Google Algorithm Recovery — Diagnose the Drop. Execute the Fix.

What Is Google Algorithm Recovery?

Google Algorithm Recovery is a structured diagnostic and remediation process for businesses that lost search traffic following a Google algorithm update. It is distinct from general SEO work in one critical way: the starting point is a traffic event that already happened, and every action taken must be traced back to a documented cause. Doing SEO work without first understanding what specifically caused the drop is not recovery — it's activity that may or may not move the needle, with no way to know which actions produced which results.

A recovery engagement begins with forensic analysis. We map the timing of your traffic decline against every confirmed Google update within the relevant window, identify which pages lost rankings and which held steady, analyze what changed in the competitive landscape around your affected queries, and evaluate your site's content, technical infrastructure, and authority signals against current quality benchmarks. The diagnostic phase produces a clear hypothesis: this is the update type, these are the affected signals, this is the remediation path.

What follows is execution — the actual work of improving the signals Google re-evaluated and found wanting. That work is different for every site and every update type, which is why the diagnostic comes first. There is no standard recovery checklist that applies to every situation. There is only the specific set of quality signals your site needs to improve, in the right order, at the right depth.

The Major Update Types That Cause Traffic Loss

Understanding what kind of update hit you matters because recovery looks different depending on the update type. Google runs several overlapping systems that can cause traffic loss, and they require different responses.

Core Updates

Broad core updates are the most common source of significant, site-wide traffic shifts. Google runs them multiple times per year, and their stated purpose is to re-evaluate which pages best satisfy user intent for a given query. A core update does not penalize pages for a specific violation — it changes how Google weights quality signals across the board, which means pages that were performing above what their quality signals justified will decline, while pages with stronger signals that were previously undervalued will rise.

Core update recovery is fundamentally about improving the quality of your content and the depth of your topical authority. If you lost ground in a core update, Google is telling you that the pages currently ranking ahead of you provide a better experience and more authoritative information for your target queries. The recovery path is improving your pages to a level that reflects genuine expertise — not cosmetic changes, not more words, but substantively better content that a knowledgeable person in that field would recognize as accurate and complete.

Helpful Content Updates

The Helpful Content system introduced a site-wide quality signal — meaning a high volume of low-quality, unhelpful content on your site can suppress rankings for your strong pages too. Unlike a core update that evaluates individual pages in context, an HCU assessment essentially asks: is this site primarily created for people, or is it primarily created to rank? Sites that had built large volumes of thin, search-optimized content saw significant suppression when the Helpful Content classifier became part of Google's core ranking infrastructure in 2023.

Recovering from a Helpful Content hit requires making hard decisions: identifying content that adds no genuine value, removing or consolidating it, and ensuring that what remains reflects real expertise and genuine purpose. This is not a technical fix. It requires editorial judgment and the willingness to reduce the size of a site in order to improve its overall quality signal.

Spam Updates

Spam updates target sites that have used manipulative tactics to inflate their search visibility — primarily scaled link building from low-quality sources, keyword stuffing, and content generated at high volume without genuine quality control. If a spam update caused your drop, the backlink profile review and potential disavow work becomes central to recovery. Google's spam detection has become significantly more sophisticated; tactics that worked three years ago are now reliably detected and penalized.

Manual Actions

A manual action is a specific, human-issued penalty from a Google Search Quality Reviewer. It is not an algorithmic assessment — it is a direct enforcement decision triggered by behavior that violates Google's spam policies. Manual actions show up in Google Search Console under Security & Manual Actions. They have a specific description, they apply to specific pages or the entire site, and they have a defined reconsideration request process. If you have a manual action, the path forward is explicit: fix the violation, submit a reconsideration request through Search Console, and wait for a human reviewer to evaluate whether the issue has been resolved.

Product Reviews Updates

Product Reviews Updates targeted sites publishing reviews that summarized existing information without providing firsthand experience, testing, or original analysis. Sites that relied heavily on affiliate-driven review content without demonstrating genuine product evaluation were downranked in favor of content that showed actual use: photos, testing methodology, comparative analysis from real experience. If your drop coincides with a Product Reviews Update, the recovery path focuses on demonstrating E-E-A-T — specifically the first E, experience — through your review content.

Why Most Businesses Handle This Wrong

Most businesses that experience a traffic drop following a Google update make the same set of mistakes. These aren't failures of effort — they're failures of method. Understanding them helps explain why recovery engagements so often stall.

