SEO for Australian online stores: ecommerce strategy, structure and product pages

Ecommerce SEO helps customers and search engines find the right category, product or buying guide in your store. Start with the pages that match valuable demand, and make the catalogue easy to discover. Keep product data accurate everywhere it appears, and measure what search contributes to sales. Let your store's actual problem set the first priority, rather than assuming categories always come first.

What ecommerce SEO covers for an Australian store

Ecommerce SEO is bigger than product page optimisation. It covers:

• how the catalogue is organised

• how search engines reach every product

• what each page tells a buyer

• how product data travels to Google

• how you count the sales that result

The ABS's final Retail Trade, Australia release covers June 2025. It put online retail sales at $4.7 billion for the month, up 13% on a year earlier.

All three come from the ABS's last monthly retail release, in Australian dollars. They describe the market, not your store's opportunity. And they say nothing about how your own customers shop. Check your own analytics for device mix and trading journeys before you plan around phones or desktops.

Our SEO for small business guide covers the foundations for any website, from keyword research to links. This one deals with what a catalogue changes.

Find the store's first priority

Start with the symptom, not a standard checklist.

Then match each kind of search to the page that should answer it. Customers search for hiking boots, wireless over ear headphones or electric standing desks. They don't search for the stock codes your warehouse uses. So name categories in their words.

Categories are important targets for broad searches. But they don't always earn most of a store's revenue. Nor is a plain product grid meaningless to Google. That's because product names, links and attributes tell Google a good deal about what a category holds.

Let the catalogue's economics set the order of work:

• where the demand sits

• which lines carry margin

• which stock you can rely on

• where buyers need the most explaining before they commit

Supporting content earns its place when it answers a real buying or ownership question. Eight functions cover most of what a store needs.

Link to a product or category where it helps the reader take the next step. Not every paragraph needs one. Our guide to blogging for SEO covers when a separate blog is worth running at all.

Build categories and internal connections

Build the hierarchy around how customers choose, from broad ranges down to narrower ones. A subcategory earns its place when people search for it as a distinct thing. You also need to stock a useful selection worth comparing. And someone has to own keeping it accurate as the range changes.

There's no magic product count. Six well chosen boots for wide feet can make a useful page. But forty near identical listings under an arbitrary label won't.

In Help Google understand your ecommerce website structure, Google recommends links from menus to category pages. It recommends links from category pages to subcategory pages too. Google says the more links a page has within a site, the higher its relative importance. It also notes that Googlebot generally doesn't submit searches into a site's search box. So a product that only internal search can reach may never be found.

Here's that kind of hierarchy for an example outdoor store.

The tree shows parent and child pages only. The links that make it work also run across it. The hiking boots buying guide links to the hiking boots category. The Ridge boot links to comparable boots and the socks people buy with them. And breadcrumbs lead back up from every product.

Our guide to site architecture for small businesses covers hierarchy and internal linking in more depth.

Give each category a short introduction above the grid that helps people choose. Cover who the range suits, the differences that matter and where to start. Keep it brief so the products stay easy to reach. Compact buying guidance and filters that narrow the range beat a long block of generic copy. The common mistake runs the other way: 600 words for search engines between the shopper and the boots.

Link every product you want indexed from at least one relevant category. Where products can't be linked, Google's structure guide suggests a sitemap or Merchant Center feed as backup.

Make product pages useful

A product page has to answer the questions someone asks just before they buy. The answers have to match the exact variant they've chosen. For each product, that means:

• The product and the selected variant, named clearly, with useful images and alternative text that describes them.

• The dimensions, materials, compatibility or fit details that decide the purchase.

• An honest account of what it suits and what it doesn't.

• The current price and availability for the selected variant.

• A realistic delivery estimate, the delivery cost and the returns terms.

• A buying action that works on a phone as well as a desktop.

Here's how that could sit on the page for the Ridge boot from the same outdoor store.

It's a layout, not a finished design or evidence the boot is any good. On the real page, the shopper must see the selected variant's price and availability. Having them in the data isn't enough.

Write original buying information first for the products that matter most:

• strong demand

• a healthy margin

• stock you can rely on

• real uncertainty for the buyer

• a steady stream of support questions

Keep accurate manufacturer specifications as they are. Rewriting every sentence just to look different isn't the goal. Templates and AI tools can speed the work up. But every claim needs a factual check. Never publish testing you didn't do or customer quotes nobody gave.

