DEV Community: Shanzy Saleem

Google Just Listed 7 Ways to Keep Your Content Out of Search. Here Is When I Would Actually Use Each One

Shanzy Saleem — Sat, 24 Oct 2026 11:14:07 +0000

Google did not invent a new off switch on October 21. It finally put all the existing controls on one page. That clarity is useful, because most sites I audit use the wrong control for the job.

I spend a lot of my week inside Search Console and robots files for client sites, and one mistake keeps showing up. Someone wants a page out of AI Overviews, so they block it in robots.txt. Or they want a page gone from search entirely, so they add a nosnippet tag. Neither does what they think it does. So when Google updated its "Control what you share with Google Search" documentation on October 21, 2026, I read it properly, because it finally groups the options in one place instead of scattering them across five help pages.

Here is the honest summary first. Google says all of these options already existed. The update brings them together, adds rows for the robots meta tag, data-nosnippet and X-Robots-Tag, and for the site-wide Search generative AI control in Search Console, and clarifies that the page covers Google Search. It does not mention Google-Extended or AI training. That last point matters, and I will come back to it.

What actually changed on October 21

Three things are worth your attention.

First, the AI control is now visible on the main overview of content controls. That control lives in Search Console settings and rolled out to all sites globally on August 31, 2026. Turning it on removes a site from AI Overviews, AI Mode and the generative features in Discover. It is a site-wide switch, not a per-page setting.

Second, Google's documentation now explains how nosnippet and data-nosnippet apply to AI Overviews and AI Mode, alongside the robots meta tag documentation for AI Mode. In plain words, the snippet controls you already knew now have a stated AI meaning too.

Third, there is a small but useful reporting change from earlier in October. If a subdomain or subfolder property overrides the parent domain's AI setting, the top-level domain property owner now sees a notification on that page. Before that, a child property could quietly opt back in and nobody would notice. If you manage a site with separate properties for a blog, a shop and a help center, check this first.

The controls, in plain English

Think of these as different tools, not different strengths of the same tool.

Robots meta tag and X-Robots-Tag. This is how you tell Google not to index a page, or not to follow its links. The meta tag goes in the page head. X-Robots-Tag does the same job from the server header, which is handy for PDFs and other files that have no head section. Use this when the page should not be in search at all.
Robots.txt. This is a crawling control, not an indexing control. Blocking a page in robots.txt stops Googlebot from fetching it, but it does not guarantee the URL disappears from search. I still see blocked URLs show up as bare links. Use robots.txt to save crawl budget and keep bots out of admin areas and faceted filter URLs. Do not use it as your only way to remove a page.
Nosnippet and data-nosnippet. Nosnippet tells Google not to show a text snippet for a page. Data-nosnippet is the finer version. You wrap one part of a page in it and Google will not use that part in a snippet. This is useful when a page should rank but one paragraph, a price, or a phone number should not be quoted. It is not a removal tool.
Search Console removals. This is the emergency lever. When something sensitive is live and indexed, a removal request hides it fast while you fix the underlying page. It is temporary by design. If you never fix the page itself, it comes back. I treat removals as first aid, not treatment.
Login or password protection. If a page genuinely should never be public, put it behind a login. No tag beats a real barrier. Staging sites, internal docs and client previews belong behind a login, not behind a noindex tag you hope someone remembers.
The Search generative AI control. This is the new headline switch. One setting in Search Console, site-wide, removes the site from AI Overviews, AI Mode and Discover's generative features. Use it only when you have made a deliberate business decision about AI search, not as a panic reaction to one bad week of traffic. It is broad, and broad switches deserve a meeting, not a click at midnight.
Per-page judgment. This is the part Google cannot decide for you. Some pages should stay fully visible, some should rank without a snippet, and some should leave AI features while staying in classic search. The documentation gives you the levers. Your content strategy decides which lever each page gets.

My decision checklist before I touch any control

I run through this list before changing anything on a client site, and it has saved me from at least one embarrassing reversal.

What exactly do I want to stop: indexing, snippet display, AI features, or crawling? Name one. If you cannot name it, you are not ready to change a setting.
Is this one page, one section, or the whole site? A site-wide switch for a one-page problem is how good content disappears.
Is the page already indexed? If yes, a robots.txt block alone will not remove it. You need a noindex tag and a recrawl, or a removal request first.
Is there a child property that could override me? Check the Search Console AI control page for override notifications, especially on multi-property sites.
Who else depends on this page? A help article that feeds AI answers might also feed your support team. Removing it has a cost.
What will I measure after the change? Note the date, the setting, and the pages affected. Otherwise you will never know if the change helped or hurt.
Is there a login barrier where one belongs? If the content was never meant to be public, fix access first and tags second.

What I would not do

I would not treat the AI control as an AI training opt-out, because Google's updated page does not present it that way. It covers Google Search features, and it does not mention Google-Extended or AI training. If training is your concern, that is a separate conversation with separate controls, and you should not assume one switch covers both.

I would not block a whole site in robots.txt to escape AI Overviews either. You would lose classic search visibility for the same pages, and the block itself does not guarantee removal. And I would not add nosnippet everywhere "just in case." A page with no snippet is a page that is harder to click, and that cost is real even when the AI benefit is unclear.

FAQ

Does the Search generative AI control remove my site from Google Search entirely? No. It removes the site from AI Overviews, AI Mode and Discover's generative features. Classic search listings are a separate question.

If I block a page in robots.txt, is it gone from search? Not necessarily. Robots.txt controls crawling, not indexing. For removal, use a noindex tag and let Google recrawl, or use a Search Console removal request for urgent cases.

Do I need new schema or a new text file for AI search? Google's own guidance has said structured data is not required for AI Overviews or AI Mode, and you do not need to create new machine-readable files for them. Good structured data still helps classic rich results and clarity, but it is not a ticket into AI features.

Should every site opt out of AI features now? No. That is a business decision, not a default. Some sites will gain from AI citations and the more prominent links Google announced for AI Mode on October 15. Others, especially publishers whose full answer is the product, may reasonably opt out. Decide page by page where you can, and site-wide only when the case is genuinely site-wide.

The bottom line

Google did not hand us a magic off switch. It handed us a clearer map of the switches we already had, with the AI control finally sitting on the same page as the rest. The win is not in clicking faster. It is in naming what you actually want to stop, picking the one control that does that job, and writing down what you changed.

If you want a second pair of eyes on your setup, my portfolio is at https://www.shanzyseoexpert.com/ . I am Shanzy Saleem, an SEO specialist and web developer based in Lahore, and this is the kind of unglamorous control check I run on client sites before any bigger SEO work starts. Get the controls right first. The rankings conversation gets much easier after that.

Google Says Site Migrations Can Take 4+ Months. Here Is the Migration Checklist I Would Run First

Shanzy Saleem — Thu, 22 Oct 2026 11:16:05 +0000

Site migrations are one of those things every developer and SEO has to survive at least once. A rebrand, a move from WordPress to Next.js, a new domain, a URL structure cleanup. Whatever the reason, the pattern is always the same: you launch, traffic dips, and then everyone asks the same panicked question. How long until Google catches up?

This month, Google's John Mueller gave an honest answer. As reported in the October 16, 2026 PageTraffic weekly SEO roundup, he explained that website migrations can take longer than four months when Google runs into problems processing redirects, crawling URLs, or understanding changes across a large website. Migration timelines, he said, depend on site size, URL changes, and technical implementation.

That is the honest version. There is no magic "reindex everything" button. But four months is the problem scenario, not the guaranteed one. The difference between a two-week migration and a four-month migration usually comes down to how much you prepared before launch. Here is the checklist I would run.

Before you touch anything: the pre-migration audit

Crawl the old site and export everything. Use a crawler and save every URL, its status code, title, meta description, headings, and word count. This is your ground truth. If you lose traffic later, this file tells you exactly what changed.
Map every URL, one by one. The single biggest cause of long migrations is bad redirect mapping. Every old URL that had traffic, backlinks, or rankings needs a 1:1 301 redirect to its closest new equivalent. Not to the homepage. Not to a category page. To the real equivalent page.
Keep URL slugs clean. Google's Gary Illyes clarified this week (same roundup) that keywords in URLs have minimal impact on rankings, so there is no need to stuff them. Use hyphens instead of underscores to separate words, and if you are rebuilding URLs during the migration, now is the time to clean up the ugly ones. Just remember: every changed URL is another redirect Google has to process, so change only what is worth changing.
Document the old site's Search Console data. Export your top queries, top pages, and backlink profile. Post-launch, you will compare weekly. Without a baseline, you cannot tell "normal settling" apart from "something broke."

