E-commerce SEO
Category pages that get found and product pages that sell.
E-commerce SEO is a structural discipline more than a writing one. The difficult parts are category architecture, what happens when filters generate thousands of URLs, and what to do with products you no longer sell.
How is e-commerce SEO different from local SEO?
Local SEO is about a small number of pages and one business listing ranking for searches near a place. E-commerce SEO is about a large number of pages, where the structural problems dominate: filters generating near infinite URLs, products duplicated across collections, out of stock items, and thin product descriptions supplied by a manufacturer and used by every other retailer.
The highest value pages are usually categories rather than individual products, because category pages match how people search when they are still choosing, and they can be made genuinely useful in a way a product page cannot.
The other half of the work is not search at all. Baymard Institute's aggregate of fifty studies puts documented cart abandonment at 70.22%, which means that for most stores the checkout leaks more revenue than the traffic does.
Why category pages matter most
Somebody searching for a specific product by model number has already chosen. Somebody searching for the category is still deciding, and that is where a store can actually win a customer rather than compete on price for one who has already made up their mind.
Category pages are also the pages a store controls completely. Product descriptions are frequently supplied by the manufacturer and used verbatim by every other retailer, which makes them close to worthless as a differentiator. A category page can contain a genuine buying guide, comparisons, and answers to the questions people ask before choosing.
The structural consequence is that category architecture is the strategy. Which categories exist, how they nest, which filter combinations get a real page and which do not, and how they link to each other, decides most of what a store can rank for.
Product pages still matter, particularly for branded and model specific searches. They are just rarely the place a store should invest first.
Where the money actually leaks
Baymard Institute aggregates fifty separate studies of documented cart abandonment spanning 2006 to 2025, and surveys the reasons shoppers give.
SourceCart abandonment rate statistics, Baymard Institute, 2025, 50 studies
None of that is a traffic problem. If your checkout is losing seven in ten carts, doubling your visitors doubles the loss as well as the sales.
What the work consists of
In the order that produces the most, which is rarely the order clients expect.
Get the crawl under control
Faceted navigation is the biggest technical problem in e-commerce SEO. Filters for size, colour, price and brand combine into an enormous number of URLs, almost all of them near identical, and a crawler can spend its entire budget on them while never reaching your new products.
- Decide which filter combinations deserve an indexable page, and index only those
- Block or canonicalise the rest rather than leaving them open
- Confirm in Search Console that the excluded URLs really are excluded
Fix duplication across collections
Most platforms generate a URL for each product inside each collection it belongs to, which means a single product can exist at six addresses. Canonical tags pointing at one primary URL resolve it, and on some platforms this needs template level work rather than a setting.
Build category pages worth ranking
A heading, an introduction that actually helps someone choose, the products, and genuinely useful content underneath: how to choose, what the differences mean, what people ask. Not four hundred words of keyword padding below the fold, which is the standard approach and is visibly padding.
See the remaining steps: What the work consists ofHide the remaining steps: What the work consists of
Differentiate product pages where it matters
Manufacturer descriptions used verbatim leave you identical to every competitor. Rewriting all of them is rarely economic, so we rewrite the ones that carry real revenue and leave the long tail alone. Reviews, real photographs, sizing guidance and honest answers to common questions do more than rewriting the specification table.
Decide what happens to discontinued products
This is a policy decision that most stores never make, which is why they accumulate thousands of dead pages. Temporarily out of stock pages should stay and say so. Permanently discontinued products should redirect to the closest replacement or the parent category, not to the home page and not to a 404.
Get the product feed and markup right
Product structured data with accurate price and availability, and a Merchant Center feed that matches the site. Mismatches between the feed and the page cause disapprovals in shopping campaigns and undermine the rich results.
Then fix the checkout
Total cost shown early, guest checkout available, the fewest possible steps, and a checkout tested on a real phone. This is conversion work rather than SEO, and for most stores it returns more than the SEO does.
Steps one and two are unglamorous and frequently produce the largest single change, because they let a search engine see the store properly for the first time.
What a store's structure should look like
The shape of the site decides what can rank, more than the content on any individual page.
A product buried six clicks deep with nothing linking to it is effectively invisible, however good the page is. Depth from the home page is a practical constraint, not a theoretical one.
The same structure that helps a crawler helps a customer. If a person cannot find the product category they want in two clicks, the architecture is wrong for both audiences.
Signs of a store with a structural problem
Most of these are visible in Search Console in a few minutes.
- Index coverage showing tens of thousands more URLs than you have products, which means facets are being crawled.
- The same product indexed at several URLs because it belongs to several collections.
- Category pages with no content other than a product grid, competing against competitors who wrote a buying guide.
- Product descriptions identical to every other retailer because they came from the manufacturer.
- Discontinued products returning 404 errors, or redirecting to the home page, which is a worse pattern than either leaving them or redirecting properly.
