How healthy are web agencies' own sitemaps?
Agencies build and maintain websites for a living, so their own sites should be a best case. We checked every URL in the sitemaps of 41 small web agencies to see how often the basics slip, and which ones slip most.
Key findings
- Half of the sitemaps send search engines to URLs they shouldn't index. 20 of 41 sites (49%) listed at least one URL that redirects, carries a noindex tag, declares a different canonical URL or is blocked by robots.txt. A sitemap should list only the final, indexable version of each page.
- Canonical mismatches are the most common sitemap problem. 37% of sites listed pages whose canonical tag points to