The technical setup that actually matters

301s, not 302s, and test them before launch. Set up redirects in a staging environment first and crawl the full old URL list against it. Look for chains (A to B to C), loops, and redirects to 404s. Every chain you leave in place is extra crawl budget burned and extra delay added.
One version of everything. Canonical tags, HTTPS, www vs non-www, trailing slashes. Pick one version and redirect the rest. During a migration is exactly when mixed signals confuse Google the most.
Update internal links to point at the new URLs. Redirects catch external links and bookmarks, but your own internal links should point directly at the new destinations. Crawlers follow internal links first, so make their job easy.
Keep the robots.txt and sitemaps honest. Do not accidentally block staging, and do not forget to unblock it. Submit the new XML sitemap in Search Console on launch day, and keep the old sitemap submitted for a while too, so Google can discover the redirects.

Launch week: watch like a hawk

Monitor indexing daily for the first two weeks. Check the Index Coverage report. A wave of excluded pages or redirect errors in week one is normal noise, but redirect errors that persist into week three are not.
Watch server logs for Googlebot behavior. If Googlebot is still hammering the old URLs weeks later, your redirects are not clean enough. If it is crawling the new URLs steadily, the processing is underway and patience is the right strategy.
Do not redesign everything at once. Mueller's point about Google "understanding changes across a large website" is important. If you change the domain, the URL structure, the design, and the content all in one go, you have given Google a puzzle instead of a map. Stagger big changes when you can.

The 4-month reality check

Here is the honest math. Small site, clean 1:1 redirects, no structural chaos: Google usually processes it in a few weeks. Large site with messy redirects, changed URL patterns, and crawling issues: this is where the 4+ month timeline Mueller described comes from. Your job before launch is to remove every item from the second list.

And one thing to internalize: a traffic dip after a migration is not automatically a penalty. Rankings need the new signals to be reprocessed, re-crawled, and re-evaluated. Panicking and rolling back mid-migration is how you turn a two-week dip into a six-month mess.

FAQ

How long does a site migration take for SEO? There is no fixed number. Google's John Mueller said in October 2026 that migrations can take longer than four months when Google hits problems with redirects, crawling, or understanding large-scale changes. Clean migrations on small sites are usually processed in weeks.

Should I change my URLs during a migration? Only if the new structure is clearly better. Every changed URL adds a redirect that Google must process, which adds time. And keep slugs clean: hyphens, not underscores, and do not stuff keywords into URLs; they have minimal ranking impact per Google's Gary Illyes.

What is the most common migration mistake? Weak redirect mapping. Sending everything to the homepage, leaving redirect chains, or forgetting URLs entirely. Crawl the old site first and map 1:1 301s before launch.

Do I need to keep the old domain? Keep it, and keep its redirects live, for at least a year. Dropping redirects too early is a classic way to lose the link equity you migrated.

I have handled the technical side of rebuilds like this on client projects, moving between WordPress, Shopify, and Next.js stacks. If you are planning a migration and want a second pair of eyes on the redirect map before launch, that is exactly what I do. You can find my work at https://www.shanzyseoexpert.com/ .

What is the scariest migration you have ever been through? I would genuinely like to hear the war stories.

A Forensic Study Claims Google Ran a Secret Core Update in August. Your Server Logs Can Check It

Shanzy Saleem — Thu, 22 Oct 2026 08:46:15 +0000

Every SEO I know watches the same two things when rankings move: Google's Search Status Dashboard and the rank trackers. This week a forensic study made both look like the wrong place to look.

A server-log analysis published on October 19 by VizzEx, a firm that studies raw crawl data, claims Google quietly ran a major core update in late summer 2026 without announcing it. Not a rumor from a forum. Their evidence is sitting in the one place most of us never check: our own server logs.

What the study actually found

VizzEx compared raw server access logs across six monitored domains. Between August 24 and September 8, 2026, Google's layout rendering effectively froze. In plain terms, the part of Googlebot that loads your page like a real browser, executing JavaScript and fetching your CSS and JS files, went quiet or dropped sharply for about 16 days.

That window sits exactly between the August 2026 spam update (Google confirmed it ran August 18 - 21, 2026; there was no August core update) and the September 2026 spam update, which finished rolling out on October 8.

Carolyn Holzman, the analyst behind the study, argues the most likely explanation is resource reallocation: Google pulled rendering capacity to run an index-wide core ranking update offline, then applied it. Google has not confirmed any of this. But unlike most update speculation, this claim is testable, because your server recorded the same thing theirs did.

Why a rendering freeze matters for rankings

When Google is not rendering, your page is judged on its raw HTML. Anything your JavaScript adds later, content, links, structured data injected client side, effectively does not exist during that evaluation.

So during those 16 days, two sites with identical content could rank differently based purely on how the content was delivered. A server-rendered page kept its full value. A page that shipped an empty shell and built itself in the browser was, to Google, an empty page.

If your traffic moved strangely in late August or early September with no announced update to blame, this is a candidate explanation.

How to check your own logs in 20 minutes

You do not need their tools. You need your raw access logs, which your host keeps (cPanel, your server, or a CDN like Cloudflare).

Pull logs for August 20 to September 12, 2026. You want the days before, during, and after the window.
Filter for Googlebot user agents. Separate the raw page fetches from the resource fetches (Googlebot requesting your CSS, JS, and images right after the HTML). Healthy crawling has a steady ratio of resource fetches to page fetches.
Compare week by week. A cliff in resource fetches while page fetches continue is the freeze pattern the study describes.
Cross-check Search Console. Did impressions or average position shift inside that window? Did your JavaScript-heavy pages move more than your static ones?
Look at your key pages as raw HTML (view-source or curl). Whatever is missing there is what Google judged during the freeze.

When I run technical audits, this raw-HTML test is the one that surprises site owners most. The rendered page they see in their browser is not always the page Google grades.

What I am changing in my own workflow

First, critical content and internal links belong in the initial HTML. Server-side rendering or static generation for anything that matters. Client-side enhancement is fine, client-side existence is not.

Second, log checks join my monthly routine. Rank trackers tell you that you moved. Logs tell you what Google actually did before you moved.

Third, I am done treating the official dashboard as the full calendar of ranking events. It lists what Google announces. This study is a reminder that the index also changes in the dark, and the footprint is measurable if you look at primary sources instead of weather reports.

The honest caveat

This is one firm's interpretation of crawl data from six domains, published as a paid press release via PressAdvantage and syndicated on press-release aggregator sites. As of October 22, 2026, no SEO trade publication has independently covered or verified it. Google has neither confirmed an unannounced core update nor commented on the rendering freeze. Treat the conclusion as a hypothesis.

But the method is the takeaway, and it is free. Your logs either show the freeze or they do not. Twenty minutes of checking beats another week of guessing why August behaved so strangely.

I write about technical SEO and AI search on my site, https://www.shanzyseoexpert.com/ , including the checklists I run when rankings move without an announcement.

FAQ

Did Google run an unannounced core update in 2026?
A forensic server-log study by VizzEx, published October 19, 2026, argues yes, pointing to a rendering freeze between August 24 and September 8. Google has not confirmed it.

How can I verify it on my own site?
Download your raw server logs for late August and early September 2026, filter Googlebot requests, and compare page fetches against resource fetches week by week. A sharp drop in resource fetching is the pattern described in the study.

Why does rendering matter for SEO?
If Google does not render your page, content and links added by JavaScript are invisible during evaluation. Server-rendered HTML is the safe baseline.

Forums Say a Google Update Is Rolling Out. Google's Dashboard Says No. Here Is What I Actually Do About It

Shanzy Saleem — Mon, 19 Oct 2026 10:58:20 +0000

Forums Say a Google Update Is Rolling Out. Google's Dashboard Says No. Here Is What I Actually Do About It

Every few months, my phone lights up the same way. Someone sends me a screenshot from a forum thread: "traffic down 40% since Tuesday," "Discover gone," "is an update rolling out?" And by dinner, three more people have asked me the same question.

So here is what happened on October 19, 2026, and exactly how I handle these moments without burning a weekend on guesswork.

What actually happened today

Search Engine Watch published a story this morning: site owners on WebmasterWorld are reporting collapsing Discover traffic, Google News showing stories days or even weeks old, and Search Console data running 52 hours to a full week behind. Some are openly predicting a core update is coming.

But Google's Search Status Dashboard says something different. As of today, the most recent listed incident is still the September 2026 spam update, which Google marked complete on October 8. Nothing dated October appears anywhere. The last core update on the dashboard is still the May 2026 core update. Search Console's data anomalies page is quiet too.