Waiting It Out Without Diagnosing the Root Cause

The first instinct for many site owners is to wait for the next update to reverse what happened. This is occasionally correct — Google does sometimes overcorrect in a broad update and partially reverses it in subsequent updates. But relying on this without understanding what specifically changed is a passive strategy with no reliable outcome. If your traffic hasn't recovered through two or three subsequent core updates, the site has a real quality signal problem that is not going to self-resolve. Every month of inaction is a month of lost revenue and compounding competitive disadvantage.

Making Random Content Changes

The second common mistake is making changes to content without a clear hypothesis about what those changes are supposed to fix. Adding word count to pages that lost rankings. Rewriting meta descriptions. Publishing new blog posts to demonstrate "freshness." These activities create the feeling of progress while producing no meaningful signal improvement. Google's algorithms are not evaluating word count or publication cadence — they're evaluating the quality and authority of the content that exists. Busy work that doesn't move quality signals doesn't produce recovery.

Removing Pages That Weren't the Problem

A particularly damaging error is removing pages or content that had nothing to do with the quality issue. Businesses see that their traffic dropped and start deleting pages they perceive as thin or low-value — sometimes removing content that was ranking and contributing positively. Without page-level analysis of what actually declined, page removal decisions are guesswork that can accelerate the problem.

Confusing Correlation With Causation

Not every traffic drop is caused by an algorithm update. Technical issues, changes to how Google interprets a query, competitive market shifts, seasonality, and incorrect tracking implementation can all cause traffic declines that look like algorithm events in aggregate data. Running a recovery process for an algorithm hit when the actual issue is a canonical tag problem or a tracking discrepancy produces no improvement — because the diagnosis is wrong. Confirming the actual cause of the decline before prescribing a remedy is non-negotiable.

The Recovery Diagnostic Process

The diagnostic phase of a recovery engagement is where the real work happens. Everything that follows — the content decisions, the technical fixes, the content roadmap — depends on getting the diagnosis right. Here is what the diagnostic process covers.

Traffic Pattern Analysis: Pre/Post Update Mapping

We pull your organic traffic data from Google Search Console and Analytics and map it against every confirmed Google update in the relevant window — going back at least 18 months, often further. This establishes the correlation between specific update dates and your traffic shifts with precision: which updates correlated with drops, which updates you survived without impact, and whether the pattern is consistent with the update types we hypothesize as causal. This phase also confirms whether the traffic loss is real or a measurement artifact — tracking issues, domain changes, and Search Console configuration problems can all create false signals in the data.

Page-Level Quality Audit

Once we understand which updates affected you, we segment your site by performance change: pages that lost significant ranking positions versus pages that held steady or improved. The contrast between these two groups is where the signal lives. We analyze what the declining pages have in common — content depth, authority indicators, page experience signals, internal link equity — and compare them against your stable pages and against the pages that are currently ranking ahead of you for the same queries.

Content Gap Analysis vs. Ranking Competitors

For each priority query where you've lost position, we analyze the pages that displaced you. What do they cover that your pages don't? Where do they demonstrate expertise that your pages assume? What format decisions — structure, depth, specificity, original research — does Google appear to be favoring? This isn't about copying competitors; it's about understanding what quality signals the algorithm is now rewarding in your specific content category and building a roadmap to meet or exceed those signals.

E-E-A-T Assessment

Google's quality evaluator guidelines describe four signals that matter for page quality: Experience, Expertise, Authoritativeness, and Trustworthiness. These aren't abstract ideals — they have concrete expressions on a web page and a website. Experience means demonstrable firsthand knowledge. Expertise means the content is produced by someone with qualifications or demonstrated depth in the subject. Authoritativeness means the site is recognized as a credible source in its field. Trustworthiness means the site presents accurate information and operates with transparency. We evaluate each of these signals systematically and identify where your site's current presentation falls short of what Google's quality systems expect for your content category.

Backlink Profile Review

The backlink profile review has two purposes: identifying whether manipulative link patterns contributed to the drop, and assessing the overall authority picture relative to competitors who are now outranking you. If the link profile shows a high concentration of low-quality, irrelevant, or clearly artificial links, we flag this as a contributing factor and build a disavow strategy as part of the remediation plan. If the profile is clean but thin relative to competitors, link authority development becomes part of the longer-term recovery roadmap.

