Five tips for ecommerce SEO

When an online store owner comes to us saying “ads keep getting pricier, and organic traffic is almost nonexistent,” we already have a good idea of what the audit will reveal. The money is rarely lost where people look for it. Not in the design, and not in competitors’ prices. It leaks out through categories that don’t match how people search for products; product pages copied from the supplier’s site; filters that spawn thousands of junk pages; a lack of reviews; and a mobile version shoppers abandon before the “Buy” button even loads.

After years of working with stores selling clothing, electronics, auto parts, building materials, and home goods, we’ve boiled it all down to five areas that have the biggest impact in e-commerce. Below is what we do in practice, in the order we recommend you do it.

Why a store isn’t just another website

A service website has a few dozen pages, each of which can be reviewed and optimized by hand. A store has hundreds or thousands of products, plus categories, brands, filters, pagination, and sorting. Search engines see dozens of times more URLs than you have products.

So with stores, we don’t work page by page — we work with rules and templates. We don’t write a title for each of five thousand product pages; we write a template that builds it correctly from the product name, brand, and key attribute. One mistake in a template costs hundreds of times more than a mistake on a single page.

The second thing we always tell clients: beating marketplaces for a query like “buy a smartphone” is expensive and nearly impossible. But a specialized store wins on depth — more precise categories, better descriptions, expertise, and a fast answer to a specific need. All five tips below build on exactly that.

Tip 1. Build your catalog around demand, not inventory

Most often we see a structure copied from the supplier’s catalog tree, or one that just seems logical to the owner. But shoppers search differently. They almost never look for an abstract “category” — they look for a need: “women’s running shoes,” “washable white matte wall paint,” “vacuum for pet hair.”

If all you have is “Sneakers” with a “Purpose” filter, the search engine has nothing to show for “women’s running sneakers” except the general category. And it loses to a competitor with a dedicated page for that exact query — with its own title, copy, and curated selection.

Here’s how we build the structure:

1. We collect keywords from Search Console, Keyword Planner, Serpstat or Ahrefs, and the site’s internal search. More on this in “Building a keyword map”.
2. We group queries by search results. If Google shows the same pages for two queries, that’s one page on the site. If different, two.
3. We stick to three levels: section → category → subcategory. Any deeper and navigation breaks while page authority gets diluted.
4. We check category depth: a category with two products is a poor landing page. Our minimum is 8–12 items.
5. We document the rules: which attributes put a product in a category, so content managers don’t sort the assortment by eye.

Then comes the technical side. We create clear URLs like /shoes/sneakers/womens-running/ and change them only once, always with 301 redirects. We add breadcrumbs with BreadcrumbList markup. Every important category gets its own H1, title, and meta description, plus helpful copy below the products — not an “SEO wall of text,” but advice on how to choose. More on this in “On-page SEO basics”.

We also set up internal linking. Categories link to their subcategories, and neighboring ones link to each other where it makes sense for the shopper: “Running socks pair well with these sneakers.” The main menu shouldn’t have two hundred items — we put key sections there and show deeper levels inside categories. That makes navigation easier for people and helps crawlers reach important pages faster.

And two mistakes we fix most often: the same product in several categories under different URLs (those are duplicates — there should be one URL), and empty categories that serve a blank page instead of similar products and a back-in-stock signup.

Tip 2. Create product pages that sell and rank

The product page is where a person decides: buy, or go to a competitor. And that’s exactly where stores cut the most corners. A manufacturer’s description, photos from a price list, half the specs missing. Thousands of stores have identical product pages — and search engines have no reason to show yours.

In our experience, a strong product page has nine essential elements:

1. A clear name: type + brand + model + key attribute.
2. 4–6 quality photos, ideally a video too.
3. Price, availability, and delivery times — above the fold.
4. A “Buy” button and a simpler alternative: “one-click order” or “request a callback.”
5. Full specs in a table, consistent across the category.
6. A unique description: who the product is for and how it beats similar ones.
7. Customer reviews and questions.
8. Shipping, payment, return, and warranty terms.
9. Related and similar products.

“We have five thousand products; we physically can’t write that many descriptions,” we hear all the time. You don’t have to write them all at once. We start with the top 100–200 products by revenue: they usually bring in most of the sales. Next come products with impressions in Search Console that are stuck on page two — a new title and description here often drive quick growth. The rest we cover with a well-designed template that builds coherent copy from real specs. That still beats a copy of the manufacturer’s description.