So we have a classic standoff: forum noise says something is moving, Google's dashboard says nothing is. Both of these can be true at the same time, and that is the part most people miss.

Why I don't panic (and why I don't dismiss it either)

Here is the uncomfortable fact I have learned after three years of doing SEO: the dashboard only records updates Google chooses to announce. Core, spam, and Discover updates, plus serving incidents. Google ships ranking changes constantly without posting them there. An empty October column means there is no announced update. It does not mean nothing moved.

But the other side matters just as much. The most specific complaints in this round are about Discover and Google News, and those are fed by systems separate from core web ranking, which is what a core update would actually target. Several members trace the timeline back to late August, a full month before the September spam update even started.

My stance is boring on purpose: treat it as noise until evidence says otherwise, and run the evidence checks before you touch a single template.

My 7-step "is this an update?" checklist

This is the exact order I work through, every time. It takes me under an hour and it has saved me from making dumb panicked changes more than once.

1. Check the status dashboard first. Bookmark status.search.google.com. If a core update is real, it shows up there with a start time. Until it does, you are dealing with a forum prediction, not a confirmed event. Five minutes, done.

2. Split Search, Discover, and News in Search Console. In the Performance report, filter each surface separately. A drop confined to Discover or News is a completely different problem from a drop in web search clicks. Today's complaints are heavily Discover and News. If your web search clicks are fine, you are not debugging a core update, you are debugging Discover.

3. Compare rankings to clicks. One publisher this week said their visibility index was "skyrocketing" while actual traffic stayed minimal. Another said rankings were unchanged but traffic fell, and blamed new AI Overviews placements. Rankings holding while clicks fall points to the results page changing around you, more AI Overviews, more ads, more local units, not your site losing relevance.

4. Check your rank tracker before you believe it. This one is underrated. Since September 20, Google has been serving rank trackers a share of completely incorrect results pages, peaking at 25% and running at 5 to 11% a day since October 1, according to data from Nozzle. They call them "poisoned" pages. If your tracking tool looks unusually volatile, some of that movement may simply not be real. Cross-check two tools before you react to anything.

5. Look for new AI Overviews on your queries. Clicks can fall with rankings fully intact when an AI Overview absorbs the click. Search your money queries in a clean browser and note what appears above your listing now. If you want a deeper playbook on winning citations inside AI answers, I keep my full process write-up on my site at https://www.shanzyseoexpert.com/, which is where I document how I run technical audits for my client sites.

6. Check server logs for bot traffic. Traffic that holds while conversions quietly die for four weeks is a pattern several members reported this month. Bot traffic distorting analytics can look exactly like a ranking loss or exactly like stability. Before you read meaning into your analytics totals, look at your server logs and your bot filtering.

7. Document, then wait. Do not redesign during volatility. If the dashboard is empty and your checks are inconclusive, write down what you observed with dates, and do nothing structural for at least two weeks. Changing templates during unconfirmed volatility is how you end up not knowing which of your own changes caused what.

A quick note on the prediction side

There is a real forecasting conversation running too. Dan Taylor's update prediction model, covered by Search Engine Journal, puts the chance of a Google update in the next 30 days at 96.8%, with late October to mid-November as the most likely window, and expects the next big one to behave like Panda, aimed at scaled, low-value AI content. Google has not confirmed any of this. It is a forecast, not a fact.

My take, as someone who audits real sites: whether or not an update lands in November, the sites most exposed are the same ones exposed in May. Thin, scaled, clearly AI-generated content with no original input. If you have been meaning to consolidate or improve those pages, this is a good week to start, not because a model predicted a date, but because the direction of travel has been obvious for months.

FAQ

Is there a Google update rolling out right now?
As of October 19, 2026, no. Google's dashboard lists nothing new since the September spam update, which completed October 8. What exists is unconfirmed forum activity and a forecast. A real core update will appear on the Search Status Dashboard with a start time.

My Discover traffic collapsed. Is that a core update?
Probably not directly. Discover runs on separate systems from core web ranking. Check your Discover performance separately in Search Console before assuming any update.

Search Console data is days behind. Is that an update?
Not necessarily. Reported lags range from 52 hours to about a week, and Google has not acknowledged a reporting delay for this period. Treat it as a reporting issue until it resolves.

Should I make changes to my site this week?
Only changes you already had evidence for. If your checks are inconclusive, document what you see and wait. The worst time to overhaul a site is during unconfirmed volatility.

The short version: forums say maybe, the dashboard says no, and my checklist says check the evidence before you panic. That approach has never once let me down. The sites I work on do not get lucky during updates; they just get prepared long before the rumors start.

Search Console Now Shows Your AI Overviews Impressions. Here Is What I'd Actually Do With That

Shanzy Saleem — Fri, 16 Oct 2026 10:58:08 +0000

I've spent three years living inside Search Console for clients across the US, UK and Pakistan. So when Google added a dedicated report for AI Overviews and AI Mode visibility, I opened it the way I'd open any new tool: excited, then suspicious, then honest about what it can and can't tell me.

The report is real. Google announced it on June 3, 2026, and it reached all websites worldwide on August 31, 2026. It shows how often your pages appeared inside generative AI experiences on Google Search, broken down by page, country, device and date. That's genuinely useful.

But it only shows impressions. No clicks. No queries. No position. Let me walk you through what that means in practice, and the exact checklist I'd run on any site I manage.

What the report actually gives you

For years, an appearance inside an AI Overview was folded into your regular Web Search totals. You could see impressions go up and clicks go down and never know which surface was responsible. Research cited in PPC Land's coverage documented a 61 percent reduction in organic click-through rate for queries where AI Overviews appeared, and an Ahrefs study put it at 58 percent for top-ranking pages. So this was a real blind spot.

The new generative AI performance report fixes one part of it. You can now see:

Impressions: how often your URLs appeared inside AI Overviews and AI Mode
Pages: which of your URLs were shown
Countries and devices: where and on what your visibility came from
Dates: trends over time, hourly to monthly

It does not tell you which search query triggered the appearance. AI Overviews and AI Mode are reported together, not separately. And clicks on those links get mixed into the regular Performance report's Web search type with no filter to separate them. Google has said it will introduce additional metrics over time, but for now, that's the deal.

Why the missing pieces matter more than the chart

Here is the part most coverage skips. Content decisions start from the question, not the URL. Knowing that your pricing page appeared in AI features 4,000 times is interesting. Knowing which buyer question it answered would be actionable. Google doesn't give you that.

John Mueller said it out loud in mid-September 2026: useful position tracking for these surfaces is hard, and the old position-1-to-10 thinking maps poorly to modern SERPs. He's right. A page can be cited inside an AI-generated answer and the concept of "ranking" barely applies.

So I treat this report as hygiene, not strategy. It answers "are we visible in AI features?" It does not answer "is this driving business?" Never confuse the two.

The checklist I'd run on any site

Here's what I'd actually do, in order:

Check whether you even have the report. If your site hasn't received enough AI-feature impressions, Google simply won't show it. No report is itself information: you're not appearing in AI Overviews yet.
Sort by page and find your most-cited content. These pages are already doing something right. Read them with fresh eyes and ask: what made this page easy for Google's AI to quote? Usually it's a direct answer near the top, clean definitions, or concrete examples.
Export both reports and compare CTR trends. Pull the generative AI impressions and the regular Performance report for the same date range. Pages with high AI impression shares and falling click-through are your citation pages: Google trusts them, but searchers get their answer without visiting. That's fine for brand visibility, but don't expect them to convert.
Rewrite your most important pages to be quotable. AI systems cite pages that give them clean, direct material. Put the direct answer in the first two sentences of key sections. Define terms plainly. Use real specifics instead of vague claims. Human editors should review anything AI-assisted before it ships, which Google explicitly recommends now.
Ask the question yourself. Since the report hides queries, I manually check how my key pages appear: I search the questions my customers actually ask and look at what the AI Overview cites. It's slow, but it's the only honest substitute for query data right now.
Don't opt out reflexively. Search Console has a control that opts your site out of AI-feature appearances without affecting your non-AI rankings. For most businesses, being named in the overview is the point. Opting out is a niche call, usually about licensing or paywalled content, not a growth move.

Where I think this is headed

DemandSphere reported that AI Overviews on its tracked branded queries jumped from about 26 percent in early September to over 80 percent by the end of the month. Google hasn't confirmed or explained that number, and the effect on clicks is still unknown. But the direction is obvious. More answers will be generated. The pages that get cited will be the ones written like they'd be quoted: direct, specific, human.