Genuine product reviews answer questions your copy can't. Ask for them honestly, with prompts about fit, intended use and limitations, and with no reward attached.

The ACCC's Online reviews for product and services guidance makes two points. It's against the law for a business to create fake or misleading reviews. A business may also mislead consumers if it suppresses or edits negative reviews, or removes genuine ones. Any incentive must apply whether the review is positive or negative, and be clearly disclosed.

Review platforms can add conditions of their own. So check theirs before you offer anything. For reviews on Google itself, our guide to getting more Google reviews legally covers the rules.

Then test the page on a phone. Variant choice, image galleries, filter controls and checkout all behave differently on a small screen. They need to work there.

Content can sit in an accordion or behind a read more control. That's fine if it's in the mobile page and people can open it. The aim is equivalent information and working rendering, not a blanket ban on collapsed content. Our guide to mobile SEO covers mobile first indexing and what Google sees.

Choose a variant architecture

Some stores sell one product in many options on a single page. Others give important variants pages of their own. Google documents both designs. Canonical and structured data decisions must fit your chosen design and the commercial intent behind it.

Google's Designing a URL structure for ecommerce websites says each variant should have its own URL. A path or a query parameter can identify it. Where optional query parameters identify variants, Google recommends the URL without the parameter as the canonical.

The markup describes the relationships. In Google's Product variant structured data documentation, ProductGroup ties variants to their parent product. It does this with variesBy, hasVariant and productGroupID. The documentation covers single page designs, where every variant is selectable without a page reload. It also covers multi page designs, where variants live on different pages.

Each variant needs a unique ID such as its sku or gtin. Done properly, the markup makes products eligible to show variant information in Google's merchant listing experiences. Have your developer map the store's actual product model before writing any of it.

Take a boot in five colours and four sizes: 20 purchasable variants. None of them automatically needs its own indexed page. Searches for the green boot itself, plus its own images and stock, may justify a colour page. Like filter pages, variant pages need distinct demand, something useful to show and someone to maintain them. Either way, the selected state has to survive the click from a listing.

Run URL Inspection on a sample of product and variant URLs. You'll see what Google does with your variants: the canonical you declared and the one Google selected. The live test shows how the page renders now. Our Google Search Console guide walks through the reports.

A site: search is a rough discovery aid at best. Google's page on the site: search operator says its results aren't always exhaustive. It points to URL Inspection instead.

Control filters and pagination

Filters help shoppers narrow a range. They can also generate more URLs than any store could maintain. Google's faceted navigation guidance warns that filters can create infinite URL spaces. That means overcrawling and slower discovery of the pages you care about.

So make two separate decisions. The first is which filtered selections deserve durable landing pages. The second is how the rest of the filter space should behave.

For the first decision, test each candidate on its merits. Does it answer distinct demand? Is there a useful, reasonably stable range to show? Can the page say something useful about choosing within it? Will someone maintain it?

Brand, colour, size and price can each qualify in the right market. Sort orders and arbitrary combinations of several filters rarely carry separate search value.

An approved selection gets a stable URL, useful content and a canonical that points to itself. Its category links to it too. Everything else stays a filter for shoppers. Give those filters a deliberate crawl design, not layers of conflicting instructions. Keep crawling, indexing and consolidation separate, because each control does one job.

A canonical is a signal, not an instruction. Google can't see a noindex on a page blocked in robots.txt, because it can't crawl the page. Google's page on blocking indexing with noindex adds that a blocked page can still appear in search results. And Google's guide on how to specify a canonical URL rules out robots.txt for choosing canonicals. That's why stacking all three controls on the same URLs without a plan backfires.

Google's faceted navigation guidance suggests preventing crawling with robots.txt for filter URLs you don't need indexed. Where filter URLs do get crawled, it asks for the industry standard ampersand between parameters. Filters should appear in a consistent order, with no duplicates. A combination that returns no results should get an HTTP 404 status at that URL. It shouldn't redirect to a generic error page.

The guidance also says rel="canonical" may, over time, reduce crawling of the duplicate versions. Check the current state with your developer before changing anything. Suppose filter URLs are already indexed and you want them out. Blocking them in robots.txt first means Google never sees the noindex you add.

Keep an empty filter combination separate from a strategic category that's temporarily empty. An empty combination can return a 404. A seasonal category you'll restock next month needs its own usefulness and stock decision. Don't sweep it up in a blanket deletion rule.

