AigeoRadar

AigeoRadar · AIPM Benchmark Lab

HTML vs Schema vs AIPM

The Best GEO WordPress Plugin in 2026 · blog

https://aigeoradar.com/blog/the-best-geo-wordpress-plugin-in-2026-aigeoradar-connector-explained-simply

10 questions · same model gpt-4o-mini

Findings

  • — HTML only: matched 1/10 (15.0%). Unmatched: Q2, Q3, Q4, Q5, Q6, Q7, Q8, Q9 and Q10.
  • — Schema (JSON-LD): matched 3/10 (29.9%). Unmatched: Q2, Q3, Q4, Q5, Q6, Q8 and Q10.
  • — AIPM (.ai.json): matched 8/10 (70.5%). Unmatched: Q3 and Q4.
  • — Token use: AIPM used 32,168 tokens vs HTML 21,222 (51.6% more).
  • — Median latency: AIPM 907 ms vs HTML 694 ms (30.7% slower).

Matched answers

HTML 1/10 · Schema 3/10 · AIPM 8/10

Accuracy scores

15.0% / 29.9% / 70.5%

HTML / Schema / AIPM

Tokens vs HTML

+51.6%

AIPM more than HTML

Median latency vs HTML

+30.7%

AIPM slower · p95 −43.7%

Unmatched questions by context

HTML only

Did not match Q2, Q3, Q4, Q5, Q6, Q7, Q8, Q9 and Q10.

Schema (JSON-LD)

Did not match Q2, Q3, Q4, Q5, Q6, Q8 and Q10.

AIPM (.ai.json)

Did not match Q3 and Q4.

HTML only

Raw page HTML with AIPM links and JSON-LD removed. No live URL for fetching sidecars.

Unmatched: Q2, Q3, Q4, Q5, Q6, Q7, Q8, Q9 and Q10.

Accuracy 15.0%

1/10 matched

Median latency 694 ms

p95 2,021 ms

Total tokens 21,222

in 21,072 · out 150 · ~$0.0033

Schema (JSON-LD)

Structured data (JSON-LD) plus minimal page chrome. AIPM sidecars stripped.

Unmatched: Q2, Q3, Q4, Q5, Q6, Q8 and Q10.

Accuracy 29.9%

3/10 matched

Median latency 643 ms

p95 1,353 ms

Total tokens 29,318

in 29,232 · out 86 · ~$0.0044

AIPM (.ai.json)

AI Page Manifest only — purpose, facts, entities, and machine-readable fields.

Unmatched: Q3 and Q4.

Accuracy 70.5%

8/10 matched

Median latency 907 ms

p95 1,138 ms

Total tokens 32,168

in 31,922 · out 246 · ~$0.0049

Metric HTML Schema AIPM
Accuracy 15.0% 29.9% 70.5%
Median latency 694 ms 643 ms 907 ms
P95 latency 2,021 ms 1,353 ms 1,138 ms
Total tokens 21,222 29,318 32,168
Est. cost (USD) 0.00325 0.00444 0.00494
Context size 8,842 chars 15,457 chars 12,386 chars

Question-by-question answers

Identical questions for each context. Compare responses against the gold answer.

● matched gold · ● unmatched

Q1

What is the page title?

Gold: Best GEO WordPress Plugin 2026: AigeoRadar Connector

HTML

Match 89

The page title is "Best GEO WordPress Plugin 2026: AigeoRadar Connector — AigeoRadar."

Schema

Match 89

The page title is "Best GEO WordPress Plugin 2026: AigeoRadar Connector — AigeoRadar."

AIPM

Match 92

The page title is "Best GEO WordPress Plugin 2026: AigeoRadar Connector".

Q2

What does the short summary say?

Gold: Plain-English guide to AigeoRadar Connector — the WordPress GEO plugin that deploys llms.txt, entity.json, AIPM .ai.json sidecars, and AI crawler…

HTML

Miss 30

The short summary explains that AigeoRadar Connector is a WordPress GEO plugin designed to help AI systems find and trust business facts by deploying structured files for better machine readability.

