- Confirm the official root URL is https://erniewood.neocities.org/.
- Confirm the canonical URL points to the same root.
- Confirm the JSON-LD WebSite URL matches the root.
- Search the source code for accidental references to neocities.org without the INKBOT username.
- Confirm all INKBOT internal links use the actual INKBOT domain.
- Confirm INKBOT Lite is reachable from the root page.
- Confirm Intelligence Architecture is reachable from the root page.
- Open Google Search Console.
- Confirm ownership of the correct INKBOT property.
- Verify that the property corresponds to erniewood.neocities.org.
- Use URL Inspection on the root index page.
- Confirm Google can access the URL.
- Confirm the inspected URL is eligible for indexing.
- Request indexing after the final index page is deployed.
- Inspect the INKBOT Lite URL.
- Inspect the Intelligence Architecture URL.
- Record indexing status for all three.
- Open Performance → Search Results.
- Enable Clicks, Impressions, CTR, and Average Position.
- Review Queries and Pages separately.
- Review Countries and Devices.
- Record the initial date range and export the baseline data.
- Confirm one primary H1 on the index page.
- Confirm meaningful H2 hierarchy.
- Confirm title tag accurately describes INKBOT.
- Confirm meta description accurately describes the experience.
- Confirm canonical URL.
- Confirm mobile viewport.
- Confirm important content exists in crawlable HTML.
- Confirm images have useful alt text.
- Confirm internal links have descriptive anchor text.
- Check for broken links.
- Check that Lite and Intelligence pages load independently.
- Check page source for accidental extension/browser artifacts.
- Record current indexed URLs.
- Record current impressions.
- Record current clicks.
- Record current CTR.
- Record current average position.
- Export available search queries.
- Export page performance.
- Record current external links that are known.
- Save the baseline date.
- Do not interpret the baseline as success or failure.
- Create a master query spreadsheet.
- Create an Action category.
- Create an Identity / Personalization category.
- Create a Conceptual Friction category.
- Create a Free / Cost category.
- Create a Rough Text category.
- Create an Idea / Dream category.
- Create a Voice / Dictation category.
- Create a Design / Engineering category.
- Record every phrase as an actual human-language hypothesis.
| Intent | Example | Primary Destination |
|---|---|---|
| Action | how can ai create an image for me | INKBOT Lite |
| Rough Idea | what ai can generate a picture from rough text | INKBOT Lite |
| Conceptual Friction | how do I use AI to create an image from an idea | INKBOT Lite |
| Technical | human intent generative AI architecture | Intelligence Architecture |
- Group “how can AI...” questions.
- Group “how do I...” questions.
- Group “can AI...” questions.
- Group “I have an idea...” questions.
- Group “from rough text...” questions.
- Group “from a dream...” questions.
- Group “create an image of me...” questions.
- Group design/invention visualization questions.
- Assign each family a destination page.
- Do not create a page simply because two phrases are grammatically different.
- Make INKBOT Lite the primary CTA on Dreamer pages.
- Use “Try INKBOT Lite” language prominently.
- Explain that no prompt-engineering expertise is required.
- Emphasize rough words, fragments, voice, and incomplete ideas.
- Make the first picture the immediate objective.
- Keep technical architecture secondary on these pages.
- Measure how many search visitors reach Lite.
- Create technical entry points around human–AI interaction.
- Create technical entry points around AI instruction architecture.
- Create technical entry points around visual reasoning.
- Create technical entry points around multimodal workflows.
- Link these directly to Intelligence Architecture.
- Provide the Beach Shell experiment as concrete evidence.
- Provide source/documentation where appropriate.
- Provide a Lite doorway for technical visitors who want to test the concept.
- Create the first 10 highest-priority question pages.
- Answer the question directly.
- Explain why the problem is difficult.
- Explain how INKBOT approaches the problem.
- Give the visitor a direct Lite CTA where appropriate.
- Link related questions together.
- Link technical questions toward Intelligence Architecture.
- Record each page in the SEO ledger.
- Identify pages that genuinely contain FAQ content.
- Put the actual questions and answers visibly on the page.
- Add appropriate structured data where supported.
- Validate structured data.
- Do not hide keyword text solely for search engines.
- Do not create fake FAQ content simply to obtain search features.
- Monitor Search Console for structured-data issues where available.
- Select the first 10–20 conversational search families.
- Create useful content around them.
- Publish the pages.
- Request indexing for priority pages.
- Share them through relevant communities.
- Use social/community posts where appropriate.
- Submit relevant articles to high-volume publishing platforms.
- Track traffic into Lite separately from traffic into Intelligence Architecture.
- Do not wait for perfect rankings before testing distribution.
- Publish “How can AI create an image for me?” style content.
- Publish “How do I get AI to create a picture from an idea?” content.
- Publish “Can AI turn rough text into a picture?” content.
- Publish “How do I explain my picture idea to AI?” content.
- Publish invention/design visualization examples.
- Link each page directly to Lite.
- Record which phrases generate impressions and clicks.
- Publish “The Layer Between Human Intent and Generative AI.”
- Publish “Why AI Image Generation Has an Instruction Problem.”