Pagination needs the same care. Google's guidance on pagination and incremental page loading says its crawlers don't click buttons. They generally don't trigger JavaScript functions either. So set pagination up like this:

• give each page in a sequence its own URL

• link the pages with ordinary <a href> links

• give each page its own canonical rather than pointing them all at page one

• put a crawlable paginated route behind any load more button or infinite scroll

Without that route, products deep in the range may never be discovered.

Sitemaps back this up. Google's guide to build and submit a sitemap says to include the URLs you want in search results. It also says submitting one is merely a hint.

So list the canonical URLs of the products and categories you want indexed. Keep redirected and noindex URLs out. And don't drop a product just because it's temporarily out of stock.

Handle stock changes and product retirement

In most stores, stock changes every week. So these decisions come up far more often than a redesign does. Each situation needs its own answer.

A page with a genuinely relevant replacement gets a permanent redirect to it. A page with no equivalent can return 404 or 410. Redirecting many old pages to one irrelevant destination might be treated as a soft 404. Sending them all to the home page is one example. Google's site move guidance explains this.

A broad category that doesn't meet the same need is no better for the shopper. Someone looking for a particular retired model gets nothing from it. And no redirect guarantees the rankings come with it. A close replacement, or one page that genuinely consolidates several, is a different case.

Keep Merchant Center in step. Google's help on the availability attribute asks for consistent availability across:

• the landing page

• the structured data

• the checkout

• your product data

The same page says not to delete an item when you temporarily stop taking orders. That's because an item added back can take a long time to show again. Set it to out of stock instead. Or use the pause attribute to stop showing it for a while.

A recall isn't an SEO decision. Product Safety Australia's page on the role of the ACCC and government in recalls gives the deadline. The Australian Consumer Law requires suppliers to notify the Commonwealth Minister within 2 days of taking recall action.

Keep structured data and Merchant Center aligned

Products can appear in Google in four different ways. It helps to keep them apart.

None of them is guaranteed. Valid markup and an approved product make you eligible. Google still decides what shows.

Start with the markup. Google's merchant listing structured data documentation requires a product name, an image and an offers entry. That entry needs a price and its priceCurrency. Availability, shipping details and return policies are recommended rather than required. Google may also try to verify the product data before showing it.

Store wide shipping and return policies can be set once in Merchant Center or Search Console. Google's merchant return policy documentation gives those settings precedence over markup. Audit what your theme and apps output, since two plugins can write product markup for the same page. Then validate a representative page from each template.

Review markup must describe the product, not a recycled store rating. Reviews a business controls about itself aren't eligible for review stars on LocalBusiness or Organization markup. That includes embedded third party widgets. Product pages with genuine product reviews still are.

Even then, Google's review snippet documentation doesn't guarantee that features using structured data will show. Our guide to schema markup covers the implementation.

Merchant Center adds the feed side. Google's free listings for products page says products can appear at no cost on:

• Google Search

• Google Maps

• Gemini

• YouTube

• the Shopping tab

• Google Images

• Google Lens

These free listings sit alongside any paid campaigns. In most cases they're on by default. Australia is one of the countries where you must set up shipping settings or add shipping costs. Judge product data health from the product statuses and issues in your own account.

Use both routes. Google's guide to share your product data with Google recommends structured data plus a Merchant Center feed. Here's why:

• crawling isn't guaranteed to find every product

• Merchant Center is mandatory for some surfaces, such as listings in the Shopping tab

• Google doesn't guarantee how long changes found by crawling take to process

• a feed can be updated weekly, daily or even hourly

So a store whose prices change often needs the feed. The detailed setup belongs to our guide to Google Shopping ads and Merchant Center.

Accurate product data matters for AI features in Search too. Google's page on AI features and your website says AI Overviews and AI Mode have no additional requirements. No special schema or AI text files are needed.

The page lists keeping Merchant Center and Business Profile information current among the practices that apply. But it guarantees nothing about selection. Our guide to generative engine optimisation covers the rest.

One catalogue, several representations

The same product lives in several places at once. That's the catalogue system, the product page, the markup, the Merchant Center feed and any marketplace listing. They must all describe the same purchasable offer.

Name the source of truth for each fact. That covers product and variant IDs, price, stock and delivery promise. And name the person who owns it.

Keep your own codes and manufacturer codes apart. Your SKU is yours to assign. A GTIN comes from the manufacturer. Google's GTIN help page says not to guess or make up a value. If you're unsure of a GTIN, don't submit one.