Google will probably add clicks and maybe queries to this report eventually. Until then, the operators who win are the ones who measure what they can, read the actual pages that get cited, and keep writing content worth quoting.

I write notes like this weekly on my portfolio at https://www.shanzyseoexpert.com/ if you want to follow along. No hype, just what the tools actually do.

FAQ

Are the Search Console AI reports available to everyone?
Yes. Google said on August 31, 2026 that the generative AI performance report and the AI feature control had rolled out to all websites worldwide, after an earlier limited test.

Why don't I see the report in my Search Console?
Google only shows it for properties with enough AI-feature impressions. If it's missing, your site likely isn't appearing in AI Overviews or AI Mode yet.

Does it show clicks or which queries triggered my appearances?
No. The report is impression-based only. No clicks, no click-through rate, no position, and no query dimension. Clicks on AI Overview links count inside the regular Performance report's Web search type without a filter.

Should I opt my site out of AI Overviews?
The Search Console control lets you opt out of AI-feature appearances without affecting your standard rankings. For most businesses, visibility inside AI answers is valuable, so treat opting out as a niche decision, not a default.

Google's September Spam Update Is Over. Here Is What I Would Check First

Shanzy Saleem — Wed, 14 Oct 2026 11:00:03 +0000

October has been a noisy month in search, and I have been reading the updates the way I always do: slowly, with my own Search Console open next to the news.

Here is the short version. Google started its September 2026 spam update on September 24, and it took nearly 14 days to finish, longer than the three earlier spam updates Google ran this year, according to Search Engine Journal. Then on October 1, Google updated its helpful content guidance with new sections around effort, originality, accuracy, and how content was created. And all of this is happening while AI Overviews keep showing up on more results pages.

None of that means you should panic and rewrite your whole site. It does mean this is a good week to check a few things properly, before you change anything.

First, what actually happened

Let me separate what is confirmed from what is only reported, because that distinction matters more than ever.

The spam update is confirmed. Google's own dashboard listed it as active, and it applied worldwide in all languages. Spam updates target sites that break Google's spam policies: thin auto-generated pages, cloaking, scaled low-value content, that kind of thing. If your site dropped during those two weeks, spam is one possible cause. It is not the only one. September also had plenty of normal ranking movement that never got a name.

The October 1 guidance update is also confirmed. Google refreshed two Search Central pages. One defines main content and talks about effort, originality, talent or skill, and accuracy, and it warns against fake author profiles. The other says AI-generated material should be checked by a person before publishing. Google is not saying you cannot use AI tools. It is saying a human needs to check the output before it goes live.

The AI Overview numbers come from third-party tracking, so treat them as directional, not gospel. AccuRanker announced a Dynamic CTR model on October 7. Its July data showed that on desktop, a result in organic position one earned about a 13.0% click-through rate with no AI Overview present. When an AI Overview appeared below that result, it fell to about 9.8%. When the AI Overview appeared above it, it fell to about 4.0%. That is roughly a 69% drop with no change in ranking at all. AccuRanker also reported AI Overviews on about 66% of its tracked desktop SERPs on September 30, up from 48% in July. That is their tracked dataset, not every search on Earth, but the direction is clear.

Two smaller October items are worth knowing. Google has introduced a UGC Fresh Data Program, reported by Search Engine Land, which lets approved platforms send fresh user-generated content signals directly to Google Search. And according to Search Engine Journal, Google quietly added UTM parameters to outgoing links so site owners can attribute referrals from Gemini more accurately.

The mistake I see after every update

The most common reaction is the worst one. People change titles, rewrite intros, delete pages, and disavow links, all in the same afternoon. Then when traffic moves again, they have no idea which change did what.

My rule is simple. Wait until a rollout ends, then compare your own data page by page and query by query before you touch anything. The September spam update is over, so now the checking can start.

Your post-update checklist

1. Confirm the dates in your own data. In Search Console, find the exact days traffic or impressions moved. If your drop started before September 24 or after the rollout ended, the spam update probably was not the cause.

2. Split the drop by page, not by site. Sort by the pages that lost the most clicks. Open the top losers and ask: would a real visitor call this page helpful? If not, that page is your work list.

3. Split by query too. Queries that lost clicks while impressions stayed flat often mean your ranking held but the result page changed around you, for example an AI Overview taking the top spot. The fix is different: clearer answers, better snippets, FAQ markup where it fits.

4. Read your top pages like a stranger. Does the page show real experience, or does it read like it was assembled from other articles? Are the facts right? Is the author real? Fake author profiles are called out in the guidance now, so fix that first if it applies.

5. Re-check any AI-assisted content. If a page used AI help, a person should have reviewed it before publishing. Look for confident claims with no source, repeated filler, and advice that sounds right but is wrong. Fix or remove what does not survive that read.

6. Look for spam-policy issues honestly. Scaled thin pages, doorway pages, hidden text, and scraped content are all in spam territory. If your site has any of that, cleaning it up matters more than any new content this month.

7. Check your AI visibility, not just your rankings. Search Console now offers generative AI reports to sites worldwide, according to recent reports. Check which of your pages show up there, then search your target queries manually and see whether your pages get cited.

8. Fix one thing at a time. Change the weakest pages first, note the date of each change, and wait. If you change ten things at once, you will never know what worked.

What I would NOT do this week

I would not rewrite content that is still performing, delete pages over a few days of dip, buy links to "recover" quickly, or publish a wave of new AI-written posts to replace traffic. All four have made things worse for sites I have seen. And I would not trust any single third-party volatility chart as proof of anything. Your Search Console is the only chart that matters for your decisions.

So what should you publish now?

The October guidance rewards the same things it always rewarded: content with real effort behind it, original points, accurate facts, and a real person standing behind it. That is hard to fake at scale, which is exactly the point.

If you want one practical move this month, write the page only you can write. The one with your own process, your own screenshots, your own mistakes and fixes. I write a lot about technical SEO on my portfolio at shanzyseoexpert.com, and the posts that hold up best are always the ones where I show the actual work instead of summarizing what everyone else already said.

FAQ

Did the September 2026 spam update target AI content?
Not specifically. Spam updates target spam policies, not a writing tool. But mass-produced thin content, whether written by a person or a machine, is exactly what spam updates tend to catch. The October 1 guidance adds the other half: AI-assisted content should be checked by a person before publishing.

My rankings look fine but clicks dropped. What happened?
That is the AI Overview effect in many cases. AccuRanker's July data showed position-one CTR falling from about 13.0% to about 4.0% on desktop when an AI Overview appeared above the result. Check whether your queries now trigger AI Overviews, and whether your pages are cited in them.

The bottom line

October gave us three messages at once: spam enforcement finished a long rollout, helpful-content guidance got sharper about effort and accuracy, and AI answers keep taking more space on the results page. None of that kills SEO. It just raises the price of lazy content. Check your own data, fix what is genuinely weak, and keep publishing work that shows real effort.

7 SEO Trends That Will Decide Rankings in 2026

Shanzy Saleem — Sat, 10 Oct 2026 10:23:37 +0000

I have been doing SEO for three years, and I have never seen the industry move this fast. AI answers, new ranking systems, shifting user behavior. If you feel like the old playbook is expiring, you are right. Here are the seven trends I believe will decide who ranks in 2026, based on what I am actually seeing in the data.

1. AI Overviews Are Eating Informational Clicks

Google's AI Overviews now answer a huge share of informational queries directly. The result: fewer clicks for "what is" and "how to" content, even when you rank number one.

What I am doing about it: I still create informational content, but I write it to be cited, not just clicked. Clear definitions, quotable statistics, structured steps. If the AI overview quotes you, that is the new position one. I covered the full playbook in my AEO and GEO guide.

2. GEO Is Becoming a Real Discipline

Generative Engine Optimization, optimizing to be cited by ChatGPT, Perplexity, Gemini, and Copilot, went from buzzword to budget line this year. My own analytics show AI referral traffic growing month over month.

The winners will be brands with clear entities, original data, and answer-first content. If you have not started tracking AI citations of your brand, start this week.

3. E-E-A-T Gets Enforced by AI

Here is the irony nobody expected: AI search made human expertise more valuable, not less. AI engines consistently prefer content with real author identities, real credentials, and first-hand experience. Faceless, generic content is getting filtered out of AI answers.

I am Shanzy Saleem, an SEO Specialist from Lahore, and everything I publish carries my name, my photo, and my real results. That used to be a branding choice. Now it is a ranking strategy.

