Google is moving the furniture again.

On 24 September, Google began rolling out its September 2026 spam update worldwide. It's global, applies across all languages, affects search rankings, and Google says the rollout may take up to two weeks to complete.

Search watchers were already reporting significant ranking volatility around the time the update began, so inevitably people are now trying to work out what's changed, who has been affected and what they should do about it.

There's just one slight problem.

Google hasn't told us exactly what it changed.

Google has confirmed the spam update, but it hasn't identified a particular spam tactic or policy being targeted by this rollout. That hasn't stopped the speculation, of course, but speculation shouldn't be confused with something Google has actually announced.

And so we enter that familiar period after a Google update. Rankings move, traffic changes and screenshots start appearing as the SEO world attempts to identify winners and losers. Then comes the inevitable effort to reverse-engineer what Google has done, often while the update itself is still rolling out.

Meanwhile, at Citewell...

We've changed nothing.

Not because we know some secret about Google's September update. We don't. And certainly not because we think Citewell users are somehow immune to changes in search rankings.

We've changed nothing because our underlying strategy doesn't depend upon predicting Google's next algorithm update.

Wouldn't that make life easier?

Instead of repeatedly asking “What does Google want this month?”, we've been building Citewell around a rather different question:

What does your organisation genuinely know, and have you made that knowledge sufficiently clear, discoverable and evidenced for machines to find and use?

That was our strategy before the September update. It's our strategy while the update is rolling out, and nothing Google has announced so far gives us a reason to change it.

So, what's actually happening at Google?

This is Google's fourth confirmed spam update of 2026, following updates in March, June and August. Those earlier updates completed relatively quickly, whereas Google has allowed up to two weeks for September's rollout. That alone makes this one worth watching.

What we shouldn't do is invent a technical explanation that Google hasn't provided.

Google hasn't announced that the September update specifically targets AI-generated articles. It hasn't announced a new penalty for “thin content”, nor has it said that businesses suddenly require some particular schema configuration to survive the update.

What Google has announced is a spam update.

Google's existing spam policies cover practices including scaled content abuse and other attempts to manipulate search rankings, and Google advises sites affected by spam updates to review those policies. But precisely what has changed inside the September rollout remains undisclosed.

So we'll watch it. We just aren't going to rebuild our strategy around guesses about it.

Because there's a bigger change happening

Search itself is changing.

Traditional search largely asked:

Which page should I show?

AI-mediated discovery increasingly introduces another question:

What information can I confidently use?

That's an important distinction.

Producing another article about something everybody else has already written about doesn't necessarily contribute additional knowledge. Neither does rewriting the same information with different keywords.

The internet doesn't have a shortage of content. Generative AI has made producing more of it extraordinarily easy.

The harder problem is finding useful knowledge, understanding where it came from, establishing who knows it and finding sufficient evidence to decide whether it deserves to be used.

That's the problem we're interested in at Citewell.

And we've started approaching parts of it from some interesting directions.

Gap-Cite: what do they show that you don't?

One of those directions is something inside Citewell called Gap-Cite.

Gap-Cite looks outward. It examines the information landscape around your organisation and your competitors, looking for gaps between the useful knowledge, information, attributes and evidence they make discoverable and what you do.

That's subtly different from conventional keyword-gap analysis.

A competitor may explain something you don't. They may expose an important attribute you've overlooked, answer a question you've never addressed, or simply make useful knowledge visible that remains buried inside your own organisation.

Finding one of those gaps doesn't automatically mean “write another blog post.”

The more interesting question might be: Do we already know this?

If you do, the problem isn't necessarily content creation. The problem is that valuable organisational knowledge hasn't been made discoverable.

And that brings us to something else we've been working on.

Take-Cite: is that really the best source?

Take-Cite approaches the landscape from another direction.

It looks at sources already earning citations around topics and competitors relevant to an organisation and asks whether the source currently being cited is actually a particularly strong one.

Because a citation isn't permanent ownership of a subject.

Sometimes the source being relied upon is shallow or incomplete. Sometimes another organisation possesses considerably deeper experience, better evidence or knowledge that goes beyond what the existing source has documented—but has never made that knowledge discoverable.

That's an opportunity.

The answer might be to create a new knowledge asset. Alternatively, it might mean taking something you already publish and pushing its Citation Line further by adding genuine organisational knowledge beyond the point at which the currently available evidence stops.

The objective isn't to copy the cited source or manufacture another version of it.

It's to build the better source.

That's why we call it Take-Cite.

And back to September...

Google's September update is still rolling out. Rankings may continue moving while it does, and businesses should be careful about drawing firm conclusions from movements that occur before the rollout has finished.

We'll be watching it too.

What we won't be doing is sitting inside Citewell wondering how quickly we can build a September 2026 Google Update Optimiser.

There isn't one.

We're still working on the same problem we were working on last week: helping organisations make more of what they genuinely know discoverable, understandable and usable.

Gap-Cite is one part of that. Take-Cite is another. The Citation Line sits underneath both.

Google will change again, and inevitably there will be another update after that. Perhaps the more important question is whether building an entire digital strategy around constantly chasing those changes is itself something that needs updating.

Wouldn't that make life easier?

One more thing before we go...

While Google has been busy rolling out its September update, we've been busy with something of our own.

There's some fairly big news coming from Citewell later this week. It's something we've been working towards for quite a while, and it says rather a lot about where we think AEO, GEO and organisational knowledge are heading.

We're not going to spoil it just yet.

Keep an eye out later this week.