NotionCue
AI Visibility Platform
All systems live
Sign in →
AEO Guidellms.txt GeneratorRobots.txtBLUF TemplatesBlogChangelogAbout
← Blog
LLM UpdatesApr 30, 2026·6 min read

GPT-4o's new citation behaviour: what changed and what it means for your AEO score

OpenAI's April 2026 update changed how GPT-4o selects sources for commercial intent queries. Here's exactly what shifted.

AM
Arun Mehta
Head of Research · NotionCue
🎯

What OpenAI changed

In April 2026, OpenAI rolled out a significant update to GPT-4o's retrieval and citation layer, affecting how the model selects sources for commercial and transactional intent queries.

What we observed

Across 3,200 tracked commercial intent queries, we measured the following shifts between March and April 2026:

  • Pages with FAQPage schema saw a +31% increase in citation frequency.
  • Pages with thin content (under 400 words) saw a -44% decrease.
  • Pages with clear BLUF structure in the first paragraph saw a +27% increase.
  • Pages blocked by robots.txt for GPTBot saw no change, uncited regardless of content quality.
  • Pages with multiple low-quality outbound links saw a -18% decrease.

The pattern: quality signals over quantity

The April update shifted GPT-4o's retrieval weighting toward content quality signals and away from raw domain authority. Sites with high DA but thin, unstructured content saw citation drops. Sites with lower DA but well-structured, schema-rich content saw citation gains.

If your AEO score dropped in April 2026, audit your content depth first. The most common cause of citation loss in this update was pages under 400 words with no structured data and no clear BLUF structure.

What to do now

  • Run an NotionCue scan and check your GPT-4o engine score specifically.
  • Identify pages that lost citations, check word count, schema status, and BLUF score.
  • Prioritise adding FAQPage schema to any page under 600 words you want cited.
  • Rewrite introductory paragraphs on key pages to lead with the answer, not the context.
Share this post
Check your AEO score
Scan your domain free — get your AI visibility score across 5 LLMs in 30 seconds.
Scan my site →
AM
Arun Mehta
Head of Research · NotionCue

Senior SEO and AEO specialist with 12+ years across e-commerce, global education, and healthcare. Building Notion Cue to track brand citations across ChatGPT, Perplexity, Gemini, and AI Overviews.

View all →
⚖️
AEO Strategy8 min read

NotionCue vs. Building AI Citation Tracking In-House: The Real Cost Comparison

This is the first in a new comparison series. Rather than a head-to-head against a named competitor, this one addresses the comparison most teams actually face first: should we buy a tool, or is this simple enough to build ourselves with a spreadsheet and some API calls.

SS
Jul 28, 2026
📖
Technical11 min read

The AEO Glossary: 25 Terms Every AI Search Practitioner Should Know

This is the first entry in a new reference series. Rather than another explainer, this is a working glossary — short, direct definitions for the terms that come up constantly across AI search, AEO, and GEO discussions, each linked to a deeper guide where one exists on this blog.

SS
Jul 27, 2026
🚫
AEO Strategy6 min read

No, ChatGPT Does Not Have a Secret "AI SEO" Ranking Algorithm

This is the first in a series of short posts correcting specific, common misconceptions circulating about AI search. Starting with the most persistent one: the idea that there is a discoverable, gameable ranking algorithm behind ChatGPT citations the way there is behind Google's organic search. There isn't, and treating it like there is leads to bad decisions.

SS
Jul 26, 2026
📆
AEO Strategy9 min read

A 90-Day AEO Walkthrough: What Actually Changes Week by Week

This is an illustrative, composite walkthrough built from the mechanics covered throughout this blog, not a single named client engagement. It maps the realistic sequence of changes a mid-size B2B site can expect over a 90-day AEO effort, and just as importantly, what does not move yet at each stage.

SS
Jul 25, 2026
Get AEO updates weekly.

Citation shifts, algorithm changes, and what's actually working.