Shopify SEO: The Complete Guide for 2026 (Including AI Search)

Published: August 24, 2026

Last updated: August 24, 2026

24 min read

By: Graeme Whiles

Every Shopify store owner has done the same thing.

You install an SEO app, watch the score tick green, and wait for Google to notice you.

Nothing happens.

Organic search stays a rounding error.

Shopify says the platform handles around 80% of technical SEO for you, and that's broadly true.

However, stores don't rank on the 80% Shopify automates. They rank on the 20% it leaves you.

That 20% is search engine optimisation: what your online store sells, how it's structured, and whether search engines can follow either.

I'm not writing from theory.

My own store, Three Putt Golf, went from new domain to 668,000 impressions in eight months on £0 of ads. This year I rebuilt REHAUS after a botched migration.

In this guide, I'll show you where that 20% lives: structure, money pages, duplicates, content, and the AI search layer other guides skip.

Real numbers throughout, and an SEO strategy you can run.

Short on time? Here are the key takeaways

Shopify handles the plumbing, not the strategy: SSL, sitemaps and canonicals come free. Structure, content and intent mapping are entirely on you.
Collections rank, products convert: collection pages are where most of your non-brand traffic will land. Treat them as your money pages.
Duplicate content is built into the platform: tag pages, variant URLs and duplicated collection paths need deliberate handling, not an app.
Structured data now feeds two audiences: Google's rich results and the AI engines deciding whether to recommend your products.
AI search is already sending buyers: REHAUS ranks number one in ChatGPT for its main category query. That didn't happen by accident.
Content compounds: Three Putt Golf grew from 474 monthly impressions to 277,913 in six months on content alone.

What Shopify gets right (and the 20% it quietly gets wrong)

Let's give Shopify its due first.

Out of the box you get SSL, an auto-generated XML sitemap, canonical tags on every template, clean crawlable navigation and mobile-ready themes.

Google's ecommerce documentation lists the technical foundations a store needs, and Shopify ships most of them by default.

That's the 80%, and it's why "is Shopify bad for SEO" is the wrong question.

First, though, let me define what I mean by 'the missing 20%'.

It's everything the platform cannot decide for you: which collections exist, what they're called, what's written on them, how they link together, what your products are marked up with, and whether anything on your domain gives Google or ChatGPT a reason to surface you over the next store.

No app makes those decisions. The green score doesn't either.

There are also a few things Shopify actively gets wrong, and I'll cover each one in this guide: forced URL prefixes, auto-generated tag pages, duplicated product paths, and no native way to add certain files AI crawlers look for.

One honest caveat before you commit hours to this.

If you run a twelve-product store in a niche nobody searches, with a settled customer base from Instagram or a market stall, you don't need this full stack, and you shouldn't hire someone like me to build it.

SEO pays back where search demand exists. Check that first.

Still here? Good. Let's get into the work.

Keyword research for stores: money pages first

The mistake most store owners make?

They start with blog keywords, because that's what generic SEO guides teach.

Ten listicle ideas later, the collections that could actually rank are still running on default copy.

Reverse it. Map keywords to money pages first.

Work through your collections one by one and assign each its target keyword.

Three page types, three kinds of search intent. Get that mapping wrong and you'll point a product page at a category term it can never win.

If two collections chase the same term, merge them. If a term with real volume has no collection, create one.

Where do the relevant keywords come from? Three places, in order of usefulness.

The search queries already earning you impressions in Google Search Console, because that's demand Google has confirmed. Your competitors' collection navigation, because their structure is their keyword research, published. And Google's own autocomplete, which hands you keyword suggestions in the exact search terms buyers type.

Keyword research tools add search volume to that list. They don't decide it.

Then, and only then, plan content.

When I launched Three Putt Golf, the domain had zero authority, so I aimed the content at research demand: brand reviews, comparisons and alternatives pieces for the golf streetwear brands people were already searching, like Malbon and Manors.

Buyers researching "Malbon alternatives" met my store while deciding what to buy.

The result? From 474 impressions in September to 277,913 in February, an average tracked position moving from 7.9 to 4.4, and 6,795 clicks in eight months.

£0 on ads.

That approach has its own playbook, and I've written it up in my guide to comparison page SEO.

For now, the principle: match every page type to the demand type it's built for, and give your money pages first pick.

Site structure: collections rank, products convert

Think of your store like a supermarket.

Nobody walks in and asks a member of staff for one specific tin by barcode. They ask where the beans are.

Google behaves the same way: far more people search the category than the individual item.

Collections are your aisles, products are your shelves.

Structure the store so the aisles get found.

