Two Versions of One URL Can Become Two SEO Problems
Bracket URLs can split SEO signals when a site alternates between raw square brackets and their percent-encoded versions. A browser may reach the same-looking content through both forms, but Google can crawl them as separate URLs. That means links, canonical signals, sitemap entries, and crawl activity may stop agreeing about which address represents the page.
This is not a reason for every business owner to inspect every character in every URL before breakfast. It is a reason to check sites that generate filter, search, booking, product, or tracking URLs automatically. A small formatting inconsistency can create duplicate paths at scale, and scale is where a harmless oddity starts eating crawl time and muddying reports.
What Changed in Google's Bracket URL Guidance
Search Engine Journal reported guidance from Google's John Mueller that raw brackets and percent-encoded brackets are treated as different URLs for crawling purposes (https://www.searchenginejournal.com/google-crawls-raw-and-encoded-bracket-urls-separately/593765/). In practical terms, these two patterns should not be mixed across a site:
- A URL containing literal
[and]characters - A URL using
%5Band%5Dfor those same bracket characters
The fresh guidance does not say brackets trigger a penalty. It says inconsistent forms can be crawled separately. That distinction matters. The problem is not the punctuation having offended Google. The problem is a site presenting two addresses and then hoping every system quietly agrees they are one.
Google's URL structure documentation recommends using readable URL structures and percent-encoding non-ASCII characters according to established URL standards (https://developers.google.com/search/docs/crawling-indexing/url-structure). Its canonicalization guidance also explains that duplicate URLs can waste crawl activity and make performance measurement harder when signals are distributed across versions (https://developers.google.com/search/docs/crawling-indexing/consolidate-duplicate-urls).
Where Raw and Encoded Bracket URLs Come From
Most owners do not type square brackets into page addresses. Software does it for them.
Common sources include:
- Product filters that serialize selected options into a query string
- Site-search forms that package several values into one URL
- Booking systems that pass dates, services, or staff choices between steps
- Analytics and campaign tools that append structured parameters
- JavaScript frameworks that encode a browser address differently from server templates
- Plugins that build links one way while the XML sitemap exports another
A product filter might create a raw-bracket link in the page HTML while the browser, CDN, analytics tool, or sitemap records an encoded version. If both remain accessible, crawlers can discover both. Add several filters and thousands of products, and the site can produce a surprisingly large collection of near-duplicate addresses. Computers remain impressively fast at multiplying small mistakes.
Why Bracket URLs Affect SEO and Customer Acquisition
Duplicate URL forms rarely cause a dramatic overnight disappearance. The damage is usually quieter.
First, crawl activity can be divided between versions. Google may spend requests checking two addresses that return substantially the same content instead of reaching a new service, category, or product page sooner.
Second, internal links can split. Some templates may point to the raw form while other components point to the encoded form. Each version accumulates its own link history until canonicalization catches up, if it catches up cleanly.
Third, reporting becomes harder to trust. Search Console, analytics, server logs, and rank-tracking exports may record the variants separately. A page can look weaker simply because its impressions, clicks, visits, or bot requests are divided between two rows.
Fourth, cache and security behavior can differ. A CDN, web application firewall, application server, or routing library may normalize one form but not the other. One version might return the page while another redirects, errors, or receives a challenge. That inconsistency can turn a duplicate-content issue into an indexing issue.
For an owner, the outcome is not "bad URL aesthetics." It is slower discovery, unreliable measurement, and more technical work between a useful page and a customer who should have found it.
A Five-Step Bracket URL SEO Check
1. Find whether brackets appear at all
Search the XML sitemap, internal link exports, server logs, crawl reports, analytics landing pages, and Search Console page data for [, ], %5B, or %5D. Check uppercase and lowercase encoding because tools may display percent-encoded characters differently.
Do not assume a clean navigation menu means the issue is absent. Filter controls, scripts, old campaign links, and third-party widgets may generate URLs only after a user takes an action.
2. Compare raw and encoded responses
Request both forms and record the final URL, HTTP status, canonical tag, indexability, rendered content, and redirect chain. The test should answer a simple question: does the site consistently resolve both forms to one preferred address, or does it serve two indexable versions?
A browser showing the same page is not enough. Browsers are designed to be helpful. SEO diagnostics occasionally require being less trusting than a browser and slightly more tedious than a normal person.
3. Choose one internal format
Pick the form your platform can generate consistently and use it in navigation, filtered links, breadcrumbs, pagination, hreflang references, structured data URLs, XML sitemaps, feeds, and canonical tags. Google's fresh guidance recommends consistency across links, canonicals, and sitemaps.
The choice matters less than the consistency. Do not create a migration simply because encoded characters look technical. Change the system only when the current implementation exposes competing crawlable forms or creates routing problems.
4. Consolidate the alternate version
When both versions exist, use a server-side redirect from the alternate to the preferred URL where that behavior is technically safe. Keep a self-referencing canonical tag on the preferred page. Update internal links and sitemap entries so Google does not have to rely on the canonical tag as a cleanup crew forever.
Google lists redirects and rel="canonical" as strong canonicalization signals, while sitemap inclusion is weaker. Signals work best when they agree. A canonical tag pointing one way while internal links and sitemaps point another is not a strategy. It is a committee meeting.
5. Verify the fix in logs and search tools
After deployment, recrawl the affected templates and monitor server logs for requests to the alternate form. Check whether Google continues discovering both versions, whether redirects behave consistently, and whether Search Console consolidates reporting around the preferred URL.
Do not judge the fix from one manual request. Test several parameters, templates, devices, and edge cases. Encoding bugs enjoy hiding in the one filter combination nobody selected during quality assurance.
What Not to Do
Do not block the alternate form in robots.txt before Google can process a redirect or canonical relationship. Blocking crawl access can prevent a crawler from seeing the consolidation signal on the page.
Do not remove every query parameter by reflex. Filters, booking states, and campaign parameters can serve legitimate jobs. The goal is to control crawlable duplicates, not break useful site behavior in the name of tidiness.
Do not rely on canonical tags while continuing to generate mixed internal links. Canonicals are useful, but clean links reduce ambiguity for crawlers, analytics systems, caches, and customers.
Do not claim recovery the moment a redirect goes live. Crawlers need time to revisit old addresses and consolidate signals. Measure the response behavior first, then watch index coverage and performance data over time.
Make Every Important Page Have One Clear Address
Bracket URL SEO is a narrow technical issue with a familiar lesson: one page should have one consistently promoted address. When raw and encoded bracket forms circulate together, search crawlers can treat them separately even if people see the same content.
Check where the URLs are generated. Compare both responses. Choose one format. Align internal links, canonicals, sitemaps, redirects, and reporting around it. Then verify the change with a crawler and server logs instead of assuming the plugin handled everything with its usual serene confidence.
If duplicate paths, conflicting canonicals, or inconsistent crawler responses are hiding across the site, an AI Visibility Audit can separate the URL problem from broader indexing, access, and content issues so the next fix protects visibility instead of merely rearranging it.
