Content Gap Finding Engine | Content Gap Analysis & Research Tool for Authors, Bloggers and Creators

You probably do not need more content. You need to understand what your existing content is missing.

Most creators feel this. You write, you publish, you keep going. Then the doubt arrives. Are these the right topics? What are readers asking that I never answered? What are competitors covering that I am not? And which gaps are real, and which are just ideas?

Answering those questions is research. Real research, with sources. Not a chat prompt that gives different answers every week.

That is exactly what this product is.

What is the Content Gap Finding Engine?

It is a local research program. You give it one content file. A book extract, an article, a transcript, a newsletter, a report, a draft, or notes in plain text or markdown.

It reads your material first. Then it investigates the landscape around that material. What do real searchers ask? What has the competition covered? Which of your topics stay unanswered? Which gaps deserve your time?

The result is a structured research brief:

• Scored keywords, each flagged as target, secondary, or skip
• Validated content gaps, with the evidence behind each one
• Ranked opportunities, scored from 0 to 100 with a separate confidence score
• A full evidence ledger, every finding tied to a source
• A research summary and verification notes, written for humans

Every run gets its own run ID. Run the same source next month, and you can compare the two runs, term by term.

It is not a website tool. It is not a dashboard. It is a research engine that runs on your computer, from a terminal, with a fixed and repeatable pipeline.

The problem it solves

The old way of content research is broken in three places.

First, most tools start from a blank box. Type a topic, get a list of terms. The list looks useful. It never tells you what your own work is missing.

Second, the numbers cannot be checked. Most tools return volumes and difficulty scores with no way to see where they came from.

Third, the results are not repeatable. Ask a chatbot for content ideas today, ask again next week, and you get a different set. Nothing ties the ideas to real search behavior.

This engine was built against those three failures. It starts from your content. It attaches a source to every finding. And its pipeline is fixed, so the same source produces the same run, with a run ID you can compare.

How the engine works

Every run follows the same sequence.

CONTENT → CONTENT INTELLIGENCE → SEARCH + AUDIENCE + COMPETITOR RESEARCH → COVERAGE ANALYSIS → GAP DETECTION → EVIDENCE + VALIDATION → OPPORTUNITY SCORING → CONTENT OPPORTUNITIES

Content intelligence reads your source before anything external. Every later step stays anchored to your actual material.

Search research pulls public signals from search engines, autocomplete, and related queries. For book-shaped sources, it adds Amazon-derived signals, always labeled as engine-derived proxies, never official sales data. It checks social and question sources such as YouTube, Reddit, and Quora where they are reachable.

Coverage analysis compares the landscape with your content. Gap detection finds what your content does not cover and why the evidence says it matters.

Evidence and validation require more than one signal before a gap qualifies. A topic does not make the list just because one competitor mentioned it once.

Opportunity scoring gives every surviving idea a score from 0 to 100 and a separate evidence confidence score.

A high-scoring idea with thin evidence is a different decision from a lower-scoring idea backed by solid sources. The report treats them differently.

The full method is documented inside the package in docs/RESEARCH_METHOD.md.

What it can help you discover

• Keyword opportunities, scored and flagged as target, secondary, or skip
• Content coverage gaps in what you already published
• Search-intent gaps, terms people search that your content does not target
• Reader questions that stay unanswered
• Competitor coverage gaps in the surrounding landscape
• Audience problems that show up in question signals
• Content expansion opportunities with verdicts: refresh, expand, or merge
• Cannibalization verdicts, seven of them, including "do not create" so you stop competing with your own site
• Evidence-backed, prioritized opportunities with confidence scores
• SEO, AEO, and GEO relevant signals from public sources

One honest note: public sources can block or rate-limit requests. When that happens, the run reports it plainly and continues with the sources that worked. Nothing is filled in with a fake result.