Crawl and Indexation Review

Technical issues often coexist with content quality issues in a traffic drop, and they can prevent recovery even when the content work is done correctly. We run a full crawl of your site to identify: pages with crawl errors, canonicalization problems that may be splitting link equity, indexation issues that are keeping good pages out of the index, Core Web Vitals failures that affect page experience signals, and internal linking structures that may be starving important pages of authority. Technical issues get remediated in parallel with content improvements — they are not blockers to starting the content work, but they must be resolved for the content work to produce its full effect.

What Recovery Actually Requires

The diagnostic tells you what to fix. What follows is the execution work — and it is worth being honest about what that work involves and how long it takes.

Timeline: Recovery Takes Multiple Update Cycles

Core update recovery does not happen immediately after you make improvements. Google re-evaluates sites on an ongoing basis, but significant quality signal changes are typically reflected in the next broad core update after the improvements are made — which may be three to six months away. In some cases, full recovery spans two update cycles, particularly for sites that required substantial content overhaul or that had a Helpful Content classifier impact. Anyone promising fast recovery from a core update is not being straight with you. The honest timeline is: do the work correctly over the next three to six months, and plan to measure the results in the following update window.

Content Rewrites vs. Page Removal Decisions

Not every page that lost traffic should be rewritten. Not every thin page should be removed. The decision framework depends on the page's traffic history, its current indexation status, its role in internal linking, and whether the query it was targeting is one where you have a legitimate path to competitive ranking. Pages with salvageable content that simply needs depth and authority improvement get rewritten. Pages targeting queries where you have no credible authority signal and cannot build one get removed or redirected. Pages that are thin but serve a structural purpose in the site get consolidated with related content to create a single authoritative page rather than several inadequate ones.

Building Positive Quality Signals

Recovery is not just about removing or fixing what's wrong — it's about actively building the positive signals Google is looking for. That means creating content with genuine depth and demonstrable expertise. It means ensuring that your site's authorship signals are clear and credible. It means building topical authority across a content cluster rather than optimizing individual pages in isolation. And it means investing in the trust signals — clear business information, transparent authorship, accurate and verifiable content — that Google's quality evaluators use as proxies for site-wide trustworthiness.

The Difference Between Algorithmic and Manual Recovery

These two recovery paths are often confused, but they are fundamentally different in structure and timeline.

A manual action is a direct enforcement decision. It is issued by a human reviewer, it appears explicitly in Google Search Console, and it has a defined resolution path: fix the specific policy violation described in the action, submit a reconsideration request, and wait for a reviewer to confirm that the issue has been resolved. Manual actions typically resolve within weeks of a successful reconsideration request. The clarity of the violation and the defined appeals process make manual action recovery more predictable — though it requires genuine remediation, not cosmetic changes. Google's reviewers have seen every possible version of incomplete fix-and-resubmit, and they do not grant reconsideration until the actual problem is resolved.

An algorithmic suppression has no appeals process, no explicit notification, and no defined timeline. You infer what happened from traffic data and update timing. You improve the signals you believe were evaluated and found lacking. And you wait for Google's systems to re-crawl, re-index, and re-evaluate your content against the current algorithm. This is why the diagnostic work is so important in algorithmic cases — without a clear hypothesis about which signals were the problem, the improvement work lacks direction and the recovery timeline becomes indefinite.

In practice, some sites have both: a manual action for a specific behavior and an underlying quality signal problem that the manual action partially masks. We check for manual actions in Search Console before anything else — they must be resolved before algorithmic work has its full effect.

Who This Service Is For

Google Algorithm Recovery is for businesses that experienced a meaningful, sustained drop in organic traffic that correlates with a known Google update date — and who have not recovered through subsequent updates without intervention.

The typical profile: a business that was generating real revenue from organic search, lost 30–80% of that traffic following one or more updates, has tried some content changes without measurable improvement, and is now looking for a structured diagnostic-first approach to understand what actually happened and build a credible path back. We also work with businesses that experienced a drop years ago and have never fully recovered — long-term suppression from an HCU or core update that left a site stuck at a fraction of its prior performance.