4. Video and Visual Search Keep Growing

More product discovery happens through video and visual search every year. Google Lens queries keep climbing, and short video dominates attention.

Practical move: add original images and short videos to your key pages, optimize file names and alt text, and mark up videos with VideoObject schema. It is low competition because most SEOs still ignore it.

5. Topical Authority Beats Random Keywords

Google increasingly rewards sites that cover a topic deeply rather than chasing scattered keywords. I rebuilt my own site around five topic clusters with pillar pages, and the ranking improvements were dramatic. Random one-off articles cannot compete with a site that owns its niche.

If your content strategy is still "publish whatever keyword looks easy," 2026 will be rough. Build clusters.

6. Technical SEO Becomes Table Stakes Again

As AI crawlers join traditional crawlers in hitting your site, technical health matters more. Sites with crawl budget waste, JavaScript rendering issues, and slow interaction times are losing ground quietly.

Core Web Vitals, clean internal linking, proper schema, and crawlable architecture are not advanced tactics anymore. They are the entry fee. My Core Web Vitals fix guide walks through the exact process I use.

7. Zero-Click Is the Default, So Diversify Traffic

Between AI Overviews, featured snippets, and direct answers, a growing share of searches never produce a click. Smart SEOs are adapting by:

Building email lists and communities (traffic you own)
Creating content designed for citation, not just clicks
Diversifying into YouTube, newsletters, and forums where audiences actually hang out
Tracking brand mentions and AI citations, not just rankings

Rankings still matter, but they are no longer the whole game.

What I Am Doing Differently in 2026

Honestly? I am spending less time chasing individual keywords and more time building things that compound: topical clusters, original data, a recognizable personal brand, and content engineered to be cited by both Google and AI engines.

The fundamentals never changed: be the best answer, prove it with real experience, and make it easy to find and cite. The formats changed. The winners adapt.

FAQ

Is SEO dead because of AI search? No. SEO is changing shape, not dying. AI engines still need quality sources, and someone has to be the source they cite. That someone can be you.

Should beginners still learn traditional SEO? Absolutely. Keyword research, technical SEO, content strategy, and link building are the foundation everything new is built on. Skip the fundamentals and the advanced stuff will not work.

What is the single most important SEO skill for 2026? Adaptability with evidence. Test everything yourself, measure what happens, and follow the data instead of the gurus.

How do I future-proof my website against AI changes? Build a real brand with real expertise, publish original data, structure content for citation, diversify traffic sources, and keep your technical foundation clean.

Will backlinks still matter in 2026? Yes, but quality matters more than ever. A few genuine editorial links beat hundreds of low-quality ones. And internal links, which you fully control, remain criminally underused.

Final Thoughts

Every year someone declares SEO dead, and every year the adaptable SEOs do better than before. 2026 rewards the same things it always has: genuine expertise, great content, and technical excellence. The packaging is new. The principles are timeless.

I am Shanzy Saleem, and I document everything I learn about modern SEO in public. You can find more of my playbooks at shanzyseoexpert.com, including the case study behind growing my own site to 1M+ monthly Google impressions.

E-commerce SEO: Category Page Optimization Guide (2026)

Shanzy Saleem — Sat, 10 Oct 2026 10:16:22 +0000

Product pages get all the attention. But category pages are the real ranking powerhouses of e-commerce SEO. They target high-volume keywords, distribute link equity to products, and convert browsers into buyers. Here is my complete guide to optimizing them.

Why Category Pages Matter More Than You Think

Think about how people search. "Running shoes" gets far more searches than any single shoe model. That query lands on a category page, not a product page. Category pages rank for the broad, high-traffic commercial keywords that actually drive revenue.

I am Shanzy Saleem, an SEO Specialist and Web Developer from Lahore. I have worked on Shopify and WooCommerce stores, and category page optimization is consistently the highest-ROI work I do on e-commerce sites.

Step 1: Keyword Research for Category Pages

Category pages should target one primary keyword each, usually a broad commercial term:

Use tools like Ahrefs or Semrush to find category-level keywords ("men's leather wallets," "organic protein powder")
Check search intent: if Google shows category-style pages ranking, your category page can rank too
Avoid targeting the same keyword on both a category page and a blog post (keyword cannibalization)

Map every important category to exactly one primary keyword and two to three secondary keywords.

Step 2: Write Unique Category Descriptions That Rank

Most category pages have zero unique content, just a product grid. That is a missed opportunity. Add 150 to 300 words of genuinely useful descriptive content:

Describe what the category covers and who it is for
Naturally include the primary keyword in the first paragraph and H1
Answer common buyer questions (materials, sizing, use cases)
Place the text where it helps users, usually below the H1 or above the footer, not shoved where it hurts UX

On Shopify, I add this through the collection description field. On WordPress/WooCommerce, through the category description. Keep it helpful, not stuffed.

Step 3: Optimize the Technical Foundation

URL structure: Keep it clean and logical: /collections/running-shoes/ not /collections/c12345xyz/. Include the keyword, keep it short.

Title tags: Front-load the keyword: "Running Shoes for Men & Women | Free Shipping | BrandName". Keep under 60 characters.

Meta descriptions: Write for clicks, not just keywords. Include the keyword, a benefit, and a call to action: "Shop premium running shoes with free shipping and 30-day returns."

Heading structure: One H1 with the primary keyword. Product names as H2s or H3s. Never skip heading levels.

Pagination: Make sure Google can crawl deep into paginated product listings. Use proper pagination handling or a "view all" page with canonical.

Faceted navigation: Filters create duplicate URL nightmares (/shoes?color=red&size=10&sort=price). Canonicalize filtered URLs back to the main category, and noindex thin filter combinations.

Step 4: Internal Linking to Category Pages

Category pages should be among the most internally linked pages on your site:

Link from the main navigation (obviously, but many stores bury categories)
Add contextual links from relevant blog posts to category pages ("we covered the best options in our running shoes collection")
Cross-link related categories ("Customers also browse: Trail Running Shoes")
Breadcrumb navigation with schema markup on every page

I wrote a full guide on internal linking strategy on my Hashnode blog if you want the deeper methodology behind this.

Step 5: Add Schema Markup

Category pages deserve structured data:

CollectionPage or ItemList schema listing the products
BreadcrumbList schema for navigation
Product schema on individual product cards where feasible

This helps Google understand the page structure and can earn rich results. My developer guide on adding JSON-LD schema covers the implementation.

Step 6: Speed Up Category Pages

Category pages are heavy: dozens of product images, filters, scripts. Speed directly affects rankings and conversions:

Lazy load product images below the fold
Compress all thumbnails to WebP
Defer filter and sorting JavaScript
Limit products per page to a reasonable number (24 to 48)

Step 7: Build Links to Category Pages

Everyone builds links to the homepage. Smart stores build links to category pages:

Digital PR around category-level content ("the ultimate trail shoe buying guide" hosted on the category page)
Resource page link building targeting the category's topic
Supplier and brand partnership links pointing at relevant categories

Even a few quality links to a category page can move it several positions for a competitive keyword.

Measuring Category Page SEO

Track these per category page:

Rankings for the primary keyword (and secondary keywords)
Organic traffic and revenue attributed to the page
Click-through rate from search (rewrite titles/descriptions for low-CTR pages)
Crawl stats: are paginated and filtered URLs wasting crawl budget?

FAQ

Should category pages have a lot of text content? Enough to be useful and unique, usually 150 to 300 words. More is not always better. UX comes first; the text supports rankings, it does not replace good navigation and filters.

How do I handle out-of-stock products on category pages? Keep the category page live, push out-of-stock products to the bottom, and suggest alternatives. Never 404 a ranking category page because products sold out.

Category page vs product page: which should rank for the main keyword? The category page, for broad commercial terms. Product pages target specific model and long-tail keywords. If a product page outranks your category for the head term, strengthen the category's content and internal links.

How many products should a category page show? Enough to look complete but not so many it loads slowly. I recommend 24 to 48 per page with fast pagination. Infinite scroll needs careful SEO handling.

Do filters hurt category page SEO? Only when mismanaged. Canonicalize filtered URLs to the main category, noindex thin combinations, and make sure the main category URL is the one that accumulates links and rankings.

Final Thoughts

Category pages are where e-commerce SEO money is made. Optimize them systematically: one keyword each, unique helpful content, clean technicals, strong internal links, schema, speed, and a few good backlinks.

Do this for your top ten category pages and watch what happens to organic revenue.

