How to Fix a Site Hit by Multiple Algorithm Updates
On this page
A site that lost ground in several updates can have more than one problem, and a costly mistake is treating them as one and applying one fix. Recovery starts with attribution: match each drop to the update that coincided with it and to the part of the site that fell. Then fix the losses in order of what each costs the business, and watch for fixes that help one part of the site while hurting another. If your history includes a fix that “helped, then hurt,” that is the pattern to look for first.
Build the timeline by hand
No report does the attribution for you. Take the dates of each drop and line them up with the entries in Google’s ranking update history, noting each update’s start and end date. Google’s core updates documentation recommends waiting at least a full week after an update completes before analyzing the result, so don’t read one update’s effect off data that still includes the rollout.
Then split every drop by section. A blog that fell in one update and a set of review pages that fell in a later one are two problems with two fixes, and only the split shows it.
One grid, several problems
Put it in a matrix: site sections down one side, update dates across the top, and in each cell the change in clicks or position for that section around that update.
| Section | Update A | Update B | Update C |
|---|---|---|---|
| Guides | fell | held | held |
| Product reviews | held | fell | held |
| Category pages | held | held | fell |
The rows and verdicts in this sample grid illustrate the layout; they aren’t data. Filled in with your numbers, it can turn “we got hit by everything” into a short list of separate losses, each tied to a date and a section. It also shows the trap: if a fix you made after Update A lines up with a new loss in another section at Update B, the grid shows both events side by side.
Know which systems exist
Google’s guide to its ranking systems is the reference for what the losses could be answering to. Three of its entries matter for a site with several losses:
- Core ranking systems. The guide lists the helpful content system among its retired systems and says that in March 2024 it became part of the core ranking systems. There is no separate helpful content update to wait for.
- The reviews system, still listed as active, which aims to better reward high-quality reviews: insightful analysis and original research by people who know the topic well.
- Spam detection systems, including SpamBrain, which deal with content and behavior that violate Google’s spam policies.
A loss concentrated in review pages, a broad loss across informational content, and a loss tied to a spam problem are different diagnoses. Match each cell in your grid to the one it resembles, and keep the uncertain ones marked as uncertain.
Fix by value, not by date
Once the separate losses are named, don’t fix them in the order the updates arrived, and don’t start with the newest because it’s freshest. Rank the losses by what each costs the business and start at the top. If an older loss hit the pages that bring in revenue and a recent one hit a low-value section, the older one comes first.
Avoid the fix that hurts another layer
“Helped, then hurt” can come from large deletions. Removing weak pages may help one part of the site, but pages removed outright can take with them links from other sites that pointed at them. Google’s core updates documentation calls deleting content a last resort. So improve first, merge overlapping pages into one strong page with redirects, and when a page must go, redirect it to its closest real counterpart if other sites link to it.
Keep the disavow file out of it. Google’s disavow guidance reserves the tool for spammy links that have caused, or likely will cause, a manual action; it isn’t a response to a quality reassessment.
Recovery can come in stages
With several losses, recovery, when it comes, arrives in pieces as you fix each one. Google’s core updates documentation says improvements don’t necessarily wait for a major core update, and adds: “If it’s been a few months and you still haven’t seen any effect, that could mean waiting until the next core update.” Measure each section against its own baseline, and don’t read a partial recovery as failure.
Frequently asked questions
Why did an earlier fix help and then hurt?
A fix aimed at one kind of loss can cause another. Mass deletion can remove pages that other sites linked to, and the grid of update dates by section shows the fix and the new loss landing in different places at different times.
Can a tool attribute each drop to an update for me?
Tools can overlay your traffic on update dates, which shows alignment. Deciding which section fell with which update, and what that means, is analysis you do from the grid.
Is the helpful content update still separate?
No. Google’s ranking systems guide lists the helpful content system as retired and says it became part of the core ranking systems in March 2024.