Ecommerce SEO / Practical guide
Which ecommerce SEO problems should you fix before adding more products?
More products can repeat a fault you already have. I would separate confirmed launch blockers from investigations and improvements that can wait.
More products can multiply a problem you already have
Before adding another range, I would test a few existing products. Does the selected size reach the basket? Are the price and stock correct? Can a customer find the range from a category page? If a shared template gets these wrong, the next import can copy the problem into hundreds more pages.
The website does not have to be perfect before you add products. A missing paragraph and a broken payment step are different priorities. I want to know which problem affects this range and whether it needs fixing before customers arrive.
This guide helps you make that expansion decision. For a wider review of categories, product information and feeds, use the existing ecommerce SEO checklist. Here, the question is what needs fixing before this range goes live. Some of those fixes affect sales directly, not just SEO.
Where would you start?
Use the three groups below to decide what needs fixing, what needs checking and what can wait.
Review the three decisionsUse this as a starting point, then check the affected products with your team.
Start with three decisions, not one long warning list
Fix before expanding the affected range
A repeatable fault prevents a legitimate purchase, misstates the offer or will be copied into the new products. Examples include the wrong variant reaching the basket, incorrect shared price data or an unintended restriction on the new product template. Confirm the scope, then repair and test it.
Investigate before choosing a repair
There is a concerning signal but no confirmed cause: important products appear difficult to discover, filtered URLs are multiplying, or search visibility has changed. Gather representative evidence before applying broad canonical, crawl or navigation rules.
Launch the unaffected products and assign the remaining work
Customers can find and buy the products, and no serious fault has been found in their template or import. Assign optional copy and image improvements for later. Your team still needs to approve the product details and normal launch requirements.
A serious fault in one template may justify holding that range while another, tested range continues. Be clear about which products the problem affects.
1. Check what the next import would copy
Follow one proposed product from the source data through the import, product page, variant selector and basket. Then test another product using the same template and one that does not. That helps show whether the problem is one record or a setting used across the range.
Check the actual customer-facing values: item identity, selected size or colour, price, currency, stock, delivery information and what is included. A correct spreadsheet does not prove the public offer is correct. Equally, one stale item does not prove every product needs rewriting.
| What you found | What I would do | What needs to pass |
|---|---|---|
| The new template sends the wrong variant to the basket. | Hold the affected products until the selector is fixed and tested. | Selected item, price and basket item agree on representative devices and cases. |
| A shared import field gives the wrong stock or price. | Correct the mapping before the next affected import. | Check source, processed record and public page; confirm the next import does not restore the fault. |
| New items are reachable only through an internal search box. | Add useful category links before relying on the new pages to attract search visits. | Useful category-to-product links, intended canonical URLs and sitemap entries agree. |
| A product has incomplete optional editorial content but accurate buying information. | Decide whether shoppers need the missing detail before you hold the launch. | Record what is missing, who will supply it and why it matters. |
If the same mismatch also affects a Shopping feed, treat the feed and website as connected checks, not interchangeable systems. The Merchant Center disapproval guide covers that investigation. A feed issue is not, by itself, proof of a Google Search penalty. For scoped feed repairs, see my Google Shopping and Merchant Center support.
2. Check whether shoppers and crawlers can reach the new range
Start from a relevant menu or parent category as an ordinary visitor. Can you reach the range, understand what belongs there and open its products through normal links? Include items beyond the first results page. A successful search inside the shop is not the same as an accessible category route.
Google’s ecommerce navigation guidance recommends links through the category structure to products. I would check that the proposed range has a useful place in the store rather than adding a menu entry purely to satisfy a crawler. The customer needs to understand the choice too.
Do not turn every filter combination into a new landing page
Adding sizes, colours and materials can multiply filtered URLs much faster than it increases the number of actual products. Before expanding the attributes, decide which combinations serve a real search task and which exist only to help a customer narrow the list.
The faceted-navigation guidance describes how excessive URL combinations can consume crawl resources. That is a reason to examine the current behaviour, not an instruction to block every parameter. Some parameter URLs identify genuine variants or useful landing pages. Review their purpose and current visibility before choosing controls.
Make the intended URL signals agree
For representative items, compare internal links, the canonical target and sitemap URL. If they conflict unintentionally, clarify the intended product or variant structure before duplicating it in a larger import. Google’s ecommerce URL guidance is a useful reference; platform-specific variant behaviour still needs inspection.