This service is not for sites that lost a small amount of traffic in normal ranking fluctuations, sites that are brand new and have not yet established authority, or sites looking for a quick fix in an industry where they have no legitimate authority signal to build on. Algorithm recovery requires that you have a real business with real expertise to demonstrate — we can help you demonstrate it more effectively, but we cannot manufacture it where it doesn't exist.

The Recovery Timeline: What to Expect Month by Month

The gap between what clients expect and what actually happens in a recovery engagement is almost always about timeline. The expectation — shaped by years of SEO vendor overpromising — is that fixes lead to visible traffic movement within weeks. The reality is that Google's quality signal re-evaluation cycle is measured in months, not weeks, and is tied to the release cadence of broad core updates. Understanding the realistic arc of a recovery keeps teams from declaring failure too early and from making impulsive changes mid-process that reset the signal clock.

Months 1–2: Diagnostic and Initial Remediation

The first two months of a recovery engagement are diagnostic-heavy. This is where we establish the precise cause of the drop, map the affected pages, identify the specific quality signals that changed, and begin building the remediation roadmap. Initial technical fixes — canonical issues, crawl errors, Core Web Vitals problems — are addressed immediately because they carry no dependency on update cycles. Content decisions are outlined but not yet fully executed; the diagnostic work has to be complete before you can make defensible decisions about which pages to rewrite, which to consolidate, and which to remove.

During this phase, expect no meaningful traffic movement. This is not a sign that the work isn't effective — it reflects the reality that Google has not yet had the opportunity to re-crawl, re-index, and re-evaluate the changes being made. Businesses that interpret this period of stability as failure and start making additional content changes without waiting to see results from the first round of improvements are among the most common cases of self-inflicted recovery delay we encounter.

Months 3–4: Content Quality Signals Begin to Register

By months three and four, the first round of substantive content improvements is live. Google's crawlers have visited the updated pages, and the quality signals have begun entering the index. At this stage you may see minor fluctuations — individual pages testing higher positions, some query categories showing modest recovery — but the overall traffic picture typically hasn't shifted materially. What you are watching for here is not traffic volume but crawl pattern changes: are the priority pages being crawled more frequently? Is Google returning to updated content within days rather than weeks? These signals indicate that the quality improvements are being registered even if they haven't yet translated to stable ranking changes.

In some cases, this phase also surfaces new diagnostic information. A page that was rewritten with significantly deeper content may test a position ten spots higher and then settle back — this tells you that the content signal improved but that other signals (authority, backlinks, internal equity) are still limiting the ceiling. That information refines the next phase of the roadmap.

Months 5–6: First Meaningful Traffic Movement

The five-to-six month window is typically where the first statistically meaningful traffic recovery appears, assuming the right signals have been addressed. This often coincides with the next broad core update in the window — Google's core updates act as forcing functions for large-scale quality signal re-evaluation. Pages that have been improved to a level that genuinely exceeds competitors for the relevant quality signals will move. Pages where the work addressed surface-level content changes without substantive signal improvement will not.

This is also the phase where the gap between "we improved some things" and "we improved the right things" becomes visible. If the diagnostic was correct and the remediation targeted the actual signals that caused the drop, you'll see recovery begin here. If the early-phase work was directionally correct but insufficiently deep — rewritten content that's better but still not at the level of what's now ranking ahead of you — this phase reveals that gap and informs what additional depth is needed.

Months 6–12+: Sustained Recovery Across Affected Pages

For sites that experienced significant, multi-page drops — particularly those affected by the Helpful Content system or multiple consecutive core updates — full recovery is typically a six-to-twelve month process. The pages that had the strongest pre-hit authority signals tend to recover first. Pages in content categories where competitors have significantly improved their own content during the same window require a higher level of remediation to displace what's now established. The long tail of affected pages — the middle-tier content that drove steady but not spectacular traffic before the drop — often takes until month eight or ten to recover meaningfully.

Subsequent core updates during this phase are important inflection points. A core update that arrives after five or six months of remediation work and produces further traffic decline is diagnostic: it suggests the current round of improvements has not yet moved the needle on the signals Google is evaluating. A core update that produces traffic recovery confirms that the quality improvements are working. Sites that see partial recovery in one update and full recovery in the next are following a normal pattern — Google's re-evaluation is not binary, and improvements accumulate across update cycles.

