DEV Community: SeoGeo Tech

I Built a Technical SEO Website. The Biggest Lesson Had Nothing to Do With SEO.

SeoGeo Tech — Wed, 22 Jul 2026 15:54:44 +0000

Traffic dropped?

Perhaps the article wasn't long enough.

I kept assuming every unusual signal meant I had done something wrong.

Over time, I realized that many of those assumptions were false.

Sometimes a page simply hadn't been crawled yet.

Sometimes search engines just needed more time.

Sometimes the page was already technically healthy.

I was trying to solve problems that didn't actually exist.

SEO taught me patience

That surprised me.

People often describe SEO as a technical discipline.

And it is.

You need to understand crawling, indexing, canonical URLs, structured data, page performance...

But another skill turned out to be just as important:

Knowing when not to touch a page.

Now, before changing anything, I ask myself three questions:

Is there evidence that something is actually broken?
Can I reproduce the issue?
Am I solving a real problem, or just reacting to a dashboard?

Those questions have saved me from making far more unnecessary changes than any SEO tool ever has.

Building in public changed how I write

Another unexpected lesson was about writing itself.

Instead of publishing articles because a keyword tool told me to, I started publishing articles because I had tested something.

That completely changed my workflow.

Every article now starts with a real observation.

A Search Console report.

A Bing Webmaster Tools status.

A structured data test.

A publishing workflow.

An AI Overview citation.

The article becomes a record of an experiment rather than another opinion piece.

I think that's much more useful—for me and for readers.

Small websites have an advantage

Large websites often publish polished guides.

Small websites can publish discoveries.

I don't have a huge team.

I don't have millions of page views.

But I can document what I'm learning this week.

I can update an article when I find new evidence.

I can admit when my original assumption was wrong.

That honesty has become one of my favorite parts of running a technical website.

I'm curious about your experience

If you run a blog, documentation site, or technical website:

What's one assumption about SEO that turned out to be completely wrong?

Mine was believing that every unexpected search engine status required immediate optimization.

It didn't.

Sometimes the best optimization is simply collecting more evidence before changing anything.

If you're interested in Technical SEO, structured data, AI Search, or GEO experiments, I document everything I test on SeoGeo Tech:

👉 https://seogeo.tech

I'd genuinely love to hear how other developers and site owners approach this. We all end up learning different lessons from the same search engines.

How My Small Website Was Cited in Google AI Overviews (And What I Learned)

SeoGeo Tech — Tue, 14 Jul 2026 04:27:19 +0000

How My Small Website Was Cited in Google AI Overviews (And What I Learned)

A few days ago, I noticed something that genuinely surprised me.

While searching Google, I found that one of my articles had been cited in Google AI Overviews.

For a small personal website, seeing that happen was exciting.

I'm not writing this because I've discovered some secret GEO strategy or because I think I've figured out how Google's AI chooses sources.

I haven't.

Instead, I wanted to document one real observation, explain what my website looked like before it happened, and share a few lessons that might be useful if you're building websites in the AI search era.

The website

I run a small project called SeoGeo Tech, where I publish practical articles about Technical SEO, structured data, and Generative Engine Optimization (GEO).

It's not backed by a company.

There isn't a marketing team behind it.

I don't buy traffic or run backlink campaigns.

The site is simply a place where I document experiments, publish implementation guides, and share what I'm learning while building in public.

What happened

One of my articles appeared as a cited source in Google AI Overviews.

When I first saw it, I refreshed the page several times because I wanted to make sure it wasn't a temporary result.

It wasn't.

Seeing a small independent website referenced by AI-generated search results was encouraging, especially because the site is still relatively new.

What I focused on

People often ask whether GEO is completely different from traditional SEO.

After this experience, my personal opinion is that many of the fundamentals are still the same.

Instead of trying to "optimize for AI," I focused on building pages that are useful, technically sound, and easy to understand.

Some of the things I consistently worked on include:

Writing original, experience-based content instead of rewriting existing articles
Organizing content with clear headings and logical structure
Adding screenshots where they improve understanding
Using descriptive page titles and meta descriptions
Implementing structured data where appropriate
Keeping internal links organized
Making pages load quickly
Publishing content around a consistent topic instead of covering unrelated subjects

None of these are secret techniques.

Most of them are simply good publishing practices.

What I didn't do

I think it's equally important to mention what didn't happen.

I didn't suddenly gain thousands of backlinks.

I didn't use any AI visibility service.

I didn't purchase traffic.

I didn't submit the article to Google in any special way.

And I certainly don't expect every article to appear in AI Overviews.

This is only one observation from one website.

It shouldn't be treated as proof that any particular optimization caused the citation.

My biggest takeaway

Before starting this project, I assumed GEO would require an entirely new set of optimization techniques.

Now I'm less convinced that's true.

To me, GEO feels more like an extension of good SEO than a replacement for it.

Helpful content still matters.

Clear structure still matters.

Trustworthy information still matters.

Technical quality still matters.

If AI systems are selecting sources from the web, these characteristics seem just as valuable today as they were before AI-generated search became popular.

What's next

I'm continuing to publish technical SEO and GEO articles while documenting everything publicly.

Some future experiments I'm interested in include:

Comparing different article structures
Testing how visual evidence affects AI citations
Measuring changes after technical SEO improvements
Observing how structured data influences AI search visibility over time

Whether these experiments produce positive results or not, I plan to share the outcomes.

I think the community benefits more from honest observations than from bold claims.

Final thoughts

This experience didn't convince me that GEO replaces SEO.

If anything, it reminded me that solid fundamentals still matter.

As AI search continues to evolve, I suspect websites that consistently publish useful, trustworthy, and technically well-structured content will continue to have an advantage.

That's the direction I'm planning to keep following.

If you're running similar experiments with AI search, I'd love to hear what you've observed.

For anyone interested, the original article (with additional screenshots and details) is available here:

https://seogeo.tech/articles/google-ai-overview-case-study/

How I Stopped Writing JSON-LD by Hand

SeoGeo Tech — Sun, 12 Jul 2026 08:21:35 +0000

How I Stopped Writing JSON-LD by Hand

If you've worked on more than a few websites, you've probably written the same JSON-LD markup over and over again.

An Article schema here.

A BreadcrumbList there.

An Organization object on every site.

None of it is particularly difficult—but it quickly becomes repetitive.

The Problem

When I started building a technical SEO website, I found myself spending more time generating structured data than I expected.

The workflow usually looked like this:

Decide which Schema.org type to use.
Write the JSON-LD manually.
Validate it.
Fix formatting issues.
Repeat for the next page.

This worked for a while, but it wasn't scalable.

What I Wanted Instead

I wasn't looking for an all-in-one SEO platform.

I simply wanted a tool that could:

Generate clean JSON-LD
Cover the most common Schema.org types
Work instantly in the browser
Require no account or setup

So I built one for my own publishing workflow.

What I Learned

Building the tool reminded me that the biggest productivity gains in technical SEO don't always come from complex automation.

Sometimes it's about removing small, repetitive tasks that interrupt your workflow dozens of times every week.

Structured data is one of those tasks.

I'd Love Your Feedback

I'm still improving the generator based on real-world use cases.

Which schema types do you use most often?

Is there anything missing from existing JSON-LD generators that slows down your workflow?

If you'd like to try the tool or share feedback, you can find it here:

https://seogeo.tech/tools/json-ld-schema-generator/

I'm always interested in hearing how other developers and technical SEOs handle structured data in their projects.