arriving with no referrer at all, from native mobile apps, logged-in sessions and copied links, so they land in Direct (Seer Interactive).
Google Analytics 4 added a built-in AI Assistant channel on May 13, 2026. It helps, but it only works going forward, it covers ChatGPT, Gemini, DeepSeek, Copilot and Grok, and it leaves Perplexity in Referral. Google also counts AI Overview and AI Mode clicks inside Organic Search, not the AI Assistant channel. A lot of AI-driven discovery still gets filed as Direct or Organic.
AI Search Buyers Arrive With Outsized Intent
Even with stripped referrers, the conversions you can isolate show AI searchers are further down the funnel. G2's The Answer Economy survey of 1,076 B2B software buyers found 51% now start their software research in an AI chatbot more often than Google, up from 29%, and 69% chose a different vendor than they first had in mind because of what an AI told them (G2). Semrush modeled the average AI search visitor in marketing topics as worth 4.4 times a traditional organic visitor (Semrush), and on Ahrefs' own site, AI search was 0.5% of visitors and drove 12.1% of signups (Ahrefs). Broader cross-sector studies land lower, closer to 1.2 times organic, so treat the high multiples as a ceiling for research-heavy categories, not a promise.
What Firewall and Edge Logs Show That Analytics Can't
If your agency is still hosting webinars about AI attribution while staring at referral traffic in Google Analytics, you're getting a rearview-mirror report.
Here's our own 30 days at Pressfit.ai. Google Analytics showed about 6,500 human visits, 99.38% of them Direct, and exactly one AI assistant (Grok, 10 visits) in the referral report. Over the same window, our firewall and edge logs recorded 7,321 requests from AI and answer-engine bots, 1.13 times our human traffic, including 127 live ChatGPT-User fetches and 57 live Claude-User fetches. Those are people asking an AI a question that made it read our pages, and none of them showed up in analytics.
We walk through every crawler, how to tell live human fetches from background indexing, and the step-by-step setup in Cloudflare, AWS, Vercel, GA4 and Search Console in Part 2: Beyond Analytics.
How to Measure AIO and GEO: The Four-Layer Attribution Stack
If your agency measures SEO and GEO only by referral traffic in an analytics dashboard, it is ignoring the thousands of machine retrieval events happening on your domain every month. Real SEO and GEO measurement connects four layers of data:
- Deterministic content and site engineering. Structure entity relationships, schema and coverage of the follow-up questions behind each query, so crawlers can read your authority without friction.
- Edge and firewall telemetry. Track user-agent volume across search indexers (Googlebot, Bingbot), model crawlers (Meta-WebIndexer, ClaudeBot), neural search APIs (Exa, Parallel) and live human-triggered fetches (ChatGPT-User, Claude-User).
- Analytics and direct-lift correlation. Track the AI referrals you can see alongside spikes in unattributed Direct traffic on the exact landing pages getting live AI user fetches.
- CRM and self-reported attribution. Ask inbound prospects how they found you before they fill out a form.
At Pressfit, we don't rely on guesswork, manual prompting or broken cookie reports. We engineer the deterministic playbooks that get your brand retrieved by AI engines, and connect the network, analytics and pipeline telemetry to prove it works.
Want the roadmap that comes before the measurement? Read The Pressfit Playbook: Your Roadmap to AI Visibility, or get your Playbook.
FAQ
Can you measure AI search visibility (AIO and GEO)?+
Yes, but not from referral traffic alone. Combine your site and content engineering, edge and firewall logs, analytics with direct-lift correlation, and CRM self-reported attribution. On Pressfit.ai, Google Analytics showed one AI assistant referrer in 30 days while our edge logs recorded 7,321 AI and answer-engine requests.
Why do prompt-based AI rank trackers give inconsistent results?+
AI answers are probabilistic. In SparkToro and Gumshoe.ai's test of 2,961 prompts, the tools returned a different brand list more than 99% of the time and the same list in the same order less than 0.1% of the time. Tracking from raw APIs adds more error, because API answers overlap with live answers far less than most teams assume.
What share of Google searches are zero-click in 2026?+
SparkToro's Similarweb data put U.S. zero-click searches at 68% from January to April 2026. First Page Sage's panel puts it at 74.2%, rising to about 83% when an AI summary appears.
Does GA4 track ChatGPT and AI assistant traffic?+
Partly. Since May 13, 2026, GA4 has an AI Assistant channel covering ChatGPT, Gemini, DeepSeek, Copilot and Grok. It is not retroactive, it leaves Perplexity in Referral, and AI Overview and AI Mode clicks stay in Organic Search. Visits with no referrer still land in Direct.
Do AI Overview citations still get clicks?+
Yes. First Page Sage found citations take 51.0% of clicks on pages with an AI answer, and the top three cited sources take 77.9% of AI answer clicks, while the first organic result falls from 22.6% to 3.6% click-through.
The Research Behind This Post
- Part 2: Beyond Analytics: What Edge Logs Reveal About AI Demand
- The AI Search Attribution Gap infographic
- Stop Prompting and Start Engineering, our research on deterministic compute
- The Graph Workflow Starter Kit
- Outside sources: SparkToro zero-click study; SparkToro and Gumshoe.ai on AI recommendation consistency; Surfer, scraped answers vs. API results; Ahrefs, AI Overview citations and the top 10; Ahrefs, 75,000-brand AI Overview study; Ahrefs, AI search conversions; Seer Interactive, Gemini 3 query fan-outs; First Page Sage, 2026 CTR by position; Conductor, 2026 AEO/GEO benchmarks; G2, The Answer Economy; Semrush, AI search traffic study
Need help putting it to work? Get in touch, or email our team at redevteam@pressfit.ai.
