The May 2026 core update dropped on May 21st, and honestly? It landed differently than March.

In March, I woke up to a full inbox and a Slack thread going sideways. This time I was following the Google I/O coverage from the day before and someone just quietly pinged me to say it had started. No sirens. Not yet, anyway.

We’re five days in as I write this, with roughly a week still to go on the rollout. I’m not going to pretend I have the full picture, because nobody does at this point. What I can share is what I’m seeing, what I think matters, and how I’m thinking about it for our clients.

This one came faster than I expected

March wrapped up on April 8th. May started on the 21st. That’s a 43-day gap, and in my experience running client sites through these cycles, that’s tight. You barely have time to assess what March did before May is already reshuffling things again.

For sites that took a hit in March and were in the middle of fixing things, this is a compounding problem. I’ve had a few conversations this week where the honest answer is: if the content work we talked about after March hasn’t happened yet, this update isn’t going to be kind.

Recovery from a core update tends to come with the next core update. This is that next one. Whether it rewards the work you put in or doubles down on the damage depends on what you actually did.

The I/O timing is worth thinking about

This update dropped one day after Google I/O 2026, where they announced what I’d describe as the most significant change to the search interface in years. A new AI-powered search box that changes how queries are formed and how results are presented.

Google hasn’t connected these two things publicly, and I’m not going to claim they’re directly linked. But I’d be naive to look at those two events a day apart and assume there’s no relationship between where Google’s product is heading and how their algorithm is being recalibrated to support it. Search is changing. The updates are changing with it.

What Google actually told us (which wasn’t much)

The official description is the same line we got in March: “a regular update designed to better surface relevant, satisfying content for searchers from all types of sites.” No companion post, no new guidance, no vertical-specific breakdown. The rollout window is up to two weeks.

One practical thing to flag: several people in our network have noticed that the Search Console link report started showing unusual data around the same time the update dropped. Big drops in reported links, or links showing as zero. Before you start making decisions based on that, check it against another source. In my experience, that looks like a reporting issue rather than actual link losses. Don’t pull the trigger on disavows or outreach based on GSC link data right now.

What I’m seeing in client data

I want to be careful here because five days is not a complete picture. Rankings are still moving. But some patterns are starting to show up, and they’re consistent with what I was already expecting based on March.

AI-generated content without a real person behind it is getting filtered. Not all AI content, but the kind that was clearly produced at scale without anyone with actual expertise reviewing it. The clients I’m seeing hold steady are the ones where AI helped with the process but a person with real knowledge shaped the final output. The clients who leaned on automated pipelines and called it done are the ones showing volatility.

YMYL content is getting the strictest look. Health, finance, legal. Every update I’ve watched over the past few years has held this category to a higher standard, and this one is no different. If you’re in one of these verticals and your author credentials aren’t clearly established on your pages, that’s worth fixing before the next update, not after.

Sites that publish on every topic under the sun are struggling. I keep coming back to the HubSpot example from March. They spent years publishing content far outside their core expertise, and they paid for it. That lesson hasn’t changed. Topical depth still matters more than topical breadth.

What I’d actually do right now

The temptation when an update is mid-rollout is to react immediately. I’d push back on that instinct pretty hard.

Don’t touch anything based on what you’re seeing this week. Rankings are still moving. What looks like a permanent drop today might stabilise in the next seven days, and what looks fine today might move again. Wait until the rollout is done, then wait another full week before you start drawing conclusions.

Set your measurement window correctly. When you do go into Search Console, you want to compare performance from before May 21st against performance from after approximately June 11th. Anything inside the rollout window is noise you shouldn’t be reacting to.

If you dropped, find out who replaced you before deciding what to fix. This is the step most people skip. If a direct competitor with similar content moved up, that’s a quality gap. If a broad authority site moved up, that tells you something different about how Google is weighting content type for that query. The diagnosis changes the fix.

Be honest about your content in Google’s own terms. Google’s people-first content guidelines include a self-assessment section worth reading through properly, not skimming. Run your affected pages against those questions. Ask someone outside your team to do the same. If the answers are mostly no, that’s where to start.

Where I think this is heading

I’ve been through enough of these to know that the updates which feel chaotic in week one often look more logical in month two. The sites that recover are almost always the ones that did the obvious thing: they made their content more useful, more specific, and more clearly authored by someone who knows what they’re talking about.

The pace is what’s different now. Two major core updates in 43 days, on top of a Discover update and a spam update earlier in the year. Google is moving faster and the window to get comfortable between cycles is shrinking. Treating content quality as something you revisit annually isn’t a viable strategy anymore.

I’ll share a proper breakdown once the rollout completes and there’s actual data to work with. In the meantime, sit tight.