- Publish “Designing a Human–AI Visual Recognition Loop.”
- Publish architecture diagrams.
- Publish actual experiments.
- Link technical articles to Intelligence Architecture.
- Link technical pages back to Lite as the working prototype.
- Prepare the Beach Shell storyboard.
- Explain the original idea in ordinary language.
- Show the survey wheel/camera arrangement.
- Show image timing and overlap.
- Show GPS association.
- Show stitched beach-walk concept.
- Show remote shell selection.
- Show technician retrieval.
- Link the case study to Lite.
- Link the case study to Intelligence Architecture.
- Prepare a HackerNoon article.
- Prepare a DEV Community article.
- Prepare a DZone technical article.
- Prepare an Instructables experiment.
- Adapt each article to its actual audience.
- Link to the appropriate INKBOT destination.
- Track publication date.
- Track referring URL.
- Track referral traffic.
- Track whether the referring page becomes indexed.
- Measure search impressions.
- Measure search clicks.
- Measure Lite page visits.
- Measure Build button usage where technically measurable.
- Measure outbound AI handoff behavior where measurable.
- Record user questions and complaints.
- Record where users abandon the experience.
- Record what users repeatedly ask for.
- Record unexpected use cases.
- Track technical search queries.
- Track Intelligence Architecture page visits.
- Track external technical article referrals.
- Track downloads/source access where measurable.
- Record technical questions.
- Record objections.
- Record requests for demonstrations.
- Record requests for architecture explanations.
- Record which technical concepts attract the strongest response.
- Rank queries by impressions.
- Rank queries by clicks.
- Rank queries by CTR.
- Rank queries by Lite visits.
- Rank queries by engagement.
- Identify low-volume/high-engagement queries.
- Identify high-volume/low-engagement queries.
- Identify unexpected query families.
- Create a “high-value intent” list.
Intent Density = relevant visitors / total search exposure
This is an internal INKBOT discovery metric—not a Google ranking metric.
- Expand winning query families.
- Create better answers for winning questions.
- Improve Lite landing paths.
- Create additional examples.
- Create additional visual demonstrations.
- Publish more external content around winning topics.
- Build additional relevant relationships.
- Retire weak pages that provide little value.
- Combine duplicate/overlapping topics where appropriate.
- Re-test after improvements.
REAL HUMAN QUESTION
↓
QUERY CAPTURE
↓
INTENT CLASSIFICATION
↓
USEFUL ANSWER
↓
┌─────────────────────────────┐
│ MASS / CREATOR INTENT │
│ ↓ │
│ INKBOT LITE │
└─────────────────────────────┘
OR
┌─────────────────────────────┐
│ TECHNICAL / AI INTENT │
│ ↓ │
│ INTELLIGENCE ARCHITECTURE │
└─────────────────────────────┘
↓
REAL USER BEHAVIOR
↓
MEASURE
↓
DISCOVER VALUE
↓
BUILD BETTER CONTENT
↓
BUILD BETTER INKBOT
↓
NEW SEARCH QUESTIONS
↓
LOOP
- Review Search Console weekly.
- Export newly discovered queries.
- Add promising conversational phrases to the query database.
- Classify new intent families.
- Create useful pages for high-potential questions.
- Route mass-market questions to Lite.
- Route technical questions to Intelligence Architecture.
- Publish external articles around validated opportunities.
- Measure referral traffic.
- Measure product engagement.
- Record unexpected use cases.
- Feed those discoveries back into INKBOT development.
- Repeat indefinitely.
Become exceptionally good at owning the conversational search territory around:
“I have an idea, but I don't know how to get AI to understand what I mean.”
Then use the traffic generated by that territory to discover what INKBOT should become next.
Strategic Guardrails
We should test lots of relevant conversational searches, publish useful material, distribute it, and put INKBOT in front of real people.
We should not manufacture hundreds of nearly identical pages, hide keywords from users, create doorway pages, or generate low-value pages whose primary purpose is ranking.
The objective is to become extremely good at answering real human questions—not to fool search engines.
A search phrase that attracts people and produces INKBOT engagement is more valuable than a theoretical keyword with impressive volume but no meaningful behavior.
We create the conditions for the market to reveal itself.
Strategy Source & Working Thread
This strategy was developed through the INKBOT strategy discussion in ChatGPT.
Use the link below as the working ChatGPT reference. If you create a ChatGPT Share link for this exact conversation, replace the URL below with that specific share URL.
Open ChatGPT — replace with this conversation's Share URL
CHATGPT THREAD URL: _________________________________________________
Master Completion Record
Strategy Start Date: ______________________________
Strategy Owner: Thomas D. Coates / INKBOT
Official Hub: https://erniewood.neocities.org/
Primary Mass-Traffic Destination: INKBOT Lite
Primary Technical Destination: INKBOT Intelligence Architecture
Primary Search Thesis: conversational long-tail queries surrounding the gap between human ideas and AI understanding
Current Strategy Completion: ______ / 20 steps
Last Strategy Review: ______________________________
Next Strategy Review: ______________________________
This document is designed as an executable working checklist. Check each step only when the underlying work has actually been completed.