How We Built an SEO System for a Small Website Using AI, Content Clusters & Technical SEO

When you launch a new website, SEO can feel like shouting into an empty room.

You publish an article.

Nothing happens.

You publish another.

Still nothing.

That was one of the challenges we faced while building SabrTime, a free Islamic digital companion focused on prayer, reflection, consistency, and useful Islamic tools.

Instead of treating SEO as simply "write more blog posts,"
We started building a repeatable system around technical SEO, search intent, content clusters, internal linking, and AI-assisted content production.

This is the workflow we are using and the lessons that can be adapted to almost any small website.

Start With the Website, Not the Blog

One of the biggest mistakes a new website can make is creating dozens of blog posts before understanding what the website actually offers.

We first looked at:

• Website structure
• Existing pages
• Search intent
• Important features
• Existing content
• Sitemap
• Metadata
• Internal links
• Indexability
• Structured data
• Open Graph information
• Canonical URLs

SabrTime isn't only a blog.

It is a web-based Islamic companion with tools and resources around prayer, Qibla, Tasbeeh, spiritual habits, and Islamic learning.

That meant the SEO strategy could connect informational content with actual useful tools.

The main website acts as the hub:

SabrTime → Tools → Resources → Blog → Educational content

This structure gave us many opportunities to build topical relevance.

You can explore the project here: "SabrTime" (https://sabrtime.in/).

Focus on Search Intent Instead of Just Keywords

Instead of targeting only highly competitive terms such as:

• Islamic app
• Prayer app
• Quran app

we started looking at specific problems users might actually search for.

For example:

• How to find Qibla direction
• Does airplane mode affect a Qibla compass?
• How to pray missed prayers
• What does Adhan mean?
• How to do Tasbeeh
• What is the Qibla direction for a city?
• How to build a daily Dhikr habit

These queries have much clearer intent.

For example, we created location-specific and educational Qibla resources rather than writing one generic article and stopping there.

The important part isn't simply putting a keyword into an article.

The article needs to actually solve the user's problem.

Build Content Clusters Instead of Random Articles

Instead of publishing unrelated articles, we started grouping content into topics.

For example, a Qibla content cluster can include:

Qibla

→ How to Find Qibla Direction
→ Qibla Direction in Degrees
→ Qibla Direction by City
→ Airplane Mode & Digital Qibla Compass
→ History of the First Qibla

For example, our "Qibla Direction in Dallas" (https://sabrtime.in/blog/qibla-direction-dallas-tx.html) article focuses specifically on the Qibla direction for Dallas.

We also created content around "airplane mode and Qibla compasses" (https://sabrtime.in/blog/airplane-mode-qibla-compass.html).

These aren't random articles.

They're different pieces of content supporting the same broader topic.

This approach can be used for almost any niche.

A fitness website could create:

Fitness

→ Home workouts
→ Beginner workouts
→ Push-up guide
→ Squat guide
→ Workout recovery

A SaaS website could create:

Email Marketing

→ Email marketing guide
→ Email subject lines
→ Open rates
→ Email automation
→ Newsletter strategy

The principle is the same:

«Build topical depth instead of publishing random keywords.»

AI Helped Us Scale the Content Workflow

Claude became particularly useful in our content-development workflow.

But we didn't simply tell AI:

«"Write me an SEO article."»

Instead, we used a structured process.

Topic

↓

Search intent

↓

Research

↓

Article structure

↓

Related questions

↓

Useful information

↓

Internal links

↓

CTA

↓

SEO metadata

↓

Human review

↓

Publish

Claude helped with research organization, outlining, drafting, semantic topic coverage, internal-link suggestions, and optimization.

But the final content still needed human review.

This distinction is important.

AI can accelerate content production.

It shouldn't replace editorial judgment.

Internal Linking Became Part of the Writing Process

Internal linking became one of the most useful parts of the workflow.

Instead of publishing an article and forgetting about it, we looked for natural connections between pages.

For example:

A Qibla article can link to another Qibla guide.

