Back to GrowthDex
Growth idea action plan

Glasp on-domain control before AEO multiple claim

Measure answer-engine work against an untreated on-domain control group before claiming an AEO lift from the platform tailwind.

rare tacticmedium budget

Why this can grow a startup

AEO numbers are getting noisy. If the whole platform is receiving more ChatGPT referrals, a treated page cluster can look brilliant even when the intervention only explains part of the lift. The Glasp arXiv study is useful because it separates treated YouTube Q&A pages from untreated pages on the same high-traffic domain. That is a more honest way to read the work. For founders, the tactic is to pick a page subset, make the answer-engine changes there, leave a comparable subset alone, and compare the treated/control ratio. Ian's operator lens: in fast consumer growth, platform tailwinds are real, but confusing tailwind for tactic makes the next budget decision worse.

Company example

A 2026 arXiv field study on Glasp analyzed hundreds of thousands of YouTube Q&A pages, comparing treated AEO pages with untreated pages on the same domain to separate intervention lift from ChatGPT referral tailwind.

Source and metric

Source: arXiv: Glasp AEO natural experiment · Browse arXiv: Glasp AEO natural experiment tactics

The study reports total ChatGPT referrals grew 5.7x, untreated pages grew 3.5x, and the treated/control ratio showed a 1.82x level increase with Google indexation preserved.

AI SearchSEOAnalyticsAEO measurementcontrol groupChatGPT referralstreated pagesplatform tailwind
GrowthDex operator note

When to use it

Use this when AI Search, SEO, Analytics is relevant to AEO measurement, control group, ChatGPT referrals and you can run a bounded test with a medium budget.

When not to use it

Do not use it as a substitute for customer evidence, a clear owner, or a measurable stop condition. Local platform rules and market behavior still need checking.

Founder checklist

  1. Read arXiv: Glasp AEO natural experiment and identify what is directly supported.
  2. Choose one channel context: AI Search, SEO, Analytics.
  3. Define the test around The study reports total ChatGPT referrals grew 5.7x, untreated pages grew 3.5x, and the treated/control ratio showed a 1.82x level increase with Google indexation preserved..
  4. Set an owner, evidence window, and stop condition before launch.

Explore the context

Advisory bridge

Apply this with an operator

Turn isolated search tactics into a crawlable visibility system tied to demand and proof.

Work with Ian