I am Shanzy Saleem, and I share the exact e-commerce SEO processes I use on client stores. You can find more of my playbooks at shanzyseoexpert.com, including the case study behind growing my own site to 1M+ monthly Google impressions.

AEO & GEO: Optimizing for AI Search (ChatGPT, Gemini, Perplexity)

Shanzy Saleem — Fri, 09 Oct 2026 15:18:06 +0000

Search is changing faster than ever. People are not just typing keywords into Google anymore. They are asking full questions to ChatGPT, Gemini, and Perplexity, and getting direct answers without ever clicking a link.

As an SEO specialist, I had two choices: panic, or adapt. I chose adapt. Over the last year, I have been testing what makes AI engines cite a website as a source. This is my practical playbook for AEO (Answer Engine Optimization) and GEO (Generative Engine Optimization).

What Are AEO and GEO, Really?

AEO means optimizing your content so answer engines pick it as the direct answer to a question. Think featured snippets, but for AI.

GEO means optimizing so generative AI models mention and cite your brand when they generate responses. When someone asks Perplexity "what is the best way to do keyword research," GEO is what gets your site cited in the answer.

Traditional SEO gets you ranked. AEO and GEO get you quoted. In 2026, you need both.

Why I Started Taking This Seriously

I am Shanzy Saleem, an SEO Specialist from Lahore, and I grew my own website to over 1 million monthly Google impressions through classic SEO. That foundation still works. But I started noticing something: my analytics showed a new traffic source labeled as referrals from chatgpt.com, perplexity.ai, and Gemini.

It was small at first. Then it kept growing. So I dug in and started treating AI visibility as its own channel, with its own rules.

How AI Engines Choose What to Cite

After months of testing, here is what I have learned about how these systems pick sources:

1. They love direct, quotable answers. AI models extract clean, factual statements. If your content buries the answer under five paragraphs of fluff, you will not get cited. Put the answer first, then explain.

2. Structure matters more than ever. Clear headings, short paragraphs, lists, and tables are easy for models to parse and quote. This is not new advice, but for AI search it matters ten times more.

3. E-E-A-T is the filter. AI engines prefer sources that demonstrate experience, expertise, authoritativeness, and trust. Real author names, real credentials, and original experience win. Generic AI-written content does not get cited. The irony is beautiful.

4. They cite the fresh and the specific. Statistics, original research, case studies with real numbers, and recently updated content get picked up far more often. My own case study content with real traffic numbers gets cited way more than my generic guides.

My AEO and GEO Playbook

Here is the exact process I follow on client sites now:

Step 1: Write answer-first content

Every article should answer its target question in the first 2 to 3 sentences. Not the introduction. The answer. Then expand with depth. This single change increased my content's citation rate noticeably.

Step 2: Add FAQ and HowTo schema

Structured data is a direct signal to AI systems about what your content covers. I add FAQPage schema to every guide and HowTo schema to every tutorial. It takes ten minutes and the payoff is real. I wrote a full developer guide on adding JSON-LD schema with Liquid on my Hashnode blog if you want the technical walkthrough.

Step 3: Make your entity crystal clear

AI models understand entities, not just keywords. Your brand needs a consistent identity across the web: same name, same description, same facts everywhere. This is a big reason I have been building out my own profiles across the internet. When every profile says the same thing about who I am and what I do, AI engines trust the entity.

Step 4: Publish original data

Nobody cites content that says what everyone else says. My most-cited pages are the ones with original numbers: my traffic case study, my technical checklists based on real audits. If you have data, publish it.

Step 5: Be present where AI looks

AI engines heavily cite community content: Reddit threads, Quora answers, forum discussions. I answer SEO questions genuinely on Reddit, WordPress.org forums, and Quora every week. Not for links. Because that is literally where AI models go to learn what real practitioners think.

Step 6: Keep the technical foundation clean

AI crawlers need to access and parse your content. Fast pages, clean HTML, no JavaScript walls hiding your main content, proper heading hierarchy. All the technical SEO fundamentals I cover in my Shopify SEO checklist apply here too.

How to Measure AI Visibility

You cannot optimize what you do not measure. Here is what I track:

Referral traffic from AI domains in analytics (chatgpt.com, perplexity.ai, gemini.google.com, copilot.microsoft.com)
Brand mention monitoring: ask the AI tools directly about your niche and see if your brand appears
Citation tracking: search your target questions in Perplexity and check if your pages get cited
Traditional rankings still matter, because AI engines heavily weight top-ranking pages as sources

FAQ

Is GEO replacing traditional SEO?
No. AI engines still rely heavily on the traditional web index. Strong traditional SEO is the foundation that makes GEO work. Think of GEO as a layer on top, not a replacement.

Do I need to block AI crawlers or allow them?
If you want AI visibility, allow them. If your business model depends on keeping content exclusive, you might restrict them. For most businesses, being cited by AI is free brand exposure. I allow them.

How long does it take to see results from GEO?
In my testing, 4 to 8 weeks for new content to start appearing in AI answers, assuming the site already has some authority. Brand-new sites take longer, same as traditional SEO.

Does AI-generated content rank in AI search?
Rarely. AI engines seem to prefer human-written content with real experience and original data. Writing your content with AI and expecting AI to cite it is a losing loop.

What is the single highest-impact GEO tactic?
Answer-first content structure with FAQ schema. It is simple, fast to implement, and it works across every AI platform I have tested.

Final Thoughts

AI search is not the death of SEO. It is SEO with a new output format. The fundamentals have not changed: be the best answer, prove your expertise, make it easy to understand and cite.

The SEOs who win in the AI era will be the ones who adapted early. Start now. Pick your five most important pages, rewrite them answer-first, add FAQ schema, and publish one piece of original data this month.

I am Shanzy Saleem, and I document everything I learn about SEO in the AI era. You can find more of my playbooks on my portfolio at shanzyseoexpert.com, where I share the exact processes behind growing a site to 1M+ monthly impressions.

Technical SEO Checklist for Shopify Stores (2026)

Shanzy Saleem — Fri, 09 Oct 2026 10:39:54 +0000

The short answer: Shopify SEO in 2026 means working around the platform's fixed constraints. You cannot change the /products/ and /collections/ URL structure, but you control canonicals, schema markup, theme speed, images, and internal linking. This technical SEO checklist covers the eight areas I audit on every Shopify store, from duplicate variants to hreflang.

I have spent three years doing technical SEO for Shopify stores. The platform makes half the decisions for you, which is a blessing when it handles canonicals and sitemaps automatically, and a trap when its defaults quietly create duplicate content or slow down your theme. Here is the technical SEO checklist I run on every store audit, with the Liquid and JSON-LD snippets I use.

How Do I Handle Shopify's Forced URL Structure and Canonicalization?

Shopify forces URL prefixes you cannot change: /products/, /collections/, /pages/, and /blogs/. Accept it and optimize what you can control.

The real issue is that every product is reachable through multiple paths, like /products/blue-tshirt and /collections/summer/products/blue-tshirt. Shopify handles this with a canonical tag pointing to the /products/ version, which most themes include in theme.liquid:

<link rel="canonical" href="{{ canonical_url }}">

Confirm the canonical tag renders correctly on product, collection, and blog pages.
Keep product handles short and keyword rich. /products/organic-acacia-honey-480g beats /products/organic-raw-unfiltered-acacia-honey-480g-glass-jar.
Never link internally to the /collections/.../products/... versions. Always use the canonical /products/ URL.
If you migrated from another platform, map old URLs to the new Shopify structure with 301 redirects under Content > Menus > URL Redirects.

Why Do Product Variants Create Duplicate Content on Shopify?

Every variant combination generates a ?variant=123456789 URL. Google usually respects the canonical, but variant URLs still get indexed more often than expected, especially when theme code or apps link to them.

Click through variant pickers and watch the URL bar. If selecting a size or color rewrites the URL with ?variant=, make sure no internal navigation links to those URLs.
Do not submit variant URLs in your sitemap or link them from collection grids.
Exception: if a color variant genuinely deserves its own rankings (think "red evening dress"), consider splitting it into a separate product.

What Schema Markup Does a Shopify Store Actually Need?

This is where I find the most missed opportunities. Most themes ship with basic Product structured data, but it is often incomplete. I have audited stores displaying customer reviews through apps like Judge.me whose Product JSON-LD had no aggregateRating or review fields at all. Visible reviews without matching schema means no star ratings in search results.

The markup I implement:

Product with offers (price, currency, availability).
Review / aggregateRating wired to your review app's data.
BreadcrumbList for the Home > Collection > Product trail.
FAQPage wherever the theme renders FAQs.
Organization with logo and sameAs links.