A Tasbeeh article can link to a digital Tasbeeh resource.

A prayer article can link to prayer-related resources.

A Qaza article can link to other relevant prayer content.

For example, we have a practical "Missed Prayers / Qaza guide" (https://sabrtime.in/blog/missed-prayers-qaza-guide.html) that can naturally connect with other prayer-related content.

Our "Tasbeeh guide" (https://sabrtime.in/blog/tasbeeh-guide.html) can similarly become part of a broader Tasbeeh and Dhikr content cluster.

This creates a path for both users and search engines:

Article → Related article → Tool/resource → Website

The goal isn't to add links everywhere.

The goal is to add links where they genuinely help the reader.

Connect Informational Content With the Actual Product

This is where content strategy can become much more powerful for SaaS and tool websites.

Don't create content that has nothing to do with your product.

Instead:

User problem → Helpful content → Relevant tool

Someone might search for information about maintaining a daily spiritual habit.

They discover an educational article.

After learning about the topic, a relevant SabrTime feature can become a natural next step.

The same model works outside the Islamic niche.

SaaS

"How to manage invoices"

↓

Invoice management guide

↓

Invoice software

Developer Tool

"How to convert JSON"

↓

JSON tutorial

↓

JSON converter

Fitness

"How to track workouts"

↓

Workout tracking guide

↓

Workout tracker

The content becomes an entry point to the product.

Optimize Pages Outside the Blog

Another lesson was that SEO shouldn't be limited to "/blog/".

Important search landing pages can include:

• Homepage
• Feature pages
• Tools
• Resource pages
• Ebook pages
• Guides
• Calculators

SabrTime also has educational ebook resources.

For example, the "New Muslim Starter Guide" (https://sabrtime.in/e-books/new-muslim-starter-guide.html) is not a traditional blog post, but it can still serve a useful search and user-intent purpose.

The same principle applies to almost any website.

A calculator can rank.

A tool can rank.

A documentation page can rank.

An ebook can rank.

A resource page can rank.

You don't have to put everything inside "/blog/".

Technical SEO Runs Alongside Content SEO

While creating content, we also worked on the technical foundation.

Our checklist included:

• XML sitemap
• Robots.txt
• Canonical URLs
• Meta titles
• Meta descriptions
• Open Graph tags
• Structured data
• Mobile usability
• Clean URLs
• Internal linking
• Indexability

The objective is simple:

Google needs to discover, crawl, and understand your pages.

A great article isn't enough if search engines can't properly discover or understand it.

Create Location-Specific Content Carefully

Another experiment was creating highly specific pages around search intent.

For example:

Qibla Direction in Dallas

Qibla Direction in Houston

Qibla Direction for other cities

This can work when each page provides genuinely useful location-specific information.

For example, the "Dallas Qibla Direction guide" (https://sabrtime.in/blog/qibla-direction-dallas-tx.html) is designed around the search intent of someone specifically looking for Qibla direction in Dallas.

The important lesson is:

«Don't create hundreds of city pages simply by replacing the city name.»

Each location page should provide unique and useful information.

Otherwise, you're just creating programmatic SEO spam.

Use Search Data as a Feedback Loop

Publishing is only the beginning.

After pages are indexed, Google Search Console can reveal:

• Search queries
• Impressions
• Clicks
• Average position
• Pages receiving impressions
• Unexpected queries

Suppose an article starts receiving impressions for:

«"qibla without internet"»

That could tell us something.

Maybe the existing article should be expanded.

Maybe a dedicated article would be useful.

Maybe the search intent is different from what we originally expected.

This creates a feedback loop:

Publish → Measure → Learn → Improve → Publish again

That's much better than publishing once and never touching the article again.

Turn Existing Content Into New Opportunities

One of the biggest benefits of content clusters is discovering new topics from existing articles.

A general Qibla guide might reveal related questions:

• Can I use a compass without internet?
• Does GPS work in airplane mode?
• Why is my Qibla compass inaccurate?
• What is Qibla bearing?
• Is Qibla always west?