Keep everything within three clicks of home: homepage to collection to product. Deeper than that and both shoppers and crawlers lose the thread. Selling in more than one country adds a layer to this, covered in my guide to Shopify international SEO.
Make navigation earn its links: your header and footer menus are the strongest internal links on the site. Spend them on collections you want ranking, not on a "Catalog" catch-all.
Control your tag pages: Shopify auto-generates a URL for every product tag inside a collection. Left alone, they pile up as thin, near-duplicate pages competing with the collection itself. Pick the handful worth keeping, write unique copy for them, and keep the rest out of the index.
Link from blog to collection to product: every article should pass a reader (and link equity) towards the money page it supports. My internal linking strategy guide covers the full pattern.

However, structure isn't only about hierarchy. It's about not competing with yourself.

When two collections, a tag page and an old blog post all target the same term, Google picks one on your behalf, and it often picks wrong.

I've covered how to find and fix that in my guide to keyword cannibalisation, and on Shopify the usual culprits are tags and near-duplicate collections.

Here's what makes structure work worth doing on this platform specifically: it scales through templates.

At REHAUS, I didn't rewrite collection pages one at a time.

I rebuilt the collection template itself, with proper H1s, entity-rich intro copy and FAQ accordions, then relaunched more than 40 priority collections on it.

Fixing a page improves one page. Fixing a template improves 246 at once.

Collection clicks rose 26.4% the following month.

One rebuilt collection, /collections/tylko, now takes 465 clicks at position 6 for a term searched 3,300 times a month.

Templates are the closest thing Shopify SEO has to a cheat code. Use them.

I've broken the full collection playbook, template and numbers included, into its own guide to collection page SEO, and the whole-store architecture, navigation and breadcrumbs included, into my guide to Shopify site structure.

Product and collection pages: the templates that do the ranking

Now for the pages themselves.

On-page optimisation is where I see the most neglect in audits, and where the fastest wins live.

When I opened up REHAUS, 790 of 791 indexable pages had metadata issues.

335 collection pages had no H1 at all. 246 of them had a meta description that read, in full, "@ REHAUS".

Eight characters, doing the job of a shop window.

Here's the standard I hold every store template to:

One H1 that names the category: "Togo Sofas" beats "Shop the collection". If the template doesn't render an H1, fix the template before touching copy.
Page titles that front-load the term: category first, brand last. Your title tag is the headline of your entry in Google search results, so around 60 characters before truncation.
Meta descriptions written to be clicked: one specific claim, one reason to choose you, no filler. I hand-rewrote 800 of these for REHAUS because nothing automated writes a reason to click. Run yours through my free meta description checker before they go live. It flags truncation and weak click triggers in seconds, and the full at-scale system is in my Shopify meta descriptions guide.
Collection intro copy above the grid: 100 to 200 words that name the category, the brands in it and who it's for. This is the copy that tells Google (and every AI engine) what the page is about. Push longer FAQ-style detail into accordions below the grid.
Product descriptions written for the buyer: materials, dimensions, what it pairs with, why it's better than the obvious alternative. No arbitrary word counts. A 60-word description that answers the buying question outranks 300 words of padding written for a robot.

None of this is glamorous.

All of it scales, because on Shopify you fix it once at template level and every page inherits the fix.

You can see what that looks like across a full store in the REHAUS case study, including the before-and-after on the collection template. And for the product-level half of the system, my product page SEO playbook goes deeper on every element here.

Image SEO: the ranking surface stores forget

Ecommerce is a picture business, and Google Images is a shop window most stores leave dusty. It's also the one set of search results where a small store can outrank a big one on craft alone.

Four habits cover it:

Name files like a human: togo-sofa-terracotta.jpg beats IMG_4382.jpg. Do it before upload. Shopify won't rename for you.
Write image alt text that describes the product: one sentence, what's in the image, category term included where it's true. It serves shoppers using screen readers, Google Images, and the vision models AI engines increasingly run over product pages. .
Let Shopify's CDN do the formats: WebP conversion and responsive sizing happen automatically. Your job is the input: upload sharp images around 2048px and don't paste in 8MB originals.
Keep one product image canonical: the image you mark up in Product schema is the one Google Shopping and AI shopping surfaces will show. Make it the best one.

Ten minutes per template, then it runs itself.

The duplicate content problem Shopify ships with

Shopify creates duplicate content by design. Nobody tells you this at checkout.

The classic example: link to a product from a collection and Shopify serves it at /collections/collection-name/products/product-name. The same product also lives at /products/product-name.

Two URLs, one product.

The platform handles this with a canonical tag (a line of code telling Google which URL is the official version), and most themes get it right by default.

