Ecommerce SEO is not service-business SEO with a shopping cart bolted on. The money sits in category and collection pages, the crawl budget gets eaten by filter URLs nobody asked for, and authority leaks across variants of the same item. We are an Australian SEO agency that works the catalogue as a system, then reports revenue per session rather than sessions.
Somebody searching a category wants to browse a range and compare. Google sends them to a single product they may not want, they bounce, and the page that should have owned the search was never built properly.
Filters, sort orders, pagination and session parameters combine into tens of thousands of near-identical pages. Crawlers spend their budget there instead of on the products you just added.
So is every competitor's, word for word. There is nothing on the page a search engine or an AI assistant can use to prefer you, so it falls back to whoever has the stronger domain.
Most store owners assume product pages do the selling, so that is where the effort goes. The search data says otherwise. Head demand in almost every vertical is category-shaped, and the page that can satisfy it is the collection page, not a single SKU. A product page answers one question. A category page answers the one people actually type.
So we map demand to page type first. A real category with real volume gets a page built to satisfy it, and where demand only supports a filtered subset we decide deliberately whether that subset earns an indexable URL or stays a parameter.
Four filters with ten options each is ten thousand combinations, and every one is a crawlable URL if nobody decided otherwise. Left alone, a crawler works through colour and size permutations of a page it has already seen while your new season stock sits undiscovered. It is the most common technical problem in ecommerce and almost never on the previous agency's list.
The fix is a set of decisions, not a plugin. Which facets may produce an indexable page because people search that way, which may be crawled but not indexed, and which should never be a crawlable href at all. Then the same discipline on sort orders, pagination and internal search URLs.
The same shirt in eight colours is often eight URLs, each with the same description, each holding a fraction of the authority one page would have had. We decide which level is the canonical buying unit, consolidate to it, and keep the variant selection working for customers without splitting the page into near-identical copies.
Then the descriptions. Manufacturer copy pasted into a thousand product pages is duplicate content at a scale no service business meets, and rewriting all of it by hand is not realistic. So it gets prioritised: products that already earn, products with demand and no position, then the templates governing the long tail.
Discontinued products are handled worst of all. Deleting a URL that has accumulated links and ranking history destroys authority you paid for. Out of stock is not a reason to remove a page, it is a reason to say when it returns and what to buy instead.
Shopping questions have moved. Ask an assistant for the best waterproof jacket under two hundred dollars in Australia and you get three or four suggestions, sometimes with a store attached. That is a shortlist, not a results page, and there is no position eleven to climb from. Whether you appear depends on whether a model can find clear, factual, quotable information about your products.
The work overlaps with good ecommerce SEO more than most people expect. Accurate structured data, real specifications, honest reviews, clear shipping and returns information, and buying guides that answer the comparison question directly. The same pages that make an assistant confident enough to name you convert better for humans too.
Traffic is the easiest ecommerce metric to inflate and the least useful. Publish enough top-of-funnel guides and sessions climb while the revenue line stays flat. We report organic revenue and transactions by landing page type, with revenue per session alongside, so added traffic has to justify itself.
Template speed belongs in the same conversation. Category and product templates render thousands of times, so a slow template is not one slow page, it is the whole commercial surface of the store. We measure templates rather than the homepage, because the homepage is rarely where the money enters.
If you want the mechanics of how an engagement runs, our approach sets out the sequence, and pricing explains how we scope a store rather than a page count.
Category demand, filter URLs, variant splits, thin descriptions and internal linking exist on every platform. Shopify, WooCommerce, BigCommerce, Magento and custom builds differ in how the fix is implemented, not in what the fix is. That is why this page is written about the strategy rather than about any one admin panel.
Shopify is common enough in Australia to deserve its own treatment. If you are on it, read our Shopify SEO agency page for the platform-level detail, then come back here for the catalogue strategy that sits on top of it. Whatever the stack, the audit is done against your real theme, apps and front end, and migrations are mapped URL by URL before anything goes live.
Stores in Sydney can meet in person. Melbourne, Brisbane, Perth, Adelaide and everywhere else runs remotely on the same reporting.
Ecommerce sits inside a wider programme. These pages cover the platform detail, the neighbouring industries and the way we run the work.
Platform-specific work for stores built on Shopify.
The main hub, including SEO by city and by industry.
How search converts differently in each sector we work in.
Bottom-of-funnel search and being named in AI shortlists.
Technical, content, local and answer-engine work in full.
For stores with physical locations and click-and-collect.
What happens in the first ninety days, written out in advance.
How we scope a catalogue rather than a page count.
A service site has maybe thirty pages and every one of them is written by hand. A store has a template, a feed and tens of thousands of URLs generated by software. That changes the job. Most of the work is deciding which pages deserve to exist, keeping the rest out of the index, and making sure the handful of category pages that carry the demand are actually built to rank rather than left as a bare grid of tiles.
Usually not. A product URL that has been live for years has links, internal equity and ranking history attached to it, and deleting it throws all of that away for the sake of a tidy catalogue. If the item is coming back, keep the page up and say when. If it is gone for good, redirect it to the closest replacement or to its category, not to the homepage. Only remove a URL outright when nothing useful ever pointed at it.
Yes, and on WooCommerce, BigCommerce, Magento and custom builds. The strategy on this page is platform agnostic because the problems are: category demand, filter URLs, variants, thin descriptions and internal linking exist everywhere. The platform decides how the fix gets implemented rather than what the fix is. We keep the Shopify-specific detail on our Shopify SEO agency page.
Revenue and transactions from organic, split by landing page type, plus revenue per session so you can see whether the traffic we added was worth having. Rankings and impressions sit underneath as diagnostics. We will not quote you an uplift figure we have not earned, so what you get instead is a plain view of which pages earn and which ones cost you crawl budget.
We will show you how many URLs your filters are generating, which category pages are losing searches they should own, and where revenue is leaking between product view and checkout. Free, and specific to your store.