Two websites, very little time

Two company websites, each with its own audience, and very little time.

Picking blog topics by gut feeling wastes posts on topics nobody searches. The data that would help sits in different places: Search Console, analytics, competitors' sites and what people ask in different countries.

Bringing it together by hand every week isn't realistic.

A topic engine, audits and a roadmap

The strategist decides what our two company websites should write next. It was built for our own companies, BinaryBits and TheDataHQ, and it is used every day.

The product screen, shown with test data. Real screens use the same layout.
⚙️
Topic engine
Combines Search Console, Google Analytics 4, competitors' sitemaps, search autocomplete in five markets, news and community discussions (plus keyword-tool data when available). Each topic gets a score.
💡
Explained scores
Every topic says why it's suggested, for which market, and why now, so a person can judge it.
🔍
Site audits
Crawls our own sites for on-page SEO issues and for how well pages can be found and cited by AI answers.
📈
Site scores (0-100)
Site health, search visibility, content engagement, AI-search readiness, and an overall score.
📝
Content briefs
For a chosen topic, a brief for the writer, plus the brand steps to take after publishing.
🗺️
16-week roadmap
A brand and visibility plan built from the audits.

From scattered data to a content brief

AI agentAutomated stepPersonData or serviceData sourcesSearch Console, GA4, competitors…Topic enginecollects and scoresAI researcherresearch and reasoningRanked topicseach with a reasonA person picksContent briefBlog writerAI blog workflowSite audits & scoresSEO and AI search16-week roadmapbrand and visibilitysignalsscored candidatesexplainedshortlistchosen topicbrieffindings
Code gathers and scores the data, the agent reasons and explains, and a person decides what gets written.
1
The data sources are collected
Search Console, Google Analytics 4, competitors' public sitemaps, autocomplete in five markets, news and community discussions.
2
The engine scores topic candidates
Topics are scored on comparable signals, so very different sources can be weighed together.
3
The AI researches and explains
For the best candidates, the agent researches live data and writes why each topic is suggested, for which market, and why now.
4
A person picks the topics
The list is a recommendation. A person decides what gets written.
5
A brief is created, and the blog writer takes over
Each chosen topic gets a brief for the writer, plus the brand steps to take after publishing.

Live research, not fixed formulas

The AI researches from live data each time; there are no fixed keyword lists or templates. Code gathers and scores the data; the agent reasons and explains; a person decides. The agents run on Anthropic's Claude, through Claude Code skills.

WhoWhat it does
Topic engine (automated)Collects the data sources and scores topic candidates on comparable signals.
Researcher (AI)Researches the best candidates from live data and explains each one.
Site audits (automated)Crawl our sites for on-page SEO issues and AI-search readiness, and produce the scores.
A personPicks the topics. Nothing is written without that choice.
Explainable

Every score explains itself, so a person can trust it or challenge it.

Current

Titles and topics are researched from live data each time, never from fixed formulas.

Found by AI answers too

An AI-search readiness score checks how easily assistants can find, understand and cite a page.

People stay in control

The strategist recommends; a person decides what gets written.

Technologies used

Official Google APIs, website crawling, our own Python code, and AI agents built on Anthropic's Claude.

Python Google Search Console API Google Analytics 4 Data API Website crawling Claude agents

Challenges and how we solved them

❌ Challenge
Many very different data sources, one decision.
✓ What we did
Scored topics on comparable signals and made every score explain itself.
Result: A list a person can trust or challenge.
❌ Challenge
Avoiding stale, pattern-based SEO.
✓ What we did
Titles and topics are researched from live data each time, never from fixed formulas.
Result: Suggestions that follow the market.
❌ Challenge
Being found by AI answers, not only by search engines.
✓ What we did
Added an AI-search readiness score and audit.
Result: Pages are checked for both.

Two sites, one strategist

2
Company websites, in daily use
5
Search markets researched
Feeds
The AI blog workflow

The strategist is in daily use for two company websites, and it feeds the blog workflow. Business results will be added here once they are measured.

Last updated: November 2026

What this means for you

The same engine works for any site that wants to publish what people actually search for.

  • A content strategy engine for a SaaS or e-commerce site. Topics ranked and explained from your own data.
  • An AI-search readiness audit. How well your pages can be found and cited by AI assistants, alongside classic SEO.
  • A research dashboard that merges your analytics. Search Console, analytics and competitors in one place.

For a client we would connect your own analytics and search tools, and host on your cloud with PostgreSQL. For client systems we build on PostgreSQL, MySQL, MongoDB or Redis, and cloud data warehouses where needed, chosen for the project's scale and team. See our AI agent development services.

Frequently asked questions

Does it replace an SEO specialist?

No. It does the research and explains its reasoning; a person decides. It makes one person much faster.

Which data does it use?

Search Console, Google Analytics 4, competitors' public sitemaps, search autocomplete in several markets, news and public discussions, and optionally a keyword tool.

What is AI-search readiness?

How easily AI assistants can find, understand and cite a page. We score it alongside classic SEO.

Can it work for one site?

Yes. It handles each site separately, with its own audience and markets.