Salesforce Commerce Cloud SEO Audit for Faceted Navigation and Crawl Waste
Audit Salesforce Commerce Cloud for faceted navigation bloat, broken canonicals, and crawl waste with a repeatable checklist built for SFCC storefronts at scale.

Salesforce Commerce Cloud sites compete with themselves. Refinement definitions, URL rules, and Page Designer conventions each multiply a single category into thousands of indexable variants, and a generic ecommerce audit will miss most of them.
An SFCC-aware audit reads crawl waste as a platform artifact, not a template problem. The levers that fix it live in Business Manager and in refinement configuration, not in a robots.txt rewrite applied at the edge.
Commerce Cloud deployments tend to be enterprise-scale, which puts them squarely in Google's crawl-budget guidance band. An enterprise technical SEO audit that treats SFCC like any other PHP ecommerce stack is how retailers end up with millions of indexed parameter URLs and declining category rankings at the same time.
Where SFCC Generates Duplicate URLs
Four mechanics in Commerce Cloud reliably manufacture crawl traps. Search refinement URLs append price, color, and size combinations as parameters when they have not been promoted to SEO URL status. Sort rules stack onto those refinements, so a single category plus a sort order plus three facets is already a four-dimensional URL space. Legacy cgid category parameters survive URL rewrites during fallback and resolve the same PDP under multiple category paths. Page Designer locales duplicate content asset URLs across language codes without coordinating canonical targets.
Each is a product of the platform's own defaults. Google's guidance is explicit that these duplicates waste crawl time that would otherwise go to pages a retailer actually wants in the index.
Refinement Definitions Carry the Largest Tax
Commerce Cloud's refinement engine generates rewritten URLs only for refinements explicitly promoted to SEO URL status inside Business Manager. Everything else gets a query string. The ceiling matters: Salesforce caps enabled SEO URL refinements at 30 per site and 5 per category. Push past that and new refinements fall back to parameter URLs, which Googlebot will still discover and request.
The math behind the tax is straightforward. A single apparel category with 15 colors, 10 sizes, 3 materials, and 8 brands generates 3,600 facet combinations before pagination and sort rules are applied. Promoting four of those as SEO URL refinements uses a seventh of the site-wide ceiling for a single category.
SiteGenesis vs SFRA Changes What Googlebot Sees
Legacy SiteGenesis templates render a different DOM than Storefront Reference Architecture. SFRA emits a cleaner HTTP response and a refinement-grid handler that defaults to a rel="canonical" pointing back to the base category. That default solves the sort-and-filter duplication case, and it defeats the case where an enabled SEO URL refinement has a legitimate landing-page claim.
SiteGenesis sites often expose the same product tile across multiple categories, each with its own cgid parameter, so one PID resolves under several category paths. Migration from SiteGenesis to SFRA is the right long-term answer, but it inherits whatever duplicate-URL inventory existed on the legacy side unless the migration plan treats URL consolidation as a first-class workstream rather than a post-launch cleanup.
URL Rules and Canonical Logic in Business Manager
Business Manager exposes two levers that enterprise retailers under-use. URL Rules let you remove or rewrite parameters at the server before SFCC generates a link; this is where tracking IDs and legacy sort parameters should be killed, not in a Cloudflare worker further downstream. Catalog Page Meta Tag Rules let you template canonical tags, meta robots directives, and hreflang per category and per refinement combination.
Most SFCC instances ship with defaults that treat every refinement URL as self-canonical. That is rarely what you want. Pair the Meta Tag Rules work with a cannibalization pass across category and refinement combinations before you touch a single template — the consolidation playbook for duplicate landing pages applies cleanly to SFCC facet groups that currently compete for the same query.
One warning the Google documentation keeps repeating: noindex does not save crawl budget, because Google still has to request the page to read the directive. Parameter URLs that should never be crawled need a robots.txt disallow, not a noindex tag.
The Crawl Waste Audit: What to Actually Pull
Audit outputs matter more than audit frameworks. For an SFCC crawl waste review, pull 30 days of server logs segmented by verified Googlebot and classify each URL against Business Manager's refinement configuration:
- SEO URL refinements — valid, keep indexable, verify canonical self-reference
- Parameter refinements — redirect or disallow in robots.txt
- Legacy cgid duplicates — canonicalize to the SEO URL path
- Page Designer locale duplicates — hreflang or consolidate at the content asset layer
- Soft 404s from empty filter combinations — return a true 410 or a 404
The last bucket is the surprise most often. Every zero-result refinement page Googlebot encounters becomes a soft 404 that drains crawl budget until you return a hard status code. On a 500,000-SKU catalog with aggressive refinement, this category can account for more waste than action parameters and session IDs combined.
The commercial case for cleaning facets rarely needs to be defended on SEO grounds alone. Baymard's product list usability research found a steep performance difference between sites with a slightly optimized filter toolset and sites with mediocre filtering, and that difference sits on the metric every retailer already reports.
What Changes in the Next Twelve Months
Google's crawl guidance has been consistent for years. The answer engines have not. ChatGPT, Perplexity, and Claude render only a subset of the HTML Googlebot can parse, and refinement URLs that resolve slowly or require client-side JavaScript to populate are increasingly dropped from retrieval sets. The SFCC audit commissioned in 2026 should include a generative engine optimization layer: which category and refinement combinations are being cited by answer engines, which are being ignored because of render budget, and which would benefit from static server-rendered snapshots served from the same Business Manager configuration. The engineering lift is smaller than a replatform, and the compounding effect across a planning year is where the margin sits.