Google's 2026 Core and Spam Updates: What Small-Business Sites Actually Need From Their Agency Now
Five documented ranking updates in 2026, all completed. What a small-business owner should ask their agency after one — and what answers should worry them.
Google’s Search Status Dashboard is the only authoritative record of when a ranking update happened, and it is public. For 2026 so far it lists five completed entries:
| Update | Started | Duration |
|---|---|---|
| February 2026 Discover update | 5 February | 21 days, 17 hours |
| March 2026 spam update | 24 March | 19 hours, 30 minutes |
| March 2026 core update | 27 March | 12 days, 4 hours |
| May 2026 core update | 21 May | 11 days, 21 hours |
| June 2026 spam update | 24 June | 2 days, 1 hour |
Two core updates, two spam updates, one Discover update. All finished. Nothing new has been confirmed since the June spam update completed in late June, which means the search results a small-business site is looking at right now have had a quiet stretch to settle.
That table is worth keeping because most of what a small-business owner hears about Google updates does not come from it. It comes from an agency’s monthly report, and the gap between the two is where a lot of unnecessary invoicing lives.
The dates matter more than the theories
The single most useful thing an owner can do after a traffic drop is compare the drop’s start date against that table.
If traffic fell on 22 May, the May core update — which began on 21 May — is a plausible explanation and the agency should be talking about content quality and site purpose. If traffic fell on 12 May, it was not a core update, and any report attributing it to one is guessing. The March pair is a particularly good test of an agency’s rigour, because the spam update on 24 March and the core update on 27 March were three days apart. An agency that blames “the March update” without distinguishing which one has not looked closely.
Note also the durations. The March core update ran twelve days; the May core update nearly twelve. Rankings move throughout a rollout and frequently move back. A panicked rewrite on day three of a core update is a decision made on incomplete information, and it also destroys the baseline needed to understand what the update actually did.
What a spam update is not
Small-business owners consistently over-read spam updates, and agencies sometimes let them.
Spam updates target manipulative practices — the kinds of things Google’s spam policies name explicitly. A legitimate local business with an ordinary website is not the target. If a site lost visibility during the June spam update, the honest first question is not “how do we recover” but “what were we doing that a spam system would catch,” and the answer is sometimes something an agency or a predecessor did: purchased links, a network of thin location pages built to spin one service across sixty towns, scraped or spun content padding out a blog.
The duration is telling here too. The March spam update completed in nineteen and a half hours; the June one in two days. These are fast, targeted passes, not broad reassessments.
Questions worth asking your agency
“Which specific dated update do you believe affected us, and what does our analytics show on that exact date?” A real answer names one of the five and shows the graph. A vague answer about “the algorithm” is not an answer.
“Did our competitors move too?” A core update reshuffles a result set. If a site dropped while everything around it stayed put, that points at something site-specific — a technical problem, a manual action, a content change the agency made — rather than at Google.
“Have you checked Search Console for a manual action?” Manual actions are not algorithm updates. They appear in Search Console with a description, and they are fixed by a specific remedy and a reconsideration request. Conflating the two wastes months.
“What did we change in the four weeks before the drop?” Redesigns, URL changes, template migrations, and plugin swaps cause traffic losses that arrive on their own schedule and get attributed to whatever update was nearest.
Answers that should worry you
A proposal for a large content rewrite delivered during an active rollout. An invoice for “algorithm update recovery” that does not name which update. A claim that the agency has a relationship or channel that lets it resolve ranking drops faster. An explanation that leans on a leaked-document theory rather than the dated record.
And most common of all: a report that treats every fluctuation as an update. Rankings move constantly without any confirmed update running. Five documented events in eight months is not a lot of weather.
The reasonable posture
Watch the dashboard rather than the commentary. When a drop lines up with a dated core update, expect the agency to assess content honestly — is this page genuinely the best answer for the person searching, or is it there because a keyword tool said so — and then to make considered changes and wait a full cycle to read the result.
When a drop does not line up with anything on that list, the cause is almost certainly closer to home, and it is usually cheaper to fix.