However, "most themes" and "by default" are doing heavy lifting in that sentence.

Apps, theme edits and custom sections break canonicals all the time, and nobody notices because the pages still load.

The other duplicates are yours to create, and store owners create them constantly:

Tag pages: every tag spawns a URL that's 95% identical to its parent collection.
Variant URLs: ?variant= parameters that get linked, shared and sometimes indexed.
Overlapping collections: "Sale", "New In" and "Bestsellers" all carrying the same products with the same default copy.

At REHAUS I consolidated 113 near-duplicate collection URLs that had accumulated over the years.

Not one of them was created maliciously. They just built up, the way they do on every store nobody's auditing.

The fix?

Open Search Console, check the Pages report for "Duplicate without user-selected canonical", and work the list: consolidate what overlaps, noindex what's thin, and check your money pages' canonical tags point at themselves.

Google's own canonicalisation documentation explains how it picks a winner when you don't.

Shopify creates the duplicates. Google picks the winner. You get no say unless you take it.

The full workflow, including how to read Search Console's duplicate reports without panicking, is in my guide to Shopify duplicate content.

Structured data: one markup, two audiences

Structured data (schema markup: code that labels your prices, stock and reviews so machines can read them) used to be a nice-to-have for rich snippets.

In 2026 it's load-bearing.

The same markup that puts your star ratings and prices into search engine results is what AI engines parse when deciding whether your product is worth recommending.

One markup, two audiences, and the second one is growing.

For a store, the priority list is short:

Product and Offer: name, price, availability, and review ratings on every product. Google's Product structured data reference lists the required fields. Price and availability are the ones AI shopping surfaces lean on.
BreadcrumbList: reinforces the collection-to-product hierarchy you built earlier.
FAQPage: on collections and guides where you've added genuine questions and answers.
Organization: who you are, where you are, and the same description everywhere. AI engines cross-reference this more than you'd expect.

One warning from the field: a single source of truth matters more than more markup.

REHAUS had Product schema coming from the theme and an app at the same time, with conflicting values. I removed the duplicate and kept one clean source.

Conflicting schema is worse than none, because machines don't know which version to trust.

If you're not hand-coding JSON-LD, my free schema markup generator builds valid Product, FAQ and Organization markup you can paste into your theme.

And if you want to understand what you're pasting, my full guide to structured data for SEO goes deeper on every type mentioned here. For the Shopify-specific install, single source and all, there's my guide to Shopify schema markup.

Technical SEO for Shopify: what actually needs checking

Let's be honest about Shopify SEO apps.

Some do a real job: structured data management, redirect handling, image compression. Most SEO tools worth using aren't apps at all, and the free ones from Google will tell you more about your web pages than any dashboard inside your admin.

However, no app fixes your structure, writes a reason to click, or gives an AI engine a reason to recommend you.

The store with 14 SEO apps and default collection copy loses to the store with none and a rebuilt template. Every time.

Apps also carry a cost nobody puts on the listing page.

Each one injects its own scripts, and script weight is the number one reason Shopify stores fail Core Web Vitals (Google's page experience metrics: loading, interactivity, visual stability).

So here's the quarterly ritual:

Audit the app list: for each app, name the job it does. No answer, no app. Delete half of them. (Yes, even some free ones. Free apps still load scripts.)
Remove leftover code: uninstalled apps often leave script tags in the theme. Search the theme code for the app's name and strip what remains.
Check the vitals: run your homepage, one collection and one product through PageSpeed Insights, on mobile devices first, because that's where most of your buyers are. Fix images and scripts first. They're 80% of the problem.
Hunt broken links: dead internal links and 404s accumulate every time a product is deleted or a collection renamed. Search Console's pages report and any crawler will list them. Fix or redirect, because broken links waste crawl budget and lose buyers mid-journey.
Leave the server alone: hosting, CDN and checkout speed are Shopify's job, and no app changes them. Spend your energy where you have control.
Protect the chain if you're moving domains: DNS, domain configuration, path-level redirects and Change of Address all have to hold. My guide to Shopify migration SEO has the autopsy of one that didn't.

At REHAUS, part of the cleanup was phasing out the SEO apps safely, checking what each one actually rendered before removing it, so rankings never wobbled.

Removing things is a strange way to improve SEO. On this platform it's usually the fastest.

For the full workflow, my technical SEO audit checklist covers all 47 points I check on every site, Shopify or not. The app-by-app verdicts live in my guide to the best Shopify SEO apps, and the performance method in Shopify speed optimisation.

Content: the channel that compounds

Ads stop the moment you stop paying.

Content is the only channel on your store that gets cheaper with age. Content marketing earned its bad name in ecommerce by producing blogs nobody asked for, which is a failure of purpose rather than of the channel.

