Search Console is full of parameter copies
Color, size, and sort stacks are indexed. Money categories are not. We sample the pattern. We rank the waste. We do not start a filter plugin tutorial in the report.
Ecommerce SEO audit
Ecommerce SEO Audit to Name Catalog Waste, Then Stop
An ecommerce SEO audit is a time-boxed look at one store catalog. We sample products, categories, variants, and filters. We rank what is wasting crawl. We score the templates that have to sell. Then we stop. This is not a website audit, not a platform tutorial, and not a quiet path into a retainer.
- Sample
- Waste
- Score
- Stop
Position
We sample the catalog. Then we stop.
A store is not one homepage with a grid. It is products, categories, variants, filters, search, and seasonal collections, all making public URLs. Search intent has to be true on the money templates. Most filter mixes are not documents. This audit exists to say which is which, with evidence, on a date.
A website, not a catalog
Marketing sites, SaaS product sites, and publishers belong on a website SEO audit. Glancing at a store homepage is not that snapshot either. If the unknown is the whole site, not the SKU tree, open that page.
Open SEO auditThe catalog program after the report
Ecommerce SEO writes which URL classes may index and which modules merchandising can keep. This audit is the dated sample that can send you there. It is not that retainer.
Open ecommerce SEOLiving inside one platform
Shopify, WooCommerce, Magento, and BigCommerce each have a page when the blocker is that admin. A catalog waste list can name the platform. It will not become a theme or rewrite tutorial.
Open Shopify SEOThe Shopping campaign
Shopping ads own titles, GTINs, and disapprovals in the auction. We will flag when HTML and the feed disagree. We will not build campaign structure here.
Open shopping adsA website SEO audit stays on marketing sites and publishers. Ecommerce SEO as a program writes the permit list merchandising keeps after this report. Shopify, Woo, Magento, and BigCommerce each own live platform work. Shopping ads own the feed in the auction. We will not fold those jobs into one store health check.
We will not index every SKU because the feed is large. We will not sell a ranking date. We will not stay on because the waste list was long. If we cannot fetch product HTML as a crawler would, that gap stays a gap.
When this fits
When crawl is busy on URLs you do not sell
This audit fits when the catalog is the unknown: too many product URLs, thin category copy, filter waste, or templates that all say the same paragraph. If you needed a website snapshot, a live platform pass, or a catalog retainer, start there instead.
Every product page shares one manufacturer block
Specs you sell are missing. Keyword targeting is not stuffing the same head term into twenty thousand titles. The audit scores whether a template can even hold a unique job.
Out-of-stock and seasonal URLs still index
Gift collections 404 in January. Dead SKUs still appear in the sitemap. We need a retire rule in the findings, not a hope that Google forgets.
The feed and the page disagree
Merchant Center says in stock. The HTML says otherwise. That is a catalog bug in this snapshot. It is not a reason to buy Shopping management on this URL.
Process
Sample, waste, score, stop
Four jobs. Each one ends with a note merchandising can refuse or fund. We do not start an ecommerce retainer in the last step. The point is a dated catalog picture, then a close.
-
Sample
We pull URL classes from the live storefront and Search Console: products, categories, variants, filters, search, and leftovers. We do not crawl every SKU if a pattern is already clear. Keyword targeting is judged on money templates, not on a homepage.
Outcome A dated sample of catalog URL classes, including what we did not fetch.
-
Waste
We mark copies: parameter stacks, thin variants, duplicate category paths, leftover seasonal URLs. Topical relevance is the test for a filter landing page: real shopping job, or a convenience mix? Keyword density on a grid intro is not a catalog score.
Outcome A waste list with evidence, ranked by crawl cost and money risk.
-
Score
We score product and category templates for unique blocks merchandising can keep. Semantic keywords belong only when they are still the same shopping job. Schema is checked as facts that match the HTML, not as a campaign.
Outcome Template scores, plus what done looks like on one example SKU and one category.
-
Stop
We hand a sequence and we close. Shipping the permit list is ecommerce SEO. A platform leak is that platform page. Fetch after deploys is technical SEO. An audit that quietly becomes weekly work has failed this step.
Outcome A closed snapshot, not an open catalog retainer.
Deliverables
What you can refuse or fund after the sample
You leave with a class sample, a waste list, template scores, and a stop note. You do not leave with a Shopify theme pass, a Shopping campaign, or a ranking date.
Class sample
Products, categories, variants, filters, and leftovers we actually fetched. Dated. Last season's crawl is not this sample.
Waste list
URL patterns that should not be documents, with evidence. Unranked rows are how catalogs sit unread.
Template scores
Product and category templates: unique blocks, missing specs, schema that matches HTML. One example SKU. One example category.
Retire rules
Out-of-stock, seasonal, and duplicate paths. Redirect, noindex, or keep. After coverage drops is too late to invent the rule.
Stop note
Which page should own shipping: ecommerce SEO, a platform URL, technical SEO, or shopping ads. The audit does not own shipping. A variant that only changes a swatch is scored as a parameter or as a SKU. We do not invent a new category for every color.
Benefits
What a catalog dump cannot give you
Crawl waste has a name
Filter copies and thin variants are listed. Money categories can get crawl back once someone ships the list.
Templates are scored, not praised
A unique spec is either in the HTML or it is not. Manufacturer paste is named.
The snapshot has an end date
Diagnosis closes. The next buy is a different page, or there is no next buy. A leftover search page that ranks for a product you no longer sell is waste, not a ranking win.
Platform leaks point at the right admin
A Shopify app, a Woo plugin, or a Magento rewrite is named. This PDF does not become that tutorial.
