SEO Strategy

The first result comes from AS400IBMSystems.com:

"IBM FlashSystem Glossary ... AS400 IBM Systems" (https://www.as400ibmsystems.com/glossary/ibm-flashsystem/)

We've recently beefed this one up considerably.

The second comes from the FlashSystem expert subdomain on AS400System.com:

"FlashSystem Glossary ... AS400System" (https://flashsystem.as400system.com/glossary/flashsystem/)

When I noticed this result, the page was ridiculously thin.

Two short blurbs and a CTA.

That's about it.

Naturally, immediately after discovering this, I had Codex turn it into something resembling a Wikipedia page.

Then realized:

Doh. I should have taken a screenshot first.

Welcome to Real SEO Life.

Fortunately, our historical search and warehouse data doesn't disappear just because I got excited and ruined my own control group.

The third result is even more interesting:

"IBM FlashSystems Glossary ... Midland Information Systems" (https://www.midlandinfosys.com/about/glossary-of-ibm-iseries-hardware-terms/ibm-flashsystems)

I created that glossary use case years ago.

Midland's glossary component never performed particularly well.

Neither did some of its directory architecture, despite the directory being something I've seen work extremely well elsewhere.

Now I'm suddenly seeing both of those old components surface for some important IBM and AS400 categories.

Some of the directory content is remarkably thin.

Nobody has been secretly optimizing it.

As far as I know, nobody there even knows how to use those components.

So what changed?

That's where my first theory started falling apart.

The three implementations aren't identical.

The content depth wasn't remotely comparable.

One page was substantial.

One was basically starving.

The Midland implementation came from an older architecture with different optimization work behind it.

So blaming the result on some magical glossary schema became harder to defend.

Then the skinny page became the interesting one.

Look at where it lives:

"flashsystem.as400system.com/glossary/flashsystem/"

That's almost comical.

FlashSystem subdomain.

AS400System domain.

FlashSystem glossary.

FlashSystem page.

Surrounded by an increasingly interconnected IBM ecosystem.

The page itself may have been weak.

The entity context around the page wasn't.

That's a much more interesting hypothesis.

This is where traditional SEO thinking can screw with how we interpret the evidence.

We see a ranking page and naturally ask:

What did this page do right?

Maybe that's the wrong unit of analysis.

Maybe the page didn't win.

Maybe the entity surrounding the page won.

The document was sitting inside an architecture sending extremely consistent signals about what that property represents and how FlashSystem relates to the larger AS400 and IBM ecosystem.

The page-level content was weak.

The entity-level context was strong.

That doesn't prove causation.

But it gives us a much better experiment than:

"Google likes glossaries."

And now, because I couldn't leave the damn thing alone, we've entered phase two.

We're not trying to make our content skinnier.

We're trying to make our entities fatter.

So the formerly skinny FlashSystem glossary page has officially entered Entity Boot Camp.

More useful content.

More semantic depth.

More explicit relationships.

Better structured data.

Stronger internal connections.

More corroborating evidence.

More ways for machines to independently arrive at the same conclusion:

This thing is about IBM FlashSystem.

Now we watch what happens.

If the thin page already earned visibility because of the entity architecture surrounding it, what happens when we combine that architecture with a substantially stronger document?

I don't know.

That's why we're doing it.

This is probably my favorite part of the experiment.

I could rewrite this story and pretend I understood what was happening from the beginning.

That's how a lot of SEO case studies get written.

Observation. Brilliant hypothesis. Perfect execution. Victory.

That's not how this happened.

I saw three glossaries.

I thought glossary signal.

Then one stupidly thin page made that explanation harder to believe.

So I changed my mind.

It's not bad to be wrong if it leads to being right.

That's the point of working from evidence instead of defending theories.

And it leads directly into the bigger lesson I took from the SERP.

Look at enough modern Google results and you start seeing roles.

The manufacturer gets its place.

Major brands get theirs.

Wikipedia gets its usual seat at the table.

Reddit seems to have a chair with its name permanently engraved on it.

Some of those results deserve to be there.

IBM absolutely should dominate certain IBM searches.

That's not really my complaint.

My question is:

Why would I build my entire strategy around trying to beat them?

That's a low-gain environment.

You can spend enormous amounts of time, links, content and authority trying to move from position five to position three in a SERP where Google has already demonstrated strong preferences about what belongs at the top.

Traditional SEO still matters.

Rankings still matter.

I'll happily take every ranking Google wants to give me.

But I'm increasingly less interested in making that the whole game.

Because there's another game sitting right next to it.

And remarkably few companies seem to understand they're playing it.