Here is the JSON-LD pattern I use for products with reviews:

{
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Organic Acacia Honey 480g",
  "image": "https://cdn.shopify.com/.../honey-480g.jpg",
  "description": "Raw, unfiltered acacia honey in a 480g glass jar.",
  "sku": "HON-AC-480",
  "brand": { "@type": "Brand", "name": "Example Store" },
  "offers": {
    "@type": "Offer",
    "url": "https://example.com/products/organic-acacia-honey-480g",
    "priceCurrency": "USD",
    "price": "24.99",
    "availability": "https://schema.org/InStock"
  },
  "aggregateRating": {
    "@type": "AggregateRating",
    "ratingValue": "4.8",
    "reviewCount": "46"
  }
}

Validate with Google's Rich Results Test, and re-validate after every theme update.

How Do I Improve Core Web Vitals on a Shopify Theme?

If a Shopify store is slow, apps are the cause about eight times out of ten. Every app injects its own JavaScript and CSS, often render blocking, often on pages where it is not used. I have seen stores running 25 apps where 10 were installed, tried once, and forgotten.

Audit installed apps quarterly. Delete what you do not use, then check theme.liquid and snippets, because uninstalling does not always remove the code.
Defer non-critical scripts and lazy load below the fold images. Older custom themes often skip this.
Give your hero/LCP image fetchpriority="high" and a proper srcset.
Keep web fonts minimal. Two font families with limited weights is plenty.
Be realistic: aim for green Core Web Vitals on mobile product and collection pages. Chasing a perfect 100 while running reviews, upsell, and chat apps is not a good use of time.

How Should I Optimize Images for SEO on Shopify?

Shopify's CDN serves WebP and AVIF automatically when you request images through Liquid filters, so this part of the checklist is mostly about using the platform correctly.

{{ product.featured_image | image_url: width: 800 | image_tag: loading: 'lazy', alt: product.title }}

Use image_url with explicit widths and srcset instead of full resolution files.
Write descriptive alt text for every product image. Describe the image, do not stuff keywords.
Rename files before upload: organic-acacia-honey-480g.jpg instead of IMG_4829.jpg.
Compress originals before uploading. The CDN optimizes delivery, but it cannot fix a 6MB source file.

What Are Shopify's Robots.txt and Sitemap Limits?

Shopify auto-generates your sitemap at /sitemap.xml, split into sub-sitemaps per resource type. It updates itself as products change, and noindexed pages are excluded automatically. You cannot edit entries manually. Submit only /sitemap.xml in Search Console, not the individual sub-sitemaps.

For robots.txt, Shopify lets you customize it through a robots.txt.liquid template. The defaults sensibly block /admin, /cart, /checkout, and internal search URLs. The customization I use most is blocking filtered collection URLs, which create crawl bloat on stores with heavy faceted navigation:

{% for group in robots.default_groups %}
  {{- group.user_agent -}}
  {%- for rule in group.rules -%}
    {{- rule -}}
  {%- endfor -%}
  Disallow: /*?*filter.p.m*
  Disallow: /*?*sort_by*
{% endfor %}

Only add disallows you understand. A wrong pattern can accidentally block paginated collection pages you want indexed.

How Do I Set Up Hreflang for Multi-Region Shopify Stores?

Shopify Markets creates localized subfolders like /en-us or /de, and this is where hreflang becomes essential. Without it, Google may show the wrong regional version.

What I implement in theme.liquid:

{%- for locale in shop.published_locales -%}
<link rel="alternate" hreflang="{{ locale.iso_code }}"
  href="{{ canonical_url | replace: shop.locale, locale.iso_code }}">
{%- endfor -%}
<link rel="alternate" hreflang="x-default" href="{{ shop.url }}{{ request.path }}">

Every localized page must list all versions including itself, with reciprocal links.
Include an x-default pointing at your primary market.
Test with a hreflang validator after launch. The most common failure I see is hreflang tags pointing to 404s after a market was disabled.

How Should I Structure Internal Linking for Shopify Collections?

Collections are your category hub pages and your best shot at ranking for competitive terms, yet most stores treat them as thin product grids with no text and no inbound links beyond the main nav.

Add 150 to 300 words of useful descriptive text below the product grid on important collection pages, and link collections contextually from product descriptions and blog content.
Implement breadcrumbs in Liquid for crawl depth and BreadcrumbList eligibility.
Make "related products" sections real links, not JavaScript-only carousels that crawlers cannot follow.

FAQ

Can I change Shopify's forced URL structure?
No. The /products/, /collections/, /pages/, and /blogs/ prefixes are mandatory on Shopify. Instead, optimize product handles, keep URLs short, and keep canonicals and internal links consistent.

Do I need a paid SEO app for schema markup on Shopify?
Not necessarily. Most schema needs can be handled with JSON-LD hardcoded into your theme, which is faster than an app and fully under your control. Apps make sense for collecting and displaying reviews, but the schema itself belongs in the theme.

Why are some of my product pages not getting indexed?
The usual suspects are variant URL duplication, thin supplier descriptions, and crawl budget wasted on filtered collection URLs. Check Search Console's page indexing report, fix the pattern rather than individual URLs, and keep your sitemap limited to canonical, indexable pages.

How many Shopify apps are too many for SEO and speed?
There is no magic number, but every app adds JavaScript weight. I get suspicious above 10 to 12 apps. Audit quarterly and remove what is unused, including leftover code in your theme files.

Does Shopify handle canonical tags automatically?
Mostly yes. Shopify outputs canonical tags on products, collections, and articles by default. Your job is to verify the tag renders, never link to non-canonical versions internally, and watch for apps or theme code that override it.

Shopify gives you a fast, secure foundation, but the technical SEO details are yours to own. Work through this checklist and you will cover the issues I find on most audits. For a second pair of eyes on your store, I offer technical SEO audits and Shopify optimization over at my portfolio.

How I Took a US Website from Zero to 1M Monthly Google Impressions

Shanzy Saleem — Thu, 08 Oct 2026 10:14:10 +0000

Three years ago, I took on a US-based website in a competitive niche with essentially zero organic presence. No rankings, no traffic, no backlink profile to speak of. Today it gets over 1 million monthly impressions in Google Search Console. This is the full story of how that happened: what I did first, what actually moved the needle, and what I would do differently.

Starting point: an honest audit

Before touching anything, I ran a full technical SEO audit. The site had no sitemap, no schema markup, thin category pages with under 100 words of content, and a backlink profile made mostly of directory spam. Page speed was poor on mobile. The good news: the niche had clear search demand and the top competitors were beatable on content quality. I documented everything in a spreadsheet and prioritized by expected impact, not by what was easiest.

Phase 1: Fix the foundation (months 1-2)

Direct answer: the foundation phase of growing a website from zero means fixing technical SEO first: crawl errors, sitemaps, canonicals, schema markup, Core Web Vitals, and internal linking. These unglamorous fixes unlock everything that follows, and impressions typically start climbing within six weeks.

The first two months were pure technical SEO. I fixed crawl errors, built a proper sitemap structure, added canonical tags, implemented schema markup on every template, and rewrote the site's metadata. I also fixed the worst Core Web Vitals issues by compressing images, removing unused scripts, and enabling caching. Then I rebuilt the internal linking so every important page sat within three clicks of the homepage. Nothing glamorous, but impressions started climbing within six weeks. Technical fixes do not go viral, but they unlock everything that comes after. When I run technical SEO audits for clients now through shanzyseoexpert.com, this foundation phase is always step one.

Phase 2: Content that deserves to rank (months 3-6)

With the foundation fixed, I turned to content. I mapped the niche's keyword landscape and found dozens of informational queries where the ranking pages were thin, outdated, or written by non-experts. I published two in-depth articles per week, each one genuinely more useful than what ranked. I wrote detailed content briefs for every article covering the target keyword, the search intent, the questions to answer, and the unique angle. Every post got original images, FAQ sections with schema, real author bios with credentials, and internal links to the money pages. By month six, informational content was driving the first real traffic, and several articles had cracked the top five.

Phase 3: Link building that compounds (months 4-12)

Content without links stalls. I started link building in month four with digital PR around original data from the site's own niche, plus targeted outreach to resource pages. My rule was simple: ten personalized outreach emails beat a hundred templated ones. I kept anchor text natural and diversified, earned links only from sites with real traffic, and said no to paid link schemes that would have risked the whole project. I also fixed unlinked brand mentions and reclaimed broken backlinks pointing at 404 pages. Link building is slow and unglamorous, but each quality link raised the ceiling for every page on the site. Around month nine, the compounding kicked in: new articles started ranking within days instead of months.