Methodology
Who can freeze merchandising for the sample week
This works when someone can say which categories are real, and someone can leave the catalog still for a few days. If both are missing, we are touring grids. We are not auditing a catalog.
We start with the money categories, the filter plugins already on, and Search Console. If any of those is missing, we name it first.
We fetch published product and category HTML. We do not trust a logged-in preview. Does this category match a real assortment? Is a variant a document, or only a swatch?
You may rename the roles below. You may not leave the sample implied. An implied full-catalog crawl is how the report never ships.
- Catalog audit lead SEOConsultants.ai writes the sample, the waste list, and the stop note. We do not become your merchandising team because the SKU count is large.
- Whoever owns store admin Can confirm templates and filters, and can freeze a launch for the sample window. A locked agency seat with no editor is a finding.
- Merchandising Names real categories and whether a collection may retire. A grid nobody will maintain is a candidate to close.
- Whoever owns the feed Helps when HTML and Merchant Center disagree. They are not asked to run Shopping from this audit.
- Leadership Funds the next page or refuses it. An unread waste list is not a failure of this snapshot. A ranking promise would be.
First working session
Start with one money category and one noisy filter
The first hour is a category you care about, plus the filter pattern filling Search Console. It is not a tour of every SKU.
We pick a money category. We list the URL classes under it. Products. Variants. Filters. Leftovers. If two live categories share one unique paragraph, we score that as a template defect before we touch titles. If the real question is a website, not a catalog, that is a website SEO audit. If the real question is Shopify admin, that is Shopify SEO.
Then we fetch published HTML and a Search Console sample by pattern. The extra rows are usually the job. A theme customizer that looks finished while you are logged in is not evidence.
We leave with a first waste slice or a stop. A slice names the pattern, the evidence, and who can freeze launches. A stop names what is missing: Search Console, a merch owner, or a store that is actually a brochure site. Both are outcomes. A shopping list of SEO apps is not.
AI search
A catalog PDF is not an AI Overview
Honest product HTML in the sample can support later extraction. This audit cannot sell AI Overviews, a shopping-chat mention, or LLM visibility as a score.
People ask whether cleaning filter waste will land SKUs in Google AI Overviews (AIO). If the catalog is manufacturer paste, or if mixes have exploded, generative search has nothing honest to lift, or too many copies. That is ours to name. If the kept pages already tell the truth and models still invent SKUs, that is an AI search audit.
The money template has to print the spec
AI search needs a page it can fetch. A field that never reaches HTML does not help. We do not sell Overview presence.
The name on the product should match the SKU you sell
We check brand and product names in the published HTML. That is entity optimization as a catalog input. It is not a Knowledge Graph project from this snapshot.
Measuring model answers is another URL
LLM visibility needs a sentence a model can quote, then measurement on LLM SEO. A waste list can make a SKU look real. It does not sample answers.
Schema
Markup is checked on the sample, not sold as the job
JSON-LD helps a machine read a product the page already states. This audit can flag two Product blocks or missing facts. It is not a schema campaign.
Facts must match the HTML
If the template and an app both inject Product markup, we name the conflict. Valid types still need price and stock that match the page.
No fake FAQ above a category grid
We refuse FAQ blocks that exist only to host more terms. A real fit question can live on the product template.
Extra fields do not buy a shopping mention
Markup will not force generative search to cite the SKU. A paid Shopping placement is a different job.
Who we work with
The person who can freeze a collection launch
Someone has to leave the catalog still for the sample. We work with that person. We do not replace them with a second merchandising retainer.
Store admin owner
Confirms templates and filters. Helps us fetch what a crawler gets. Rechecks nothing until the sample date is set.
Merchandising
Names real categories and retire rules. Holds the waste list when a campaign wants every mix indexed.
Feed or ads already in motion
Can keep going once HTML and feed facts are named. If they try to run this as Merchant Center only, two jobs collide.
Questions
Frequently asked questions
How is this different from a website SEO audit?
A website SEO audit diagnoses one marketing site, then it stops. This audit diagnoses a catalog: products, categories, variants, filters, and out-of-stock URLs. If the risk is a homepage and a few landing pages, stay on the website audit. If the risk is tens of thousands of SKUs, stay here.
How is this different from ecommerce SEO as a program?
Ecommerce SEO writes the rules merchandising has to keep after the report: which URL classes may index, which modules stay unique. This page is a dated sample, a waste list, and a stop. We will not quietly become the catalog retainer because the PDF was long.
Do you need Shopify, Woo, or Magento admin for this?
Admin helps when templates and filters are the issue. Search Console and a crawl of the live storefront can still show waste. If the blocker is a platform setting, we name that platform page. We do not turn this snapshot into a Shopify theme pass.
Will the audit include Merchant Center and Shopping ads?
Only where the HTML and the feed disagree on the same SKU. Campaign structure, bids, and disapprovals in the auction stay on shopping ads. A stock mismatch on the product page is a catalog finding here.
Will a cleaner catalog get us into AI Overviews?
No. Honest product HTML can help AI search later. This snapshot does not buy a place in Google AI Overviews (AIO), a generative-search mention, or an LLM visibility score. If the catalog already tells the truth and models still invent SKUs, that work sits on an AI search audit.
One catalog, then a stop
Are filter URLs indexed while money categories starve?
Send the storefront, Search Console, and a merchandising tree if you have one. We will say if a catalog snapshot fits, or whether a website audit, a platform page, or shopping ads should go first.
- One catalog sample
- Waste ranked with evidence
- Templates scored
- Then a stop
Sample, waste, score, stop. Not a ranking date.