Instead of stuffing every question into one enormous article, we can decide whether each deserves:

• A section
• An FAQ
• A separate article
• Or simply an internal link

This keeps the website organized while allowing the content library to grow naturally.

External Publishing Can Support SEO and Awareness

We also started publishing useful educational and technical content on external platforms.

Platforms such as DEV Community can be useful when the article genuinely provides value to developers, builders, marketers, or website owners.

But the goal shouldn't be:

«"I need a backlink."»

The better approach is:

«"I have a useful case study that other website owners can learn from."»

This article itself is an example.

SabrTime is the real-world case study, while the actual subject is the SEO process.

That makes the references contextual rather than purely promotional.

Useful external content can potentially generate:

• Referral traffic
• Brand awareness
• Relevant backlinks
• Community engagement
• New relationships
• Search visibility

But repeatedly publishing thin promotional articles with the same anchor text is not a sustainable link-building strategy.

AI Content Still Needs Human Review

AI makes publishing faster.

But speed isn't the same as SEO quality.

A website could generate hundreds of articles with AI.

That doesn't automatically create authority.

Our approach is closer to:

AI-assisted research + human strategy + useful content + internal linking + technical SEO + continuous improvement

The human still needs to decide:

• Is this topic worth covering?
• Does the article satisfy search intent?
• Is the information accurate?
• Is the article actually useful?
• Are the internal links relevant?
• Is the CTA appropriate?
• Is anything repetitive or generic?

AI is the accelerator.

The website owner still needs to be the driver.

What Other Website Owners Can Copy

You don't need an Islamic website to use this system.

SaaS

Problem → Guide → Tool → Product

E-commerce

Question → Buying guide → Category → Product

Developer Tools

Problem → Tutorial → Documentation → Tool

Education

Question → Explanation → Resource → Course

Local Business

Question → Local guide → Service → Contact/Booking

The niche changes.

The framework stays almost the same.

The SEO System We Are Building

After putting everything together, our basic system looks like this:

Research

↓

Find specific search intent

↓

Create genuinely useful content

↓

Build topic clusters

↓

Connect pages with internal links

↓

Optimize technical SEO

↓

Publish useful external content

↓

Earn relevant backlinks and referral traffic

↓

Monitor Search Console

↓

Improve existing pages

↓

Repeat

Final Takeaway

Our biggest SEO lesson from building SabrTime wasn't a particular keyword or backlink.

It was learning to think of the website as an ecosystem.

A blog post shouldn't exist alone.

A useful article can lead to another article.

That article can lead to a tool.

The tool can lead to the main website.

External educational content can introduce new people to the ecosystem.

And search data can tell us what to build next.

The basic principle is:

«Don't build a pile of articles. Build a connected knowledge system.»

AI can help you build that system faster, but strategy, accuracy, usefulness, and originality still matter.

That's the approach we're continuing to use while growing "SabrTime" (https://sabrtime.in/) from a small website into a broader collection of free Islamic digital tools and educational resources.

Explore the project: "sabrtime.in" (https://sabrtime.in/)

Explore the content: "SabrTime Blog" (https://sabrtime.in/blog/)

Comments

[Dmitry Sergeev]

{"title":"How We Built an SEO System for a Small Website Using AI, Content Clusters & Technical SEO","comment":"how much of the content clustering did you actually automate? i am worried the ai just generates generic filler that hides behind good titles"}

[Yunus Khan]

That’s a fair concern. We didn’t fully automate content clustering and publish whatever AI generated.

AI helped with keyword grouping, identifying search intent, finding topic relationships, and suggesting cluster structures. But the actual topics, overlap, usefulness, and final content were manually reviewed.

The goal was to use AI for the research and organization layer—not to generate filler just to target more keywords. We also tried to make each article answer a distinct user question and link naturally to related content.

For SabrTime, the biggest test is still whether the pages provide genuine value and gain impressions/clicks over time. If a page doesn’t add something useful, there’s no reason to keep it just because AI produced it.