Methodology

The JSON Burns methodology for AEO, GEO, and WebMCP: extractable answers, connected entity graph, honest provenance, and machine-readable discovery artifacts.

1. Extractable answer, not buried prose

Every answer page opens with the question as an H1 and the answer as the first self-contained sentence. That sentence is the acceptedAnswer in QAPage JSON-LD, character-for-character. Models cite what they can lift cleanly.

2. Connected entity graph

One Person node (Jason Burns), one Organization node (Adolicious LLC), one ProfessionalService node (JSON Burns), one Dataset node (the answer corpus). Every page references those @ids so citations resolve to the same real entities across the site.

3. Honest provenance

Author, publisher, dates, and license stamped on the page, in the JSON-LD, and in the .md twin. Content is licensed under the Citation License 1.0 — citation with attribution is welcomed; wholesale republication is not.

4. Machine-readable discovery

Every published answer ships as HTML, a Markdown twin at {url}.md, an entry in sitemap.xml, llms.txt, llms-full.txt, and questions-index.json. There is also a dedicated /ai-instructions page for crawlers and agents.

5. Additive-only builds

New capability never replaces working routes. Discovery artifacts are generated from the database at request time, so publishing or updating an answer propagates everywhere at once — no manual sitemap edits, no orphaned links.

See the corpus at /answers.