See the full checklist: Signs of a store with a structural problemHide the full checklist: Signs of a store with a structural problem
- A Merchant Center feed with disapprovals nobody has looked at, which silently removes products from shopping results.
- No guest checkout, when forced account creation is one of the most cited abandonment reasons.
- Extra costs appearing only at the last checkout step, which is the single most cited reason people abandon.
The last two are not SEO problems and will usually make you more money than any of the others.
How this is measured
Revenue by landing page, which tells you which pages actually earn. Organic revenue as a share of total, which tells you whether the channel is growing or whether the store is. Index coverage, because the structural work either reduces the junk or it did not happen.
Then the conversion side, which for a store is inseparable from the traffic side: conversion rate by device, cart abandonment rate, and checkout step completion. A store with rising traffic and a leaking checkout is spending more to lose more.
Average order value and repeat purchase rate matter too, because they change what you can afford to pay for a visitor, which changes the entire economics of the paid channels.
Ranking reports on individual product keywords are close to useless here. Revenue by page is the measure.
What varies by store
Catalogue size changes everything. A store with two hundred products can have every product page written properly. A store with forty thousand cannot, and the strategy becomes category level with selective product investment.
Branded versus unbranded matters. A store selling its own products controls the descriptions and has no direct competition on its exact product names. A store reselling other brands is competing with the manufacturer and every other retailer for the same terms.
Marketplace exposure changes the calculation. If a substantial share of your sales come through a marketplace, investing in your own site's search visibility is partly a margin decision rather than a volume one.
Seasonality is usually extreme. Work done in the quiet season is what capitalises on the peak, and a store that starts in October has already missed the window it was aiming at.
What we need from you
Access to the store admin and the product data, and someone who understands how your catalogue is actually organised, including the parts that are historical accident.
Decisions about discontinued products. This is a policy we need from you once, and then it applies automatically.
Margin data, or at least relative margin by category, because it decides where effort is worth spending. Ranking a low margin category is a poor use of a budget.
Willingness to change the checkout. It is often the highest return work available and it is frequently the hardest to get approved.
We do not publish a price for this piece of work on its own, because the right scope depends on what already exists. What is published is the bundle pricing: 2,400, 3,600 or 4,800 dollars a month depending on which channels are running. You can read the full breakdown on the pricing page, and you will get an exact number in writing before anything starts.
How this connects to the rest
This sits inside our SEO and GEO program and leans heavily on technical SEO, because the hardest problems here are crawling and duplication rather than content.
Shopify design and development is where the structural fixes get implemented if you are on that platform, and conversion rate optimisation is where the checkout work lives.
Paid advertising interacts closely with an online store: shopping campaigns depend on the same product feed, and the search term data tells you which categories are worth building.
Content strategy applies to the buying guides that make category pages worth ranking, which is usually where the editorial budget belongs rather than on a blog.
Find out whether your problem is traffic or checkout.
We will look at your index coverage and your conversion funnel before the call. For a lot of stores the answer is that more traffic is the expensive way to fix a cheaper problem.
Straight answers.
Should I write unique descriptions for every product?
For a small catalogue, yes. For a large one it is rarely economic, and the return is concentrated in your highest revenue products.
Customer reviews, real photographs and genuine buying guidance usually do more than rewriting a specification table.
What should happen to products we no longer sell?
If it is coming back, keep the page and say it is out of stock, with an alternative offered. If it is gone for good, redirect to the closest replacement or to the parent category.
Returning a 404 wastes any authority the page had, and redirecting everything to the home page is worse than a 404 because it is misleading.
Do filters hurt SEO?
Uncontrolled ones do. Filter combinations generate enormous numbers of near identical URLs, and a crawler can spend its entire allowance on them without reaching your new products.
The fix is deciding which filter combinations deserve an indexable page and blocking or canonicalising everything else.
Is SEO or paid better for an online store?
Paid gives you control and immediate volume, and it stops when you stop. Search work compounds and is slower. Most stores that are working run both.
The deciding factor is usually margin. If your margin cannot support the cost per click in your category, search work is not optional.
How long does e-commerce SEO take?
Structural fixes to crawling and duplication can show within a few weeks, because they change what the search engine can see rather than how good it is.
Category ranking in a competitive niche is measured in quarters, and seasonality means a fair read often takes a full annual cycle.
Where this comes from.
Primary documentation and published research behind the guidance on this page.
- Google Merchant Center Help (opens in a new tab)
- Structured data general guidelines (opens in a new tab)
- Google Search Console Help (opens in a new tab)
- Baymard Institute cart abandonment research (opens in a new tab)
- Google Search Essentials (opens in a new tab)What Google requires, recommends and forbids.
Talk to the team
A short call, a look at how the business currently shows up, and a straight answer on what we would do first.