Three years ago a client rang me at 7am on a Friday. Traffic down 60% overnight. Revenue halved. He was convinced someone had hacked the site. It wasn't a hack. Google had just rolled out a broad core algorithm update, and his 340-page affiliate site had taken a direct hit. He'd already deleted half his content by the time he called me.
That's the worst thing you can do. And yet it's the first thing almost everyone does.
After building and recovering sites across 12 years and well over 12,000 projects at Seahawk Media, I've developed a specific process for core update recovery that actually moves rankings rather than just producing a 40-slide audit deck nobody acts on. This is that process.
---
Step One: Confirm It's Actually a Core Update Drop
Before you touch a single line of content, you need to be certain the traffic drop is algorithm-related and not something mundane. I cannot tell you how many times a client has blamed Google when the culprit was a broken canonical tag, a noindex left on production after a staging migration, or a server going down for four hours during Googlebot's crawl window.
Pull Google Search Console and look at the Performance report. Filter by date to isolate the 72-hour window around the update rollout. If impressions and clicks dropped together, it's almost certainly algorithmic. If clicks dropped but impressions held, check your CTR and look for a title tag change. If impressions fell off a cliff but clicks were already near zero, crawling or indexation is your problem, not ranking signals.
Cross-reference with Google's confirmed update history. The dates matter. If your traffic started falling two weeks before any confirmed update, stop looking at content quality and start checking your tech stack.
I run this diagnosis in Screaming Frog first, always. A quick crawl to verify canonical tags, index status, internal link chains, and redirect health takes 20 minutes and rules out 40% of the "it must be the algorithm" cases I see.
---
Step Two: Identify Which Pages Actually Lost Rankings
Most people look at overall traffic. That's too broad. You need page-level data.
In Search Console, go to Pages (under Performance), sort by clicks, and compare the last 28 days against the equivalent period before the update. Export it. Then sort your biggest losers into three buckets:
- Pages that lost impressions AND rankings (Google demoted them in quality assessment)
- Pages that lost clicks but kept impressions (they're still indexed, titles or meta descriptions are underperforming post-update)
- Pages that disappeared entirely from the index (crawl or canonicalisation issue)
Each bucket needs a different fix. Treating them all the same is why most recovery attempts fail.
Back in 2022, Seahawk had a travel client where 80% of their traffic loss came from bucket two. Their pages were still ranking at positions 4-8, but Google had started showing AI-generated snippets and rich results above them. The "rankings didn't actually drop" revelation saved us from a pointless content rewrite project and led us to fix structured data instead. Traffic recovered by about 35% in six weeks without touching a word of content.
---
Step Three: Run a Genuine Content Quality Audit
Here's the thing. Google's own guidance on core updates is frustratingly vague ("produce helpful content for users") but the Quality Rater Guidelines are 170 pages of specificity. Read them. Specifically the E-E-A-T section.
What I actually look for when auditing content post-update:
- Thin content with no original insight. If the page says the same thing as the top three results, Google has no reason to rank it above them.
- Author credibility signals. No byline, no bio, no linked credentials. Google's quality raters check this manually.
- Topical depth mismatches. A 400-word page trying to rank for a query that clearly demands long-form treatment.
- Factual outdatedness. Dates, statistics, product names, pricing. A page referencing "2021 data" in 2024 reads as neglected.
- User intent mismatch. This one kills more sites than any other issue. If someone searches "how to fix a leaking radiator" and your page sells radiator replacement services, the intent is wrong. Rewrite or create a separate informational page.
I use a scoring sheet with 12 criteria rated 1-3. Any page scoring under 24/36 goes into the rewrite queue. Pages scoring under 18/36 I consider for consolidation or removal.
Consolidation (merging two thin pages into one strong one) is almost always better than deletion. I learnt this the hard way in 2020 when I deleted 60 pages from a client's site on the assumption that fewer, better pages would improve overall domain quality. We lost 20% more traffic before recovering. Deletion destroys internal link equity. Consolidation with a 301 redirect preserves it.
---
Step Four: Fix Technical Signals That Compound Content Issues
Content quality and technical health are not separate problems. They're multiplied together in Google's assessment.
Core Web Vitals Still Matter
I know the "CWV is overrated" crowd is loud right now. Ignore them. I've seen enough A/B tests to know that a page loading in 1.8 seconds outranks the same page at 4.2 seconds when content quality is otherwise equal. Use PageSpeed Insights to identify your LCP bottlenecks. Nine times out of ten it's unoptimised images or a bloated third-party script (hello, chat widgets and cookie banners).