The Content Quality Signals Google Measures

The most actionable guidance Google has published for recovery from core updates is consistently some version of: improve your content quality. This is accurate but incomplete as a practitioner instruction. "Better content" means nothing without specificity about which content quality signals Google is actually measuring and how those signals are expressed on a page. What follows is how we operationalize content quality assessment in a recovery context — not at the level of editorial theory, but at the level of specific signals that can be audited and improved.

Original Research and First-Hand Experience

Google's emphasis on the first E in E-E-A-T — Experience — is a direct response to the proliferation of content that synthesizes existing information without adding anything new. Pages that aggregate what others have already written, even if they do it accurately and thoroughly, represent a lower quality signal than pages that contain information Google cannot find elsewhere. Original research — surveys, proprietary data, testing results, case studies with specific outcomes — creates a quality signal that cannot be replicated by competitors without doing the same work. First-hand experience expressed specifically — not "our clients have found" but concrete accounts of specific situations and outcomes — functions the same way.

In a recovery audit, we look at whether your existing content contains any information that is genuinely original — and whether that originality is expressed in a way that signals firsthand knowledge versus synthesis. The bar here isn't academic research. It's specificity that could only come from someone who has actually done the thing the content describes.

Depth of Topic Coverage and Topical Authority

Google's quality systems assess topic coverage at two levels: depth within a page (does this page comprehensively address the query and its related subtopics?) and breadth across a site (does this site demonstrate authoritative command of this subject area across its content cluster?). Shallow coverage — pages that address the main topic but don't engage with the nuances, edge cases, and related questions that a truly knowledgeable treatment would include — is consistently downranked relative to content with genuine depth.

Topical authority is the site-level version of this signal. A site with ten well-developed pages on a subject sends a stronger topical authority signal than a site with one excellent page surrounded by thin content on loosely related subjects. Recovery audits that reveal a pattern of shallow coverage across a topical cluster typically require a content architecture decision: consolidate thin pages into fewer, deeper treatments rather than attempting to improve ten inadequate pages individually.

Author Expertise Signals

Expertise signals are the visible indicators that the content was produced by someone with genuine knowledge in the relevant field. These include: clear author attribution with a byline that links to a credible author page; author pages that articulate actual credentials, experience, and track record rather than generic bios; content that reflects domain-specific knowledge in its framing, terminology, and treatment of nuance; and editorial decisions that a non-expert would not make — the choice to acknowledge complexity rather than oversimplify, the inclusion of caveats where they're warranted, the willingness to say "this depends on your situation" rather than providing a generic answer.

Sites that publish content without clear authorship, or with author attribution that links to empty or generic profile pages, are sending a negative expertise signal. In YMYL categories — health, finance, legal, and similar high-stakes subjects — the absence of credible author signals is a significant quality issue. Recovery work for these sites typically includes building out genuine author infrastructure: real author pages with verifiable credentials, consistent attribution across content, and enough editorial transparency that a quality evaluator could confirm the expertise behind the site's content.

Editorial Standards Signals

Factual accuracy, citation quality, and content recency are editorial signals that quality evaluators and Google's systems use as proxies for overall trustworthiness. Pages that contain factual errors — even minor ones that don't affect the core value of the content — undermine the trust signal of the entire page. Pages that cite authoritative sources and link to them signal that the content has been produced with editorial diligence. Pages that carry a last-updated date, and where that date reflects genuine content review rather than an automated timestamp, signal that the information is being actively maintained.

In a recovery audit, we look for factual accuracy issues as potential signals dragging down page quality, outdated information that no longer reflects the current state of the subject, and citation patterns that either strengthen or undermine credibility. This isn't editorial perfectionism — it's signal management. Every avoidable quality red flag is a reason for Google's systems to weight the page lower than it would otherwise be rated.

User Engagement Signals

Google has been deliberate about not confirming specific user behavior metrics as ranking factors, but the pattern of evidence from core update winners and losers consistently points to engagement quality as a factor in how content is evaluated. Pogo-sticking — users clicking a search result, immediately returning to the results page, and clicking a different result — is a signal that the content did not satisfy the query. Extended dwell time, scrolling engagement, and return visits are the inverse signal. We don't treat these as direct ranking factors in the technical sense, but we do treat them as indicators of whether the content is actually satisfying user intent — which is what Google's quality systems are trying to assess.