Who it is for

Ranked by strongest fit:

1. Authors and book marketers. A manuscript becomes the seed. The engine researches reader questions, search intent, and competitor coverage around the book, and returns the content plan for the launch.
2. Bloggers and niche-site owners. An existing article becomes the seed. The engine returns the missing subtopics and search opportunities, with a cannibalization check.
3. Content strategists and content marketers. A content library becomes the seed. The engine returns coverage gaps and prioritized expansion targets.
4. Freelance researchers and consultants. You can hand a client a reproducible research process with sources attached, instead of a list they have to take on faith.
5. Newsletter creators and YouTube researchers. Past issues or scripts become the seed for the next topics, with evidence.
6. Digital-product creators, small agencies, and publishers. Anyone who researches content before creating it.

This is not a product for everyone. It is for people who create content and want research they can check.

Real-world use cases

AUTHOR: Feed it a book extract. It returns reader questions, keyword opportunities, and the gaps your book does not cover, each with evidence. That becomes the content plan around the book.

BLOGGER: Feed it an existing article. It returns the missing subtopics and search opportunities around that article, and flags topics you should not create because you already cover them.

CONTENT STRATEGIST: Feed it a content library. It returns coverage analysis and a prioritized expansion list with cannibalization verdicts.

RESEARCHER: Feed it source material. It investigates the surrounding questions, entities, and competitor coverage and returns an evidence-traced research brief.

NEWSLETTER CREATOR: Feed it past issues. It returns the reader questions your archive never answered and the topics worth building next.

What makes this approach different

Most keyword tools return numbers you cannot check. This engine returns a ledger. Every finding carries a source, an evidence class, and a confidence score.

Blocked sources are reported, never faked. Nothing gets a fabricated search volume or an invented difficulty score.

The engine is deterministic. The pipeline is fixed, so the same source produces the same run. Chat tools cannot do that.

The engine is local. Your source file and your results stay on your machine. It does not log into your accounts and does not publish anything on your behalf.

It works before your site exists. Domain-based tools like Ahrefs or Semrush need a live website to compare. This engine works from your material, published or not. That matters for authors, for new sites, and for anyone researching before they publish.

And it costs one payment, not a monthly subscription.

What you receive

A complete, self-contained package:

• The engine itself, a Python program that runs locally
• Configuration files and an environment template
• Engine documentation covering architecture, data model, adapters, limitations, installation, research method, skill architecture, and routing
• 79 unit tests for the documented self-check
• The skill layer, including operator, master writing, and social media skills, with documentation
• 22 ready-to-use prompts covering gap research, book research, keyword research, briefs, outlines, chapter work, image planning, QA, and troubleshooting
• 4 sample inputs, including a book extract, article, transcript, and URL notes, so you can test before using your own material
• Documentation in Markdown, PDF, and Word formats
• Product concept images
• A thank-you page with support contact and free resources
• START HERE, README, VERSION, CHANGELOG, and MANIFEST entry files

After purchase, you get the download instantly, plus a receipt with the download link.

How to get started

1. Download the package and extract the ZIP.
2. Open the package root and read START_HERE.md.
3. Install Python 3.10+ if you do not have it.
4. Open a terminal in the engine folder.
5. Run the environment check:

python -m adhil_engine init

1. Run the doctor:

python -m adhil_engine doctor

You want 0 FAIL at the end. WARN lines are normal; they mean optional credentials are not configured.

1. Optional: run the documented self-check:

python -W ignore -m unittest discover -s tests -v

1. Add your research input as a .txt or .md file. The samples in 02_SAMPLE_INPUTS let you test right away.
2. Run your first research and open the reports in the output folder. Start with research-summary.md.
3. Use the findings with the included prompts, or inside your own content workflow.

The user guide walks a complete beginner through a full 10-step first run, including a test with the sample book.

Technical requirements

• Python 3.10 or newer
• Standard library only, no third-party packages required
• Windows, Linux, or macOS
• Basic terminal comfort. You do not need to know how to code
• An internet connection for any step that pulls from a live public source
• Source files as plain text or markdown
• No account, login, or API key required to start

Optional API / adapters

The basic engine needs no credentials. It runs fully with public signals.