Internal Linking After a Drop
When pages lose rankings, they also lose PageRank flow from external links that used to land on them. You need to compensate with internal links. Audit every high-authority page on your site and find opportunities to link internally to the demoted pages. I use Ahrefs for this: filter by referring domains > 20 on the internal pages, then manually check whether they link to the recovering pages. Most don't. A couple of strong internal links from authoritative pages on the same domain moves rankings faster than most people expect.
Crawl Budget on Large Sites
If you're running 500+ pages, crawl budget matters. After an update, Googlebot sometimes reduces crawl frequency on sites it's demoted. Fix your crawl traps first: infinite scroll pagination, faceted navigation without proper parameter handling, duplicate content from URL parameters. These aren't glamorous fixes. They work.
---
Step Five: Build Your Recovery Timeline Honestly
This is where I have to be straight with people, especially clients paying monthly retainers hoping for a miracle.
Core update recoveries are slow. Not slow like "a few weeks". Slow like Google officially says sites may not recover until the next core update rolls out, which is typically every three to four months. I've seen well-executed recoveries take six to nine months for highly competitive niches.
What I tell clients:
- Expect early signals (small ranking improvements, CTR upticks) within 6-8 weeks of implementing fixes.
- Significant traffic recovery usually comes in waves aligned with Google's recrawl and reprocessing cycles.
- Do not revert fixes because you don't see results in week three. You're resetting the clock.
One of our SaaS clients at Seahawk got impatient at week five and switched SEO agencies mid-recovery. The new agency undid our content consolidation work, republishing the thin pages we'd merged. Traffic dropped another 15%. They came back four months later. We started over. Honestly, I wasn't even annoyed. I've seen it enough times that it just becomes a standard part of the conversation upfront now.
---
Step Six: Rebuild Authority Signals Post-Recovery
Once the on-site work is underway, you can't ignore off-site signals. Particularly link quality.
If your site took a big hit, audit your backlink profile in Ahrefs or SEMrush. Look specifically at link velocity changes around the update date. Did Google's update correlate with a batch of links being discounted? Patterns to flag: lots of links from domains with zero organic traffic, links from topically irrelevant sites, links acquired in an obvious burst during a specific month.
I don't recommend the disavow tool as a first resort. Google is fairly good at ignoring bad links these days. But if you've had a history of link schemes or bought links (don't lie to yourself), disavow the obvious offenders and document it.
The more productive investment is earning new, relevant links. Guest posts on sites that actually rank. Digital PR around original data. Partnerships. Nothing complicated. Just coverage that a real journalist would plausibly produce.
---
What Doesn't Work (and I've Tried Most of It)
Brief list, no commentary needed:
- Publishing 30 new blog posts in a month immediately after a drop
- Submitting sitemaps repeatedly hoping Google recrawls faster
- Changing your domain name or moving to HTTPS if you're already on HTTPS
- Hiding thin content with CSS instead of fixing or removing it
- Adding fake author bios with stock photos and no verifiable credentials
- Begging for manual review via Search Console feedback forms
The last one is particularly common. Google's manual review queue for algorithmic issues is not a customer service line. Submitting feedback doesn't expedite reconsideration for core update demotions, which are not manual penalties.
---
FAQ
How long does Google core update recovery actually take?
Realistically, three to six months for most sites making genuine improvements. Highly competitive niches can take longer. Google's John Mueller has said publicly that improvements may only be reflected after the next core update, which gives you a sense of the cycle time. Build your expectations around that, not around weekly ranking checks.
Should I revert site changes if rankings haven't improved after a month?
No. One month is not enough signal. Core update recoveries don't move in straight lines. I've had sites show no improvement for eight weeks and then recover 70% of lost traffic in a two-week burst. The pattern is lumpy, not linear.
Does publishing more content help after a core update?
Only if the new content is genuinely better than what's already ranking. Volume for the sake of volume actively works against you. Google's helpful content system (now folded into the core ranking system) specifically assesses whether a site seems to produce content primarily for search engines rather than people. More thin posts make this problem worse.
Can I recover from a core update without building new backlinks?
Yes. Most of the sites I've recovered didn't need new backlinks at all. The issue was almost always content quality, user intent alignment, or technical signals. Fix those first. External links help if your domain authority is genuinely lower than competitors, but they're not the first lever to pull.
What's the single biggest mistake site owners make during core update recovery?
Changing too many things at once. If you update content, fix technical issues, change your URL structure, and disavow links simultaneously, you can't identify what's working. Do it in phases. Prioritise content and intent alignment first, technical health second, off-site signals third.
---
The honest truth about core update recovery is that it mostly rewards patience and specificity. Google isn't targeting your site personally. It's recalibrating quality signals across billions of pages, and your site got caught in the rebalancing. If the content is genuinely useful, technically accessible, and credibly produced, it comes back. Not always to the same position, sometimes to a better one.
Do the work. Give it time. Don't delete things in a panic at 7am on a Friday.