When auditing recovery candidates against competitors who are now outranking them, we look at content structure and presentation as proxies for engagement quality: does the page answer the actual question quickly and then provide depth, or does it bury the answer in preamble? Does the page structure match how users actually consume this type of content? Is the writing clear enough that a reader gets value without unnecessary effort? These aren't aesthetic questions — they're quality signals that correlate with the engagement patterns Google's systems appear to weight.

When Recovery Isn't Possible: Knowing When to Rebuild vs. Recover

The most uncomfortable conversation in a recovery engagement is the one where the data points toward a conclusion that recovery of the existing domain — while technically possible — will take longer and produce worse results than starting correctly from scratch. This conclusion is not common. Most sites with legitimate businesses behind them have a viable recovery path. But it exists, and treating every site as recoverable regardless of the evidence does the client a disservice. The rebuild-vs-recover decision should be made objectively, based on specific indicators, not on sunk cost psychology.

Domain-Wide Quality Signal Failure

A site where 80% or more of the indexed pages have meaningful quality signal problems — thin content, no genuine expertise signals, content that exists to rank rather than to serve — faces a different calculus than a site where 20–30% of pages need improvement. The Helpful Content system introduced a site-wide classifier that penalizes entire domains for high concentrations of unhelpful content, not just the individual pages that are low-quality. For sites in this category, the remediation math is brutal: removing or significantly improving 80% of a site's indexed content is often a multi-year project that still leaves you operating under an established negative quality signal. A new domain built correctly from day one starts without that signal weight.

Manual Actions That Were Not Properly Resolved

Manual actions have a defined reconsideration process, but the window to complete that process correctly matters. Sites that had a manual action issued, made partial fixes, submitted a reconsideration request before the work was genuinely complete, received a denial, and then let the action expire without resolving it have accumulated a trust deficit that is difficult to overcome algorithmically. Google's systems retain signals about sites' historical behavior, and a pattern of manual action followed by incomplete resolution followed by expiration is a signal pattern that tends to persist in how the domain is evaluated. If the business has evolved significantly since the manual action period — different content strategy, different link practices, different business model — a new domain that represents the current business accurately may be a cleaner path than attempting to rehabilitate one with this history.

Severe Toxic Link Profiles Without Clean Disavow Records

Backlink toxicity that accumulated over years of aggressive link building is addressable through disavow work, but the effectiveness of disavow has limits. Sites with thousands of toxic domains in their backlink profile, no historical disavow file, and no record of which links were intentionally built versus naturally acquired face a link cleanup project of significant scope. The disavow file is a tool Google provides and respects, but it is not a reset button. A domain with severe link toxicity that has been through multiple spam updates continues to carry the signal history even after disavow, particularly if the toxic link patterns are concentrated in the anchor text distribution rather than just the domain count. Evaluating whether the cleanup effort is worth undertaking requires an honest assessment of the domain's authority ceiling given its link history — and in some cases, that ceiling is low enough that a fresh start is the more efficient path.

Sites Built Primarily on AI-Generated Content

The 2023–2024 Helpful Content Updates were, in substantial part, a response to the explosion of AI-generated content that had been indexed at scale. Sites that were built primarily — not partially, but primarily — on AI-generated content and were flagged by the HCU classifier are in a category where algorithmic recovery is demonstrably difficult. The classifier doesn't evaluate individual pages; it evaluates the overall character of the site's content. A site where the majority of pages were generated without genuine human expertise, experience, or editorial judgment cannot be recovered by rewriting 20–30% of the content. The signal pattern of the site's history is established. If the business has a real story to tell and real expertise to demonstrate, that story is often better told on a domain that doesn't carry the classifier weight of the original site's content history.

Starting a Recovery Engagement

The first conversation is a diagnostic intake — not a sales pitch. We want to understand the timeline of your traffic drop, what updates you've correlated it with, what changes you've already made, and what data you have access to. If we can take on your recovery engagement and we're confident in a viable path forward, we'll tell you that. If the data suggests a different diagnosis — technical issue, tracking problem, competitive market shift — we'll tell you that too, even if it means a different scope of work than you came in expecting.

Traffic recovery from a Google algorithm hit is achievable for most sites with legitimate expertise and a real business behind them. The work is not fast, and it is not simple. But it is methodical, and when it's done correctly, the results are durable — not a temporary ranking spike, but a site that performs well through subsequent updates because its quality signals are genuinely strong. That's the standard we work to.