I would not mass-redirect variants or change a site’s robots rules just to simplify an audit export. First establish which URLs customers use, which need to remain distinct and what the proposed change would remove. The audit prioritisation guide helps turn a confirmed finding into proportionate work.
3. Get the product facts right before writing more descriptions
New products need accurate facts, not a catalogue-wide word-count target. For a replacement part, compatibility may be decisive. For homeware, dimensions, material, care and what is included may matter more. Ask the catalogue owner which facts they can verify and where those facts come from.
If the new range is missing essential specifications, get those from the catalogue owner before ordering hundreds of descriptions. A writer or automated tool must not invent capacity, certifications, compatibility or guarantees to fill the gaps.
Check a mobile buying journey too. Can someone select the intended option, read delivery information and continue to checkout without an obstructive overlay or misleading stock message? Use the business’s approved test procedure; do not create real orders, payments or customer messages just to complete an SEO checklist.
Platform changes need a platform-aware owner. The implementation may differ on Shopify and WooCommerce. I would inspect the actual theme, apps and import behaviour before proposing a repair that belongs to a different store setup.
A launch decision: add the mugs, or fix the template?
Here is a fictional example of how I would approach the decision.
A homeware retailer plans to add a new mug range. During a controlled check, selecting a larger size changes the image but sends the smaller item to the basket. The same behaviour appears on another product using that selector, while a simple product without options works correctly.
What I would hold back, fix and test
What I would hold back: the new products that depend on the affected variant selector. The checks point to a problem with that selector, not every product in the store.
Who needs to fix it: the developer investigates the selector and basket mapping; the catalogue owner verifies the intended variant IDs and prices; the release owner approves testing and the deployment window.
What I would test: representative options, a second affected product and the unaffected product behave correctly. The selected item and basket must agree. Record the change and agree how to undo it if another problem appears.
What I would review later: a later review of discovery, impressions and suitable purchases. The correct item reaching the basket proves a repair, not a future ranking or sales improvement.
If another new range uses a different, tested purchase path, the team can consider it independently. The point is to contain the known risk, not to postpone every merchandising decision until the whole site is perfect.
Agree what must pass before the products go live
Before the next import, agree which products are affected, who will fix the problem and what needs to pass. End the meeting with a decision about the range, not just another spreadsheet.
- Accurate product details: the catalogue owner has verified the required facts, variants, stock and offer.
- Working templates and imports: the relevant template, import and purchase-path checks pass.
- Discovery: the intended category links and product URL signals are in place.
- Exceptions: record what is still unresolved, which products it affects and who will deal with it.
- Follow-up: somebody records the release and reviews search and business evidence later.
A tool score or an indexing delay is not enough to decide whether to launch. Check the affected products and buying journey. If the store has little traffic, you may need those checks before you have enough sales data for a useful comparison.
My ecommerce SEO work starts with the range, the buying task and the constraint that deserves attention. If you are deciding between a larger catalogue and repairing the existing one, share the store URL, platform and intended launch. Tell me what you plan to launch and what is holding it up. We can start with that decision.
Common questions before expanding a product catalogue
Should I stop adding products until every SEO warning is fixed?
No. Check whether it is a real problem that will affect the new products. A broken selector or checkout step can justify holding that range. Optional polish or an unconfirmed warning should not automatically stop an otherwise sound launch.
Does a larger catalogue automatically bring more organic traffic?
No. New products need a useful buying purpose, accurate information and a discoverable, working page. More URLs can also repeat existing faults. Treat traffic and suitable purchases as measured outcomes, not an automatic consequence of SKU count.
Should every filtered category be indexable?
No blanket rule fits every store. Some combinations serve a useful search task; others only narrow an on-site list and can create excessive URLs. Review demand, content, URL behaviour and the platform before choosing crawl or indexing controls.
Is a product feed enough if the category navigation is weak?
A feed can help a platform discover product URLs, but it does not repair a shopper’s navigation or replace a useful category structure. Check the customer journey and the intended internal links as separate requirements.
What should I send for an initial ecommerce review?
Send the store URL, platform, planned range and the exact behaviour you are concerned about, with a few affected examples. Do not send passwords, payment details or private customer records through an enquiry form.
Sources
Google’s ecommerce navigation, URL and faceted-navigation guidance was checked on 17 October 2026. Use the checks above with the people responsible for your product data, website and launch.