A quick illustration of why the name matters. “Gel-Nimbus 25 1012B356-020” is how a product often arrives from the supplier. “Asics Gel-Nimbus 25 Women’s Running Shoes, Gray” is how a person searches for it. The first tells nothing to either a shopper in the catalog or a search engine. The second answers the query right away and looks better in search results. We always rework the naming template before touching descriptions.

Three more things we check in every store:

• Product variants. A T-shirt in 12 colors and 6 sizes is one product with a selector, not 72 near-identical pages. The canonical URL (rel="canonical") should point to the main product page. More than once we’ve seen canonicals on all product pages mistakenly point to the homepage — and products simply dropped out of the index.
• Product markup. Price, availability, and rating in structured data show stars and price right in search, and such snippets get noticeably more clicks. But we add ratings only where reviews actually exist on the page.
• Out-of-stock products. Temporarily unavailable — we keep the page and offer a back-in-stock signup and alternatives. Discontinued but with a replacement — we set up a 301 redirect. No replacement and no traffic — we return a 404 or 410. We don’t recommend anyone mass-delete pages with history and backlinks.

Tip 3. Get your filters under control

Filters are the most useful part of a store for shoppers and the most dangerous for SEO. Every combination creates a new URL. Crawlers waste time on junk and reach new products less often, duplicates and empty pages appear, and a filter page “steals” a query from the category, so neither reaches the top.

Our solution is to split filters into two types:

• SEO filters — combinations with real demand: “women’s,” “running,” “Nike,” “winter.” We give them static URLs (/sneakers/womens/), their own H1, title, and copy, and open them for indexing. Essentially, they’re full-fledged categories.
• Utility filters — price ranges, sorting, rare combinations. Useful for people but without search demand. We close these with noindex, follow or a canonical to the base category.

Our working rules: we index combinations of one or two filters at most; we fix the order of parameters in the URL; we don’t open pages with fewer than 3–5 products; we keep utility filters out of the sitemap. Pagination stays crawlable, with a self-referencing canonical, and category copy appears only on page one. If the site uses infinite scroll, it must have regular links to pages 2, 3, 4 underneath.

One last tip: review your internal search report once a month. If people regularly search for “sneakers for wide feet” and there’s no such page, that’s a ready-made idea for a new SEO filter. The on-site search results pages themselves we always block from indexing.

Tip 4. Collect reviews systematically

In a physical store, a shopper can hold the product and talk to a salesperson. Online, they only have a screen. Reviews make up for the missing human contact. They’re also one of the strongest SEO tools: they add fresh, unique text to the page, written in the same words people use to search — “does it run small,” “will it fit wide feet.”

Most happy customers don’t leave reviews not because they don’t want to, but because nobody asked. Here’s what works in our projects:

1. An automated message 5–10 days after delivery — ideally via WhatsApp, Viber, or Telegram, with a direct link to the review form for that product.
2. Minimal fields: stars, text, name. No mandatory sign-up.
3. The option to add photos — these reviews are the most convincing.
4. A small reward for any honest review, not just positive ones.
5. A separate “Questions & Answers” section on the product page.

We never advise deleting negative reviews. A site where everyone is thrilled looks suspicious. A calm public reply that solves the problem works far better — it shows the store won’t vanish if something goes wrong. Often it’s a well-handled negative review that convinces someone to order. Fake reviews are a risk not worth the short-term boost.

Beyond reviews, we check basic trust signals: an “About Us” page with real photos, full contact details and messengers, clear shipping, return, and warranty terms, and a Google Business Profile. And a blog: “how to choose” articles capture queries category pages don’t answer and lead readers to products. How we plan this content is covered in “Writing content for your website”.

Tip 5. Speed, mobile, and honest analytics

You can have a perfect structure and hundreds of reviews and still lose sales on a slow mobile site. Most store visits today come from smartphones, and Google evaluates sites primarily by their mobile version.

Our first test is always the same: we open the store on an ordinary smartphone over mobile data and follow the shopper’s path — from search to checkout. Are the price and button visible without scrolling? Does a signup pop-up cover everything? Does the numeric keypad open for the phone field? Is there one-tap payment? Every extra step loses some buyers. More in “Succeeding with mobile-first indexing”.

For speed, we follow Core Web Vitals:

Metric	Target	Most common culprit in stores
LCP — main element load	up to 2.5 s	Heavy images, slow hosting
INP — interaction response	up to 200 ms	Heavy filter scripts, widgets, chats
CLS — layout stability	up to 0.1	Banners and images without set dimensions

Quick wins we implement almost every time: images in WebP or AVIF with responsive sizes, lazy loading for everything except the main photo, removing old ad pixels and widgets, caching, and a CDN. If the server takes longer than 600 ms to respond, it’s time to talk about hosting. We handle this as part of our website speed optimization service.

