What Google's 2026 Core Updates Actually Reward

Google ran two core updates in 2026, and said essentially the same thing both times: write helpful, people-first content.

That sentence has been repeated so often that it has stopped meaning anything. Business owners nod at it, then go back to publishing the same pages they were publishing before. So let us make it concrete: what it means, what these updates changed, and how you can test your own pages against it in an afternoon.

The 2026 core updates at a glance

The second update was reported to affect health, finance and other "your money or your life" topics again. That matters if you write about anything where a wrong answer costs someone money or wellbeing.

A core update is not a penalty

This is the single most useful thing to understand.

A spam update punishes rule-breaking. A core update does something different: it re-evaluates which pages best satisfy a search, across the whole web, and adjusts accordingly. When your page drops, it usually means other pages were judged more useful, not that you did something wrong.

Two consequences follow. There is rarely a single "fix", because you are not being punished for a specific thing. And recovery is not guaranteed to happen at the next update, because Google's assessment of the competition keeps moving too.

What "helpful" actually means

Google publishes a set of self-assessment questions for exactly this. Stripped of jargon, they ask four things.

Is there original value? Does the page offer information, research, analysis or a perspective that is not simply a rewording of what already ranks?

Is there real expertise behind it? Can a reader tell who wrote it, what they know, and why they should be believed? Does it read as if the author has done the thing?

Is it complete? Does the page answer the question properly, or does it make the reader go and search again?

Would someone trust and bookmark it? Not "would it rank", but "would you send it to a colleague?"

Notice what is absent: word count, keyword density, publishing frequency, the number of headings. Those are what people optimise when they do not know how to answer the four questions above.

Where business sites usually lose

In our experience of reviewing business sites, pages that lose ground tend to share a handful of weaknesses.

• They are generic. A service page that could belong to any competitor by changing the logo.

• They hide the author and the evidence. No named people, no case studies, no proof that the claims are true.

• They answer a different question from the one searched. A page about "web design cost" that talks about the agency rather than the cost.

• They are padded. Seven hundred words that say what three hundred would.

• They are stale. Statements about "2023 trends" in a page still live in 2026.

None of these is a technical issue, and that is the point. Core updates reward the thing that is hardest to fake: a page that is actually better.

A one-afternoon audit

Pick your ten most important pages: the ones that should bring in enquiries. For each, answer these questions honestly, in a spreadsheet.

• What does this page say that the top three results do not?

• Who wrote it, and is that visible?

• What evidence supports the claims? A case study, a number, a photograph, a quote?

• Does it answer the question in the opening lines?

• When was it last meaningfully updated, and does it say so?

• What would a sceptical visitor still not know after reading it?

A page that fails three or more is a rewrite candidate. A page that fails all six should probably be merged into something stronger or removed.

An example of the difference

Take a corporate website page that says: "We deliver innovative digital solutions tailored to your needs." It could belong to anyone, so it is no use to a reader and no use to Google.

Compare it with how a case study can say it. When we rebuilt a cement manufacturer's website, the central design problem was getting two entirely different audiences, dealers and end customers, to the right place in under three clicks. A page that explains that, with the decisions made and the reasoning behind them, is the kind of thing nobody else can publish.

That is the practical meaning of "people-first": say the specific, true thing only you can say.

What to do this month

• Run the six-question audit on your ten key pages.

• Rewrite or merge the weakest three, and add a named author and a real example to each.

• Update any page that mentions a year or a trend.

• Leave the rest alone. Constant tinkering is not a strategy, and a core update's effects can take weeks to settle.

• Re-check Search Console a few weeks after the update finishes, comparing like-for-like periods.

Frequently asked questions

What is a Google core update? It is a broad change to the systems Google uses to rank pages. It is not aimed at specific rule-breakers but re-evaluates which pages best satisfy a search.

How long does a core update take? Google typically says up to two weeks. The March 2026 update took about 12 days.

Why did my traffic drop after a core update? Usually because other pages were judged more helpful for the queries you were ranking for. Compare Search Console data from before and after, and look for which pages and queries fell.

Can I recover from a core update? Often, but not always at the next update. Improving the pages that lost ground helps, and results can take several months to show.

Do I need more content to recover? Not usually. Better content on fewer pages tends to beat more content of the same quality.

Related reading: the 12-point spam self-check and our web design and development service.

Want an outside view of your weakest pages? Our free audit highlights where your site is losing visitors and trust, and ranks what to fix first. Request your Free Website Audit →