How long does AEO take to show results?

Technical AEO fixes (schema markup, AI crawler access, llms.txt) can show results within weeks once AI crawlers re-index your content. Content-based AEO (building topical authority through consistent publishing) typically takes 2 to 4 months to influence citation frequency. Training-data-based improvements (affecting ChatGPT without browsing) have longer timelines tied to model update cycles. Track Perplexity and AI Overview citations first — they show real-time changes fastest.

What AEO experiments are worth running?

High-value AEO experiments include: adding FAQPage schema to existing pages and measuring citation frequency before/after, publishing direct-answer content on target queries and tracking AI recommendations, opening AI crawler access and measuring Perplexity citation rates, and adding Organization/LocalBusiness schema and checking entity recognition across AI engines. Track results in Perplexity (which shows sources) for the fastest feedback loop.

How do you measure AEO performance?

AEO measurement is less standardized than SEO. Useful methods include: manually querying ChatGPT, Perplexity, and Bing Copilot with target queries and recording whether you're cited; using Perplexity's source citations to identify which pages are being pulled; tracking brand mention volume in AI conversations via tools like BrandMentions; and monitoring Google Search Console for AI Overview appearances. Combine at least two methods for a reliable signal.

3 Months of AEO Experiments — What Actually Moved the Needle (Recap)

This is the recap post I promised in week 9 — a straight account of what I tried over the past 3 months, what moved AI citation frequency, and what was a waste of time. No false positives, no cherry-picking.

The experiments ran across bisyri.co, starting from the week the AEO Analyzer launched. The baseline: a site with a score of [BISYRI: initial score] on its own tool, appearing in AI recommendations for [BISYRI: number] of the target queries tested at baseline.

What I Tested

Over 12 weeks I made the following changes, each isolated where possible:

Week 1–2: Added Organization schema + sameAs links to homepage JSON-LD
Week 3: Added /llms.txt with brand positioning and key URLs
Week 4–8: Published 8 blog posts with FAQPage schema, all on AEO-related topics
Week 9: Updated existing service pages to direct-answer writing format (answer first, context after)
Week 10–11: Guest post published on [BISYRI: relevant industry site] with backlink
Week 12: Tool shared in [BISYRI: communities/newsletters] where it got organic mentions

Each week I ran the same set of test queries across ChatGPT (with browsing), Perplexity, and Bing Copilot, and recorded whether bisyri.co appeared in the response or citations.

What Moved the Needle

1. FAQPage Schema + Blog Content (Biggest Impact)

[BISYRI: Insert finding — e.g., "Publishing 8 blog posts with FAQPage schema over 5 weeks was the single biggest driver of Perplexity citation frequency. Before: cited in 1 of 10 target queries. After week 8: cited in 4 of 10. Perplexity started pulling FAQ answers directly from the blog posts as sources."]

This was the highest-ROI change — publishing direct-answer content with FAQPage schema, consistently, in a focused topic area. The effect built week over week rather than appearing all at once.

2. Organization Schema + sameAs Links

[BISYRI: Insert finding — e.g., "After adding Organization schema with LinkedIn and GitHub sameAs links, ChatGPT-with-browsing began correctly identifying Bisyri as an AEO consultant specifically, rather than a generic web designer. This is an entity clarity improvement — harder to measure by citation count alone but observable in how accurately the AI described the business when it did cite it."]

3. External Mentions (Tool Sharing)

[BISYRI: Insert finding — e.g., "When the AEO Analyzer was mentioned in a newsletter with ~8,000 subscribers and a few blog posts, Perplexity started citing bisyri.co for queries about 'AEO analyzer' and 'AEO check tool' within 2 weeks. External mentions are fast-acting signals for real-time AI engines."]

What Didn't Move the Needle

llms.txt

[BISYRI: Insert honest finding — e.g., "I could not attribute a measurable citation increase to adding /llms.txt. It's a low-cost, forward-looking signal, and I'll keep it in place, but as of this writing it doesn't appear to produce a detectable effect in citation frequency on any of the engines I tested."]

Meta Tag Optimization

[BISYRI: Insert finding — e.g., "Improving meta descriptions for click-through on traditional search didn't correlate with any change in AI citation frequency. AI engines aren't reading meta descriptions when deciding what to cite."]

The Honest AEO Score Progress

[BISYRI: Insert before/after scores across the 12 weeks — e.g.:

Week 1 (baseline): [score]/100
Week 4 (after schema fixes): [score]/100
Week 8 (after blog content): [score]/100
Week 12 (after external mentions): [score]/100

]

The main lesson: technical AEO fixes (schema, crawler access) are fast and high-ROI — fix them first. Content-based AEO (FAQPage schema + consistent publishing) is the compound return. External mentions are the multiplier. All three together are required for reliable AI citation.

What's Next

The next quarter's experiments will focus on:

Testing FAQPage schema on client sites and documenting the before/after across different industries
Tracking citation frequency for specific query clusters rather than just brand queries
Experimenting with video content and whether AI engines cite YouTube transcripts in recommendations

I'll publish a follow-up recap at 6 months.

Run Your Own Baseline

Before running AEO experiments, you need a baseline. The AEO Analyzer gives you a scored snapshot of where your site stands across 10 AEO factors. Run it now, note the score and which checks fail, then run it again in 8 weeks after making changes. That's the feedback loop.