But the most important thing is analytics. Without it, you don’t know what from this article is already working. Our minimum for a store: GA4 with e-commerce, Search Console, tracking of calls and messenger orders (in many markets there are lots of them, and without them SEO looks weaker than it really is), and separate reports for categories, filters, product pages, and the blog.

Once a month we also check technical hygiene: the sitemap contains only indexable pages with a 200 status code; robots.txt blocks the cart, account area, and search but not styles and scripts; there are no redirect chains or broken links. If the store runs in several languages, each version needs its own URL and correct hreflang tags. A sudden jump in page count in the Search Console indexing report almost always means a filter problem.

Every month we look first at organic clicks by page type, the number of queries in the top 10, product page CTR, conversion, and organic revenue. Not at rankings for three “main” keywords — at the money.

What this looks like in practice

A typical case: a home goods store with about three thousand products, steady advertising, but organic traffic flat for years. In the audit we find a catalog copied from the supplier, with categories like “Miscellaneous”; over forty thousand indexed pages due to open filters; descriptions identical to dozens of other sites; almost zero reviews; a mobile version that loads in six seconds; and analytics without e-commerce tracking.

First, we set up analytics, close utility filters, and speed up the site. Then we rebuild categories around keyword research, create a few dozen SEO filters, rewrite product pages for the two hundred most profitable products, and launch automated review requests. At first, the number of indexed pages drops — which is good: the junk is going away. Then the number of queries in the top 10, product page CTR, and mobile conversion gradually grow. Exact numbers depend on the niche, so we don’t promise “plus 300% in a month” — but the trend shows up in analytics, in revenue.

Seven mistakes we see most often

If you recognize your store in at least three of these, that’s where we’d start.

1. Relying only on ads. They deliver fast results but stop when the budget does. Organic traffic accumulates and over time makes every order cheaper.
2. Descriptions copied from the supplier. A page identical to dozens of others has no chance at top rankings.
3. Uncontrolled filters. Tens of thousands of junk pages in the index instead of a few hundred useful ones.
4. Changing URLs without redirects. After a redesign or a move to a new CMS, the store loses authority its pages built up over years.
5. Identical titles for all products. A “Buy at our online store” template with no product name — and search engines can’t tell what the page is about.
6. Testing only on a big monitor. Even though most shoppers come from smartphones.
7. One-off pushes instead of a system. A month of active work, then six months of silence. Search engines value steady growth.

Where to start: our six-month plan

You can’t do everything at once. Here’s the sequence we suggest to clients:

Period	What we do
Weeks 1–2	Audit, analytics, Search Console, keyword research
Weeks 3–6	Technical fixes: indexing, filters, duplicates, speed
Months 2–3	New category structure, SEO filters, meta tag templates
Months 3–4	Top product pages, Product markup, review collection
Months 4–6	Blog, internal linking, external mentions

The first changes after technical fixes usually show up within 4–8 weeks, and a noticeable rise in organic revenue within 3–6 months. We covered realistic timelines in detail in “Growing traffic in six months”.

Frequently asked questions

Can you compete with marketplaces?

For broad queries, it’s hard. For narrow, expert ones, absolutely. A better structure within the niche, more detailed descriptions, articles, and personal service — that’s what we build strategies for smaller stores on.

Does a store need a blog?

Yes, if it answers shoppers’ real questions and leads them to products. “How to choose” guides, model comparisons, and care instructions bring in informational traffic and build trust. A box-ticking blog of discount news doesn’t.

Should you switch CMS for SEO?

Rarely. WooCommerce, Shopify, Magento, OpenCart, and other popular platforms let you do everything described here. A migration always risks losing rankings, and we recommend it only when the platform truly can’t handle critical tasks.

What comes first on a limited budget?

Set up analytics so you can see the money. Fix technical issues that block indexing. Then the meta tags and copy for the 10–20 categories that bring in the most revenue. These three steps pay off fastest.

The bottom line

The five pillars of store promotion are a demand-driven structure, product pages that sell, managed filters, trust, and technical quality with analytics. None works alone: great product pages won’t help if crawlers get lost in filters, and a perfect structure won’t sell anything on a site that takes eight seconds to load.

The good news: most stores don’t do even half of this. So systematic work is a real chance to overtake competitors who still rely on ads alone.

If you’d like us to review your store and draw up a plan, get in touch with us. We run audits and provide end-to-end SEO services, and if you don’t have a store yet, we’ll help with e-commerce development with SEO built in from day one.