Google Search Console AI Search Reporting Is Broken: What D2C Brands Should Track Instead

Search Console's AI Search Report Is a Black Box — And Google Knows It

You've been cited in AI Overviews. Your brand name appears in a ChatGPT answer. You've watched a Perplexity thread pull three lines from your product page. But when you open Google Search Console's AI search report, all you get is a number: impressions. No clicks. No queries. No way to connect that visibility to a single session in GA4 or a rupee of revenue in Shopify.

Google admitted as much in mid-2026 — the reporting is inadequate. For Indian D2C brands pouring resources into answer engine optimization, structured data and AI-friendly content, that's a serious problem. You're optimising blind, unable to prove ROI or even confirm whether an AI citation drove a visit. This post lays out what Search Console does show, why it's not enough, and the parallel measurement stack you should build right now.

What Google Search Console Actually Shows for AI Search (and What It Doesn't)

As of September 2026, the Search Console AI search report includes:

• Impressions — the number of times your URL appeared in an AI answer.
• Average position — unclear how Google calculates this for multi-source answers; treat it as directional.

What's missing:

• Clicks — you can't tell if anyone tapped through from the AI answer to your site.
• Queries — you don't see which questions or prompts triggered the citation.
• Citation context — was your brand the primary answer, or buried in a footnote among eight others?
• Session attribution — GA4 won't tag these as AI-sourced; they roll into organic or direct depending on how the user navigates.

This isn't a temporary gap. Google has signalled that building comprehensive AI search attribution is complex — linking user prompts to citations to clicks involves privacy, consent and technical architecture that doesn't exist yet. Meanwhile, your festive budget meeting is next week, and the board wants proof that your AEO work is delivering.

Why Impression Counts Alone Are Useless for D2C Measurement

Impressions tell you you're in the game. They don't tell you if you're winning.

Consider this: your skincare brand gets 12,000 AI search impressions in August 2026. Sounds promising. But:

• 10,000 of those could be citations buried at position seven in a ten-source answer — effectively invisible.
• 1,800 might be for queries like "harmful ingredients in niacinamide serums," where you're cited as a cautionary example.
• Only 200 were for high-intent queries ("best vitamin C serum under ₹800"), and even then, no one clicked.

You've spent three months optimising schema, rewriting FAQs and publishing Reddit threads to feed AI engines. Without click or conversion data, you can't prove it worked, and you can't justify the next quarter's budget.

Worse, if you're seeing high AI impressions but declining traditional organic clicks, you may be cannibalising your own traffic — AI answered the question, so the user never visited. Impressions alone can't surface that trade-off.

How to Spot AI-Driven Traffic in GA4 (Even Without a Referrer Tag)

GA4 won't label AI search as a source — at least not reliably. But you can infer it by session behaviour.

Set up a custom exploration in GA4 with these dimensions:

Create a segment: Landing page = /products/vitamin-c-serum AND Session engagement time >60s AND Pages per session = 1. Compare weekly volume. If you see a spike that coincides with an AI impression spike in Search Console, you've likely found your AI traffic.

It's not perfect. You'll catch some non-AI sessions. But it's the closest proxy you have until Google ships proper attribution.

Build a Manual Citation Log (Yes, Really)

If the official tools won't track citation context, you track it yourself.

Set up a simple Notion or Airtable base with these fields:

1. Date — when you spotted the citation.
2. Query / Prompt — the exact question that triggered it.
3. AI engine — Google AI Overview, Perplexity, ChatGPT, Bing Copilot.
4. Position — primary answer, secondary source, footnote.
5. Snippet text — what the AI quoted from your page.
6. Screenshot URL — use a tool like Stillio or manually save.
7. Landing page — which URL was cited.
8. Outcome — did you see a traffic or conversion spike in GA4 within 48 hours?

Assign a junior team member or an intern to search 20 high-value queries weekly — "best moisturiser for oily skin India," "niacinamide vs retinol," "quick commerce COD delivery" — and log every citation. Over three months, you'll have a dataset that reveals:

• Which content formats AI prefers (FAQ schema? Reddit threads? Product descriptions?).
• Which queries drive citations but zero clicks (cannibalization risk).
• Which engines cite you most (prioritise schema efforts accordingly).

It's tedious. It's manual. But it's the only way to measure what Search Console won't show you.

What Bing Webmaster Tools Gets Right (and Why You Should Use It Anyway)

Bing's AI Performance Report — launched in mid-2026 — is everything Google's should have been.

It shows:

• Impressions and clicks — actual click-through from Copilot answers.
• Queries — the prompts that triggered your citation.
• Position distribution — how often you appear as the primary vs. secondary source.

Yes, Bing is a fraction of the volume. But the insight is transferable. If your schema drives citations in Copilot, it'll likely work in Google AI Overviews too. Use Bing's data to validate your AEO approach, then apply the same principles to Google-targeted content.

Set up Bing Webmaster Tools if you haven't. Verify your Shopify domain. Check the AI Performance report weekly. It's the best proxy measurement you'll get until Google ships a real solution, and working with our SEO team can help you extract patterns that apply across both engines and feed into a unified content strategy.

The Four KPIs D2C Brands Should Track Instead of Search Console AI Impressions

If impressions alone are a vanity metric, what should you measure?

1. Citation-to-Session Lift

Track sessions from your top 20 product and collection pages week-over-week. When AI impressions spike, does session volume follow within 48 hours? If not, your citations aren't driving traffic — they're absorbing it.

2. Branded Query Volume in Traditional Search Console

If AI answers are increasing brand awareness, you should see a lift in branded queries ("your-brand-name moisturiser," "your-brand-name review") in the traditional Performance report. Compare month-over-month. A flat or declining trend means AI isn't building brand equity — it's just answering questions.

3. Zero-Click Session Percentage

In GA4, measure sessions where the user lands on a product page, stays >60 seconds, then exits without a second pageview. If this percentage climbs alongside AI impressions, users are treating your page as a fact-check for the AI answer — they're not browsing, not adding to cart. That's not inherently bad, but it's not a sale either.

4. Conversion Rate by Landing Page (AI-Cited vs. Non-Cited)

Segment your top 50 landing pages into two groups: those cited in AI (from your manual log) and those not cited. Compare conversion rates. If AI-cited pages convert lower, it suggests AI is sending informational traffic, not buyers. Adjust your schema and content accordingly — add clear CTAs, product schema and pricing detail that AI engines can surface.

What to Do When Google Ships Better Reporting (Spoiler: It Won't Be Soon)

Google has acknowledged the gap but has not committed to a timeline. Expect iterative improvements — maybe query sampling, maybe click visibility for a subset of users — but don't expect full parity with traditional Search Console in 2026 or early 2027.

In the meantime:

• Keep optimising for structured answers. Good AEO lifts traditional snippet capture, voice search and featured-snippet visibility — all of which are measurable.
• Build your own data layer. The brands that win AI search in 2027 will be the ones who started logging citation context, testing schema variants and measuring session behaviour in 2026.
• Lobby for better tools. If you're part of a brand consortium, a D2C trade group or an agency that manages ₹10+ lakh in monthly ad spend, make noise. Google listens to aggregate demand — one brand asking for better AI reporting gets ignored; 50 brands asking together get a product roadmap mention.

Sources: Search Engine Journal reports citing Google statements and practitioner observations; Bing Webmaster Tools documentation for AI Performance Report features.