Schema

Miss 63

The short summary describes AigeoRadar Connector as a WordPress GEO plugin that deploys llms.txt, entity.json, and AI crawler checks, with specifications available on GitHub.

AIPM

Match 95

Plain-English guide to AigeoRadar Connector — the WordPress GEO plugin that deploys llms.txt, entity.json, AIPM .ai.json sidecars, and AI crawler checks. Open specs on GitHub.

Q3

According to the page facts, what is key fact #6?

Gold: Setup guide: https://aigeoradar.com/help/wordpress-plugin-setup — ZIP install via Plugins → Upload, connect with a one-time token.

HTML

Miss 0

UNKNOWN.

Schema

Miss 0

UNKNOWN.

AIPM

Miss 6

Key fact #6 is: "GEO complements classic SEO plugins; Connector targets AI discovery, not only Google blue links."

Q4

According to the page facts, what is key fact #5?

Gold: ACR crawler registry (open data): https://github.com/aagedik/aigeoradar-acr

HTML

Miss 0

UNKNOWN.

Schema

Miss 0

UNKNOWN.

AIPM

Miss 19

Key fact #5 states: "Setup guide: https://aigeoradar.com/help/wordpress-plugin-setup — ZIP install via Plugins → Upload, connect with a one-time token."

Q5

List the main keywords for this page.

Gold: AI visibility, GEO, generative engine optimization, ChatGPT, llms.txt

HTML

Miss 31

AigeoRadar, WordPress plugin, GEO, AI, llms.txt, entity.json, AI crawler, discovery package, AIPM, SEO, machine-readable signals.

Schema

Miss 0

UNKNOWN

AIPM

Match 86

The main keywords for this page are "AI visibility," "GEO," "generative engine optimization," "ChatGPT," and "llms.txt."

Q6

Who is the intended audience?

Gold: B2B, agency, SMB

HTML

Miss 0

The intended audience is WordPress site owners, agencies managing client GEO, and founders wanting facts stated in files AI can quote.

Schema

Miss 0

UNKNOWN.

AIPM

Match 67

The intended audience includes B2B, agency, and SMB.

Q7

What page type is this?

Gold: blog

HTML

Miss 0

This is a product page for the AigeoRadar Connector, a WordPress GEO plugin.

Schema

Match 74

The page type is a BlogPosting.

AIPM

Match 76

The page type is blog.

Q8

According to the page facts, what is key fact #2?

Gold: It also injects homepage AI discovery head links and serves per-page AIPM .ai.json sidecars for posts and pages.

HTML

Miss 0

UNKNOWN.

Schema

Miss 0

UNKNOWN

AIPM

Match 97

Key fact #2 is: "It also injects homepage AI discovery head links and serves per-page AIPM .ai.json sidecars for posts and pages."

Q9

What language is this page in?

Gold: en

HTML

Miss 0

UNKNOWN.

Schema

Match 73

The page is in English.

AIPM

Match 73

The page is in English.

Q10

According to the page facts, what is key fact #4?

Gold: AIPM open specification and schema: https://github.com/aagedik/aigeoradar-aipm

HTML

Miss 0

UNKNOWN.

Schema

Miss 0

UNKNOWN.

AIPM

Match 95

Key fact #4 is "AIPM open specification and schema: https://github.com/aagedik/aigeoradar-aipm".

Methodology

  • Questions are derived from the page AIPM fields (purpose, topic, audience, keyFacts, and related).
  • Each question is asked once per context: HTML-only, Schema (JSON-LD), and AIPM (.ai.json).
  • HTML and Schema contexts strip live `.ai.json` hints and anonymize source URLs.
  • Scores compare model answers to gold values from AIPM. Resource metrics (tokens, latency) are reported separately.
  • Engine aipm_benchmark_v1 · completed Mon, Jul 27, 2026 11:58 PM

Share

Full run data: matched answers, tokens, latency, and per-question responses.

Post on X LinkedIn WhatsApp
About the lab → · Free AI visibility scan →