Optional Google Ads credentials can unlock the keyword planner adapter for official volume data. The engine checks credentials honestly and reports them as valid, invalid, or not provided. It never prints their values.

Two adapters are documented as not implemented in this release: the X API and the YouTube Data API. Public signals are used instead, and the source status table says so plainly.

If a service is blocked or unavailable, the run reports the status honestly and continues.

What it does not do

• It does not guarantee Google rankings or traffic
• It does not replace large proprietary SEO databases, rank trackers, or index-scale suites
• It does not automatically publish your content anywhere
• It does not pretend a blocked source was successfully researched
• It does not write your finished article. It stops at the research brief
• It is not a generic AI chatbot. It is a research engine with a fixed, checkable pipeline

If a writing skill is installed, the finished brief can flow into a first draft. Writing the final piece is your work, using the brief as its evidence base.

Honest limitations

• PDF is not a supported input in the current release. Convert it to plain text or markdown first
• One source file per research run. For several files, merge them into one text file or run them separately
• No graphical interface. It runs from a terminal
• It does not crawl your live website. The cannibalization check uses a content table you supply
• Public sources such as Reddit, Quora, and Google Trends can block requests. It is live behavior; it is reported honestly, and the run continues
• Official keyword volume requires optional Google Ads credentials. Without them, volume is marked as unavailable

Before you buy

This is a research engine, not a one-click SaaS dashboard. It requires a supported environment and a user who is comfortable following setup instructions.

The package ships with complete documentation, including START_HERE, a beginner user guide with a 10-step walkthrough, a technical how-to, and a troubleshooting prompt. Support is available by email.

If you are looking for a dashboard, automatic publishing, or guaranteed rankings, this is not the product for you.

If you create content and want research you can actually check, this is.

FAQ

What is a content gap?

A topic your content does not cover, while the surrounding evidence says it should, based on what people search for and what competitors already address. Keywords and gaps stay as separate outputs on purpose.

Who should use this engine?

Authors, book marketers, bloggers, niche-site owners, freelancers, consultants, content strategists, newsletter creators, and digital-product creators who research content before creating it.

Can authors use it?

Yes. Feed it a book extract as a text or markdown file. It adds book-specific research, including Amazon-derived signals labeled as proxies.

Can bloggers use it?

Yes. Feed it an existing article. It returns missing subtopics, search opportunities, and topics you should not create because you already cover them.

Can I use an existing article?

Yes. Articles, transcripts, newsletters, reports, drafts, and notes all work as text or markdown sources.

Can I use multiple research files?

One source file per run. Combine several files into one text or markdown file, or run them in separate runs. Both work fine.

Do I need an API key?

No. The engine runs fully without credentials. Optional Google Ads credentials can add official volume data later. Most buyers never configure them.

What happens when a source is blocked?

The run reports it honestly as blocked, rate-limited, or unknown, and continues with the sources that worked. Nothing gets a fake result.

Does it write the final article?

No. It produces a research brief with keywords, gaps, scored opportunities, and the evidence ledger. Writing is your work, or the work of a separate skill.

Does it guarantee rankings?

No. Nothing can honestly guarantee rankings. The engine helps you make better content decisions with evidence.

What operating system do I need?

Windows, Linux, or macOS, with Python 3.10 or newer. Standard library only, no extra packages.

What is included in the download?

The engine, configuration, documentation, unit tests, skill layer, ready-to-use prompts, sample inputs, documentation files, product images, thank-you resources, and the required entry files.

Can I use the research with another writing skill?

Yes. The research brief is the deliverable. Any compatible writer workflow can use it as its evidence base.

How is this different from a normal AI prompt?

A prompt gives you a new answer every time, with no source attached. This engine runs a fixed pipeline, saves an evidence ledger, writes a run ID, and reports blocked sources. You can check where every finding came from.

Final word

The engine takes something you already made and tells you, with sources attached, what it is missing and what is worth building next.

Then it stops exactly where your judgment needs to take over.

If you write and you want research you can check, start here.

Get the Content Gap Finding Engine.