Each variant has its own GTIN. Products that genuinely have none don't need one. A missing or incorrect GTIN can limit visibility. So never enter a borrowed or guessed code just to clear an error.

A local shop's Business Profile is one more representation: a shop window, not a store or a feed. Google's About product editor page says that tool suits a local store selling physical goods. That's particularly true for one with no online checkout, no barcodes or no current stock data. A profile linked to Merchant Center manages its products there. Our Google Business Profile guide covers the profile itself.

Around every promotion and stock change, run a small reconciliation.

It's a checking sequence, not a claim that every system pulls from the website or updates instantly. In practice, the check runs like this:

• sample a product

• open its destination with the selected variant

• compare the visible price with the checkout

• inspect the markup and the Merchant Center data

• record any mismatch and who owns the fix

Look hardest at expired sale prices, backorders, the wrong currency and landing pages that don't load. A feed that was valid last week proves nothing about today.

When product data goes wrong

Google checks for mismatches. Its help page on a mismatched product price says Googlebot routinely crawls landing pages. The crawler then compares them with your product data. The price in the page's HTML needs to match exactly.

A mismatched product may be disapproved. The page also warns that mismatches may lead to an account suspension.

The ACCC's price displays guidance gives an Australian reason to move fast as well. Say the checkout price doesn't match the displayed or advertised price, even by mistake. The business must sell at the lowest of those prices. Or it must stop selling the item until the price is corrected.

When it happens, work in this order:

• Confirm the offer a customer can actually buy right now.

• Stop publishing the wrong data through the affected channel.

• Correct the source of truth, not just the symptom.

• Refresh the downstream copies: the page, the markup and the feed.

• Verify the processing status, and request a review where the account offers one.

Google says it recrawls disapproved products automatically over the following hours or days. So don't promise anyone a reinstatement time. Merchant Center's automatic item updates can correct price and availability from what Google finds on your pages. But Google says they aren't a replacement for regular updates of your product data. Keep a record of every incident.

Keep account identifiers and customer details out of any copy you share. A record like this also answers the first question after a sales dip. Was the product available, at the right price, when people looked?

Australian buyers and marketplaces

Australian shoppers judge a store on the price they'll actually pay. They also judge when the parcel will arrive and whether they can send it back. Get those right on the page and in the data. And say them the same way everywhere.

The minimum total price rule covers the product page. The same total should flow into your product data. Google's help on the price attribute asks for an amount and currency matching the landing page and checkout. For Australia, that means GST included and shipping costs left out of the price. Google also asks you not to change a landing page price based on the visitor's location.

Drip pricing is where fees appear late in the checkout. It's one of the practices our guide to dark patterns and Australian consumer law covers. Launch compliance, from GST to payments, sits in our ecommerce website checklist.

Marketplaces are a separate decision. Amazon Australia and eBay are discovery and sales channels of their own. Each has its own fees, customer relationship and listing controls.

A marketplace can reach buyers your store wouldn't. But being on one doesn't replace an owned store strategy, or prove you should list everywhere. Keep your own store worth visiting, with selection advice, service and clear product data. And plan around the marketplaces that still operate.

Platform, international and migration checks

Your platform decides how much of this you control. No platform is best for rankings. What matters is who controls URLs, filters and markup, plus redirects, product feeds and performance. Just as important is who checks them after every change.

The platforms' own documentation spells out the traps. Shopify's robots.txt.liquid documentation says its default file works for most shops. It also says the template, once added, replaces the generated file.

Shopify's help page on creating and managing URL redirects adds another trap. A redirect won't work while the URL still loads a valid page. That matters when you retire a product.

WooCommerce's permalinks documentation warns that changing the permalink structure updates the URLs of existing products. Links and search results pointing at the former URLs then return 404 errors unless you set up redirects.

The detailed comparison sits in our guide to Shopify vs WooCommerce. Our guide to the best ecommerce platform in Australia covers the rebuild decision.

Test performance on the templates that carry revenue: category pages, product pages, filtered views and checkout. Test them on the devices your customers use. Don't assume every speed problem is an image problem. Read the field data for each template. Our guide to Core Web Vitals explains the measures and how to read them.

Selling overseas

Give each overseas market stable regional pages with the right currency, delivery terms and language. Google's guide to managing multi regional and multilingual sites lists clear geotargeting among a country code domain's advantages. It doesn't recommend URL parameters for locale. And it says not to use IP analysis to adapt your content.

