Case Study · Service Business / Multi-Channel NDA

First AI-Search Citations on a 944K-User US Service Brand

GA4 now reports 583 users from "AI Platforms" — direct ChatGPT, Perplexity, and Gemini referrals. Combined with 21K Organic Video and 8.6K Organic Social, organic acquisition is 93% of new-user growth on a 106K-user property tracked in parallel.

Client: US Service Brand (Multi-Channel)
Timeframe: Q1 2026
Starting baseline: 0 AI Platform users (channel did not appear)

583 AI Platform users in GA4 (a new acquisition channel that didn't exist 6 months ago)

The numbers

Every metric, side by side

AI Platform users

583

new channel

Total users

944K

site-wide

Email channel users

195

newsletter ramp

Engaged sessions / user (parallel content site)

avg 0.96

+58% vs cat. avg

Channel mix

Where users came from

First-touch acquisition by primary channel — organic dominates the funnel.

Organic Search 65.38% 69,327
Organic Video 19.84% 21,037
Organic Social 8.14% 8,627
Paid Search 2.88% 3,053
Cross-network 2.11% 2,242
Direct 1.35% 1,435

Methodology

What I did, step by step

01

Entity-first content

Audited content for entity coverage (people, places, products, concepts), restructured top 30 pages around answer-first paragraphs LLMs can lift cleanly.
02

AEO/GEO schema

Layered FAQPage, HowTo, Article, and Speakable schema. Added clear question-answer pairs in headings to maximize LLM extraction surface area.
03

Citation surface

Distributed source-of-truth content to 6 high-authority surfaces (Wikipedia-adjacent, industry directories, niche publications) so LLMs encounter the entity in trusted contexts.
04

Brand SERP cleanup

Hardened the brand SERP (Knowledge Panel, sameAs graph, Wikidata signals) so when LLMs disambiguate the entity, they reach a clean canonical reference.
05

Multi-channel reinforcement

Coordinated organic search, organic video (YouTube SEO), and organic social so each channel reinforces the entity signal LLMs read.

Evidence

Real dashboards, not stock charts

Direct screenshots from Google Search Console and GA4 — exactly what I see when I log in to the client's account.

Source: GA4 — Q1 2026.

Supporting view — Q1 2026.

Client identifying details obscured per NDA. Full unredacted dashboards available on a verification call.

The challenge: GA4 now reports 583 users from "AI Platforms" — direct ChatGPT, Perplexity, and Gemini referrals. Combined with 21K Organic Video and 8.6K Organic Social, organic acquisition is 93% of new-user growth on a 106K-user property tracked in parallel.

What I did

Entity-first content

Audited content for entity coverage (people, places, products, concepts), restructured top 30 pages around answer-first paragraphs LLMs can lift cleanly.

AEO/GEO schema

Layered FAQPage, HowTo, Article, and Speakable schema. Added clear question-answer pairs in headings to maximize LLM extraction surface area.

Citation surface

Distributed source-of-truth content to 6 high-authority surfaces (Wikipedia-adjacent, industry directories, niche publications) so LLMs encounter the entity in trusted contexts.

Brand SERP cleanup

Hardened the brand SERP (Knowledge Panel, sameAs graph, Wikidata signals) so when LLMs disambiguate the entity, they reach a clean canonical reference.

Multi-channel reinforcement

Coordinated organic search, organic video (YouTube SEO), and organic social so each channel reinforces the entity signal LLMs read.

The outcome

GA4 now reports 583 users from "AI Platforms" — direct ChatGPT, Perplexity, and Gemini referrals. Combined with 21K Organic Video and 8.6K Organic Social, organic acquisition is 93% of new-user growth on a 106K-user property tracked in parallel.

AI Platform users: 0 → 583 (new channel)
Total users: 944K (site-wide)
Email channel users: 195 (newsletter ramp)
Engaged sessions / user (parallel content site): avg → 0.96 (+58% vs cat. avg)

Client name available on request — currently under NDA.