It's tempting to treat the blog as a place for launch announcements.

The stores winning organic treat it as an engine with three moving parts.

A cluster plan, not a topic list: every article exists to support a money page, and they interlink deliberately. One pillar, many spokes, all pointing at collections. I've mapped the full approach in my content cluster strategy guide, the Shopify-specific engine is in my Shopify blog SEO guide, and this very article is the model in action.
A template, not a blank page: at REHAUS, all 28 articles follow one rigid review structure: same sections, same schema, same declarative summary up top. A rigid template sounds like the death of creativity. Instead, it's how sixteen articles shipped in a single month without quality slipping, and why AI engines can quote them cleanly.
A refresh cycle, not a publish-and-forget: content decays. Pages that ranked in March are quietly losing clicks by September, and nobody notices because the graph everyone watches is total traffic. My content decay detector finds the pages slipping before the loss shows up in revenue. Run it quarterly and refresh the winners first.

Does it work on a store with no authority? That was the entire Three Putt Golf experiment: a new domain, no backlinks to speak of, content-led from day one.

668,000 impressions and 6,795 clicks in eight months.

And on an established store, the same engine moved faster.

REHAUS editorial clicks rose 267% in a single month once the template articles started landing, with the Made.com review sitting at position 3.5 on 50,866 impressions.

The pattern holds at both ends.

Money pages give content somewhere to send buyers. Content gives money pages the authority to rank.

Neither works alone.

Links and authority: the honest version

Every Shopify SEO guide ends this section the same way: write guest posts, do the skyscraper technique, build links at scale.

I'll give you the honest version instead.

Three Putt Golf grew to 668,000 impressions with almost no deliberate link building.

The content earned its authority slowly, and for a store in a niche with soft competition, that's often enough.

Links accelerate what's already working. They don't rescue a store with default templates and no content worth citing.

When you do go after them, stores have angles most sites don't:

Your products are the pitch: journalists and niche publications write gift guides, roundups and "best of" lists year-round. A genuinely distinctive product plus a short, specific pitch beats any template email.
Your suppliers and brands already link out: stockist pages, "where to buy" lists, brand directories. If you stock a brand, ask to be listed. These are the easiest relevant links in ecommerce.
Your data is linkable: sales trends, search trends in your category, a survey of your customers. One useful number gets cited by the same publications that ignore product pitches.
Buy nothing: paid link packages are the fastest way to spend £500 turning a clean profile into a risky one. If a link costs money and isn't editorial, walk away.

Priority-wise, this comes after structure, money pages and content.

An hour spent on a collection template usually beats an hour of outreach at this size.

AI search: where your next customers are already shopping

Your customers stopped waiting for ten blue links.

They ask ChatGPT which brand to buy, they read an AI Overview instead of clicking, and they get a shortlist from Perplexity before they ever see your ad.

Adobe's data showed AI-referred retail traffic converting 31% better than other traffic over the 2025 holidays, and 42% better by March 2026.

Most Shopify SEO guides don't mention any of this.

It's the biggest gap on this SERP and the reason this section exists.

Before the acronyms pile up, let me define the term.

'Answer engine optimisation' (AEO) is the work of making your store the answer AI systems give: being cited, quoted and recommended when a buyer asks a question you should win.

For a Shopify store, that work is concrete:

Let the right crawlers in: the retrieval bots that fetch pages to answer live questions (OAI-SearchBot and ChatGPT-User, PerplexityBot, Claude-SearchBot) respect robots.txt, which Shopify lets you edit through robots.txt.liquid. Blocking them means opting out of the answer.
Rewrite the file Shopify generates for you: llms.txt (a plain-text index that points AI systems at your most important pages) is now auto-generated for every storefront, and the default describes Shopify's ecosystem more than your store. Override it by editing agents.md.liquid in your theme, not llms.txt.liquid, because Shopify mirrors one into the other. My guide to llms.txt on Shopify covers the how and the trap.
Say what you are, in one canonical sentence: pick the description of your store you want machines to repeat, then use it consistently: homepage, about page, Organization schema, everywhere. Declarative beats clever.
Keep schema to a single source: conflicting Product markup reads as an unreliable narrator. One clean source, as covered above.
Earn third-party mentions: AI engines weight what others say about you. Reviews, press and directory listings all feed the answer.
Feed the shopping surfaces too: your Merchant Center product feed is the structured twin of your schema, and my guide to Google Merchant Center for Shopify covers what it now reaches.

Does any of this move product?

Ask ChatGPT for the best second-hand designer furniture.

REHAUS comes back as the number one "Best Overall" pick, above Vinterior, Pamono, 1stDibs and eBay UK, and the answer quotes the site's own canonical description almost word for word: a specialist reseller that collects, authenticates, prepares and stores.

That sentence was written to be quoted.

The good news? Almost none of your competitors have done this work.

Check where your own store stands with my AEO readiness score, then read my guide to AI Overview optimisation for the citation playbook. And for the ecommerce-specific play, my guide to getting your products recommended by ChatGPT covers both AI surfaces in full.

Keep it in proportion: AI referrals are still a small share of total traffic today, and your search visibility now spans two systems rather than one.

They're also the fastest-growing share, and the buyers arrive pre-sold.

Being early here costs little. Being late costs the answer itself.

Measurement: prove it or stop

SEO earns its budget the same way any channel does: with numbers a finance person would accept.

Not organic traffic in aggregate. Not search engine rankings in isolation. Revenue you can trace back to a page.

Google Search Console out of the box shows you everything blended together, which mostly means brand traffic flattering the graph.

Segment it, and the truth appears.

Filter the Performance report by page path (/collections/, /products/, /blogs/) and by queries containing your brand name versus not.

Two tools, and they answer different questions. Search Console covers everything up to the click: impressions, position, which queries you appear for. Google Analytics covers everything after it: what those visitors did, and whether they bought.

Five numbers, once a month, and together they're your SEO performance:

Non-brand clicks: the growth number. Brand clicks were coming anyway.
Non-brand share: at REHAUS, this moved from 38.2% to 51.1% in a month. That crossover, more than any total, is what "the SEO is working" looks like.
Queries in the top 10: the truest read on your search rankings, and a leading indicator. Clicks follow position with a lag.
Money-page CTR: if a collection ranks top 5 and CTR sits under 1%, the metadata is failing, not the ranking.
Pages indexed vs pages that matter: growth in indexed junk (tags, variants) is a structure problem announcing itself.

Before you spend a penny or an hour, sanity-check what the traffic is worth.

My SEO ROI calculator turns your conversion rate and order value into a projected return, and it's deliberately conservative.

If the numbers don't clear your bar, don't do the work.

Results, not reports. If a metric doesn't change a decision, stop tracking it.

Final Thoughts

Shopify gives you a faster technical start than any platform I work on.

That start is worth nothing if you treat the green app score as the finish line.

The stores that win on Shopify aren't the ones with the most apps installed. They're the ones that treat the missing 20% as the actual job: structure first, money pages second, the content engine third, and the AI layer before their competitors get there.

Work in that order. Every step compounds the one before it.

If you'd rather run it as a tick-list, my 40-point Shopify SEO checklist is this entire guide in checkbox form.

And if you'd rather see exactly where your own store's 20% is leaking, that's my day job. I offer a free SEO audit covering structure, metadata, schema and content quality, with findings back within five business days.

Get in touch, and I'll take a look, or email me directly.

Author Bio

Graeme Whiles is an independent SEO and AEO consultant at GWContent. He has worked with enterprise and SaaS brands, including Originality.ai, Connecteam, 6sense, Practice Better, and Peppr, growing organic traffic and AI search visibility across some of the most competitive categories in B2B. He also built Three Putt Golf Clothing from a blank domain as a live proof of concept for his methodology.

Shopify SEO Frequently Asked Questions

Is Shopify good for SEO?

Yes, with a caveat. Shopify handles the technical foundations (SSL, sitemaps, canonicals) better than most platforms out of the box. Rankings are decided by what it can't automate: structure, page copy, content and authority. The platform won't hold you back. It won't carry you either.

Do Shopify SEO apps actually work?

For specific jobs, yes: structured data, redirects, image compression. No app fixes structure, writes persuasive metadata or builds authority, and every app adds script weight that slows the store. Audit the list quarterly and keep only the ones doing a nameable job.

How long does Shopify SEO take?

Expect first movement in 6 to 12 weeks and compounding results over 6 to 12 months. Three Putt Golf went from 474 monthly impressions to 277,913 in six months on a new domain. REHAUS impressions grew 5.9x in four months on an established one.

Can you add llms.txt to a Shopify store?

Yes. Since May 2026, Shopify auto-generates one at yourstore.com/llms.txt, alongside agents.md and llms-full.txt. The default is generic, so override it by adding an agents.md.liquid template to your theme, which Shopify mirrors into llms.txt, and write a curated index of your most important pages instead.

How long should product descriptions be?

As long as the buying decision needs, and no longer. A 60-word description answering the real question beats 300 words of padding. Cover materials, sizing, what it pairs with and why it beats the obvious alternative, then stop.