Avoid automatic redirects between versions. Link to the other versions so customers can change market themselves. Regional pages can share a language and very similar content. For those, Google says to pick a preferred version with rel="canonical". It also says to add hreflang so the right regional URL is served.

Those hreflang annotations have to be reciprocal. Google's guidance on localised versions of a page says each version must list itself and every other version. Use x-default as the fallback when no version matches the visitor.

The annotations aren't a duplicate content cure either. The same page says localised versions count as duplicates when the main content stays untranslated. So keep the canonical and regional signals compatible.

Ecommerce migration acceptance checks

A replatform puts every one of these decisions at risk in a single release. Our website migration SEO checklist covers the full process. A store adds these acceptance checks:

• Map every valuable product, category, guide, filter and discontinued product URL to a relevant outcome. Include pages with links or support value.

• Keep product identifiers and review relationships where they still apply. Test that Merchant Center and other live destinations open the right variant.

• Check redirects, canonicals and robots rules on a representative page from each template. Do the same for pagination, sitemaps and structured data.

• Verify stock, price, shipping and payments through an approved test purchase. Check enquiries, consent and analytics the same way.

• Name the owners for feed continuity, release monitoring and rollback before launch day.

Test a small set that covers each template and exception first, then validate more widely. Nobody should paste every product into the Rich Results Test by hand. Automated sampling and developer checks scale to a large catalogue in a way manual testing can't. A store that needs a rebuild instead can see how we approach it in ecommerce website design.

Measure sales contribution

Measure the sales search contributes, not the traffic it brings. Group your pages into categories, products and supporting content, then report each group separately. Take each number from the system that holds it.

Merchant Center's guide to track your product performance explains where free listings sit in its reporting. Our guide to GA4 for small business covers the setup.

Define the rules before you compare anything. Settle the attribution model and whether you count by order date or dispatch date. Agree how channels are grouped as well. A session's landing page isn't necessarily where the customer first found you.

Then track the measures that connect to money:

• stock availability

• product views

• basket additions

• checkouts started

• completed orders

• cancellations

• returns

• contribution

Reconcile transaction IDs between analytics and the order system where you can. Two rules stop the obvious double counting. Don't add assisted revenue to direct revenue for the same order. And don't assume Search Console and Merchant Center counts can simply be added together.

The same split makes a sales drop readable. When organic revenue falls, check categories, products and buying guides separately. Then check stock levels, the match between feed and site, seasonality and any change to the checkout. Do all that before you blame an update. Our guide to Google algorithm updates covers how to tell whether one was involved.

An example monthly store ledger

Here's how the reconciliation works for one Australian retailer. All amounts are in Australian dollars, excluding GST, for one month's reconciled order cohort. Recorded gross order value is $50,000. After $5,000 of cancellations and refunds, net sales are $45,000.

The return on the SEO work is contribution after its cost, divided by its cost.

Treat that figure carefully. Attribution isn't proof the SEO work caused every one of those sales. The figure leaves out fixed overheads and other excluded costs. And 142.9% is arithmetic, not a target. Our guide on how to measure SEO ROI goes further into the method.

Then allocate the same $45,000 once, across landing page groups that don't overlap. Under this model, categories took $27,000. Products took $13,500 and guides $4,500.

That's 60%, 30% and 10% of the same net sales. It's a teaching allocation, not evidence that categories always contribute 60%. If you compare several stores or periods, report each cohort and its method separately rather than pooling them.

A ninety day plan

Ninety days is enough to set baselines, fix the worst problems and start building. It isn't a guarantee of rankings or revenue inside that window. Give each phase an owner and the evidence that says it's done.

The evidence decides the order. A generic SEO score doesn't.

When to bring in specialist help

Get help when the catalogue is too big to check by hand, or a migration is close. Do the same when product data problems keep coming back and nobody owns the fix. A good specialist starts with your order data and your product data, not a keyword list. They also put acceptance checks in writing before any work starts. That's how we approach ecommerce SEO support.

What we recommend at Elev8d

FAQ

Next steps

Written by Ajay K.Co-founder, Elev8d

Ajay is the co-founder of Elev8d and writes most of what you will read here. Psychology grad turned marketer, which is a fancy way of saying he cares about why people click, not just where things rank. Every guide is written in plain English for Australian businesses, tested on real campaigns before it earns a spot on this site.