Phase 4: Scale what works (months 9-18)

Once I knew which content types earned links and traffic, I doubled down on them and cut what did not perform. I expanded winning articles, built content hubs around the highest-converting topics, and refreshed older posts quarterly. I also started tracking revenue per page, not just traffic, and pruned or merged pages that attracted visits but never converted. This is the phase most people skip. They keep publishing instead of improving.

What the numbers looked like

Direct answer: the site went from zero to 10,000 monthly impressions by month 3, 100,000 by month 6, 500,000 by month 12, and over 1,000,000 by month 18. Growth was not linear. It stayed flat for months, then turned steep once technical fixes, content, and links started compounding.

Month 3: first 10,000 monthly impressions. Month 6: 100,000. Month 12: 500,000. Month 18: over 1,000,000 monthly impressions, with clicks growing proportionally. The growth curve was not linear. It was flat, flat, flat, then steep. That is how SEO works when the foundation is right: every improvement multiplies the ones before it.

The mindset that made it work

I treated SEO like a product: weekly metrics reviews, a backlog of experiments, and a rule that every change got documented. When traffic dipped after a core update in month eleven, I did not panic. I compared the losers against the winners, found the pattern (thin sections on two hubs), fixed them, and recovered within six weeks. Calm, data-driven iteration beats frantic tactics every time.

What I would do differently

Two things. First, I would start link building in month one, not month four. I underestimated how long links take to compound. Second, I would publish less and promote more in the early months. Ten articles with real distribution beat thirty articles that nobody sees.

The takeaway

There was no secret tactic and no single hack. It was a technical SEO audit, better content than the competition, consistent white-hat link building, and eighteen months of patience. If you are starting from zero, that is genuinely good news: the playbook is known, it just requires execution. And if you want help running the same playbook on your site, that is exactly what I do. You can find me at shanzyseoexpert.com.

FAQ

How long does it take to go from zero to 1M monthly impressions?

In this case, 18 months. The timeline depends on niche competition, publishing velocity, and link building consistency. Expect 3 to 6 months of nearly invisible progress before the curve turns. Most people quit during the flat phase, which is exactly why the opportunity exists.

What mattered more: content or backlinks?

Both, in sequence. Technical SEO first, then content to give the site something worth ranking, then backlinks to raise the ceiling. Content without links stalls, and links to a broken site waste money. The order matters as much as the effort.

How many articles did the site publish to reach 1M impressions?

Roughly 150 in-depth articles over 18 months, about two per week. But the top 20 percent of articles drove 80 percent of the traffic. Publishing less and promoting more would have been smarter in the early months.

Can this playbook work in 2026?

Yes. The fundamentals have not changed: fix technical issues, publish content that beats the top results, earn real backlinks, and iterate with data. AI-generated content has made genuine expertise more valuable, not less.

Shanzy Saleem is an SEO Specialist and Web Developer based in Lahore, Pakistan. She specializes in technical SEO, link building, and growing organic traffic for US and international brands. Website: shanzyseoexpert.com.

Technical SEO Checklist for Next.js Websites (2026)

Shanzy Saleem — Thu, 08 Oct 2026 10:13:40 +0000

Next.js is one of the best frameworks for SEO when it is configured correctly, and one of the easiest to get wrong when it is not. I have audited dozens of Next.js sites over the past few years, and the same issues show up again and again. This is the exact technical SEO checklist I run through on every Next.js project in 2026.

1. Choose the right rendering strategy for each page

Direct answer: for SEO, use static generation for pages that rarely change and server-side rendering for dynamic pages. Both send fully rendered HTML to Googlebot. Reserve client-side rendering for logged-in dashboards that do not need to rank.

Next.js gives you static generation, server-side rendering, and client-side rendering. For SEO the rule is simple: content that must rank should be rendered on the server or at build time. Use static generation for marketing pages and blog posts, server-side rendering for pages that change often, and keep client-side rendering for logged-in dashboards that do not need to rank. I have seen entire blogs invisible to Google because they rendered behind a loading spinner. Always verify what Google actually sees with the URL Inspection tool in Search Console.

2. Nail metadata with the Metadata API

Every page needs a unique title tag and meta description. In the App Router, generate them dynamically with the Metadata API from your CMS or database. Keep titles under 60 characters, descriptions between 150 and 160, and never let two pages share the same metadata. I also set canonical URLs explicitly on every page template, because filters and tracking parameters create duplicate URLs fast on dynamic sites.

3. Keep sitemaps and robots.txt self-updating

Generate your sitemap dynamically so new pages appear automatically, and split large sitemaps by content type. Your robots.txt should reference the sitemap and block routes that waste crawl budget: API routes, preview routes, admin paths, and internal search pages. I verify both files after every major deploy, because one bad deploy can silently break them.

4. Fix Core Web Vitals properly

Direct answer: the highest-impact Core Web Vitals fixes for Next.js are the next/image component for images, next/font for self-hosted fonts, and removing or deferring third-party scripts. These three changes fix most LCP, CLS, and INP failures. Target LCP under 2.5 seconds on real field data, not just lab scores.

Next.js makes performance easier but not automatic. The wins that matter: use the next/image component for image optimization, use next/font to self-host fonts and prevent layout shift, and be ruthless about client-side JavaScript. Every third-party script hurts Interaction to Next Paint. I audit the bundle on every release and lazy load everything below the fold.

5. Put structured data on templates, not pages

Add JSON-LD schema to your page templates so every new page ships with valid structured data automatically. Article schema for the blog template, Product schema for products, FAQPage for Q and A sections, BreadcrumbList everywhere. Then validate with Google's Rich Results Test. Broken or missing schema is one of the most common findings in the technical SEO audits I run. I perform these audits for site owners through shanzyseoexpert.com, and structured data problems appear in almost every report.

6. Keep URLs clean and redirects flat

Short, descriptive slugs with no dates or IDs. Pick one trailing slash convention and stick to it. Next.js makes redirects easy in the config file, so there is no excuse for chains of three or four redirects. I keep a redirect map for every migration and test the top URLs before launch day.

7. Control pagination and filters

Sites with filters, sorting, and pagination can generate thousands of near-duplicate URLs. Canonicalize to the main category page, noindex the filtered variants, and make sure new product pages get discovered quickly. I have seen crawl budget burned on 40,000 filter combinations while new pages waited weeks for their first crawl.

8. Build internal linking into components

Your component architecture should make internal linking automatic: related post blocks, breadcrumbs, category hubs, and contextual links inside content. I check link depth on every audit. Important pages should be reachable within three clicks of the homepage. Orphan pages do not rank no matter how good the content is.

9. Get internationalization right

If you serve multiple languages or regions, hreflang has to be reciprocal and point at live URLs. One x-default, one canonical per language version, and no hreflang tags pointing to pages that 404. I validate the full set with a crawler before trusting it, because hreflang mistakes silently split your ranking signals.

10. Monitor after every deploy

SEO is not a one-time task. Track indexation of new pages weekly, watch the Core Web Vitals report in Search Console, and run scheduled crawls that compare against your baseline. One bad deploy, one wrong robots.txt, one broken sitemap can undo months of work. This checklist only works if you keep running it.

Work through these ten items before you spend on link building. Technical SEO is the foundation that everything else stands on.

FAQ

What is the most important technical SEO factor for Next.js?

Rendering strategy. If Googlebot cannot see your content without executing JavaScript, nothing else matters. Use static generation or server-side rendering for every page that needs to rank, and verify with the URL Inspection tool.

How do I fix Core Web Vitals on a Next.js site?

Use next/image for all images, next/font to self-host fonts, and audit third-party scripts ruthlessly. Most Next.js sites I fix pass Core Web Vitals after these three changes. Measure with field data in Search Console, not just lab scores.

Does a Next.js site need a sitemap?

Yes. Generate it dynamically so new pages appear automatically, split it by content type for large sites, and reference it in robots.txt. I verify both files after every major deploy because a bad deploy can silently break them.

Can a Next.js site rank without server-side rendering?

It can, but it is harder. Google renders JavaScript, but rendering costs crawl budget and delays indexation. For competitive keywords, server-rendered pages consistently outperform client-rendered ones in my experience.

Shanzy Saleem is an SEO Specialist and Web Developer based in Lahore, Pakistan. She helps brands grow organic traffic with technical SEO, link building, and high-performance websites built on Next.js, WordPress, and Shopify. Learn more at shanzyseoexpert.com.