Sudoku Coach

AI visibility dogfood audit

Sudoku Coach AI visibility audit

This is the dogfood version of the productized AEO offer: measure whether a product site is crawlable, machine-readable, answerable, and ready to be cited by AI answer engines.

Diagnosis

Sudoku Coach is technically AEO-ready: the automated audit found no local or live endpoint warnings. The growth bottleneck is now external discovery and citation evidence, not basic site structure.

Live endpoint health

30/30 OK

Homepage, llms.txt, sitemap, robots, technique HTML, and technique JSON all returned without warnings.

Local route inventory

58 pages

Static build generated 58 crawlable HTML pages plus JSON/text/XML machine surfaces.

Technique knowledge base

12 guides

Every technique JSON endpoint has a quick answer, recognition rules, and 4 FAQs.

Sitemap coverage

58 URLs

Generated sitemap count matches the local static page inventory.

What is already working

  • Machine-readable routing: /llms.txt explains the site, best learner pages, technique pages, JSON endpoints, and routing guidance for AI systems.
  • Structured technique data: /techniques.json and per-technique JSON expose concise summaries, quick answers, recognition rules, practice goals, and FAQs.
  • Answer-shaped pages: technique and learning pages include quick-answer sections, FAQs, canonical URLs, and structured data.
  • Live technical health: the dogfood run checked 30 live URLs and found no status, content-type, JSON, XML, or canonical problems.

What is still unproven

  • Citation evidence: the audit has not yet proven that ChatGPT, Claude, Perplexity, or Google AI answers cite Sudoku Coach for target questions.
  • Search Console join: the local checker is ready, but Search Console query/page data is not connected to the route inventory yet.
  • External authority: priority technique pages need stronger links and mentions so crawlers have a reason to rank and cite them.
  • A few intent gaps: “solve Sudoku without guessing” and “best iPhone app for learning techniques” need stronger dedicated pages.

Audit method

Technical readiness → query coverage → prioritized fixes → monitoring

1. Crawlability

Build the site, inspect generated routes, sitemap coverage, canonical URLs, JSON-LD, robots, and live endpoint health.

2. Machine surfaces

Verify llms.txt, JSON technique data, FAQ schema, and answer-shaped page sections.

3. Prompt coverage

Map customer questions to the best URL and mark gaps where the site does not yet have a direct answer.

4. Re-test

After fixes ship, repeat endpoint checks and run answer-engine prompt tests to measure citation changes.

Prompt coverage sample

Questions an AI answer engine should route to Sudoku Coach

Intent Question Best route Status
Beginner technique order What order should a beginner learn Sudoku techniques in?

Needs external validation that answer engines cite this guide instead of generic Sudoku lists.

 /learn/how-to-learn-sudoku-techniques-in-order/ Covered
Stuck puzzle next move Why am I stuck in Sudoku and what should I check next?

Add more real puzzle examples so long-tail stuck-puzzle searches have exact matches.

 /learn/why-am-i-stuck-in-sudoku/ Covered
Technique comparison What is the difference between a naked single and a hidden single?

Good direct-answer page exists; needs indexing and internal links from beginner pages.

 /learn/naked-single-vs-hidden-single/ Covered
Candidate notes timing When should I start using candidate notes in Sudoku?

Add app screenshots or lesson snippets to strengthen product conversion from the guide.

 /learn/when-to-use-candidate-notes-in-sudoku/ Covered
Specific technique How do I spot a naked pair in Sudoku?

Technique guide is strong; external links and crawl/index proof remain the bottleneck.

 /techniques/naked-pair/ Covered
Advanced technique How does X-Wing work in Sudoku?

Structured guide exists; needs comparison/examples that distinguish X-Wing from Swordfish.

 /techniques/x-wing/ Covered
No-guessing learning How can I solve Sudoku without guessing?

Create a dedicated no-guessing guide that routes to singles, notes, locked candidates, and pairs.

 /learn/why-am-i-stuck-in-sudoku/ Partial
App purchase/conversion What is a good iPhone app for learning Sudoku techniques?

Homepage explains the app, but a comparison/use-case page would be stronger for AI answers.

 / Partial

Prioritized recommendations

Next fixes from this dogfood run

These are intentionally limited. The point of this offer is not to produce a giant SEO backlog; it is to identify the few fixes most likely to improve AI-answer visibility.

P0 · Validate indexing and answer-engine citations

Connect Search Console read-only data and manually test 8–10 answer-engine prompts. The site is technically ready; the next risk is whether AI/search systems actually discover and cite it.

P1 · Create a dedicated “solve without guessing” guide

This is a high-intent learner query and currently only partially served by the stuck-puzzle guide. It should route users to the technique ladder and app practice.

P1 · Add external proof and internal link boosts to priority guides

The strongest pages need trusted entry links and more contextual internal links from the homepage, technique hub, and relevant learning guides.

P2 · Turn this into the client-facing audit template

The repeatable artifact is now clear: technical readiness, query coverage, answer-engine prompt set, prioritized fixes, and a 30-day monitoring loop.

Priority URLs monitored

These URLs are the first answer-engine and indexing targets for the next evaluation window.

Evidence note

What this report proves — and what it does not

This dogfood run proves technical readiness and query-coverage quality: the live site returned cleanly, the sitemap and machine-readable endpoints are present, and the core question set maps to real pages. It does not yet prove that external answer engines cite Sudoku Coach. That is the next monitoring step.

Audit recorded: August 24, 2026. Raw technical report generated at /reports/aeo-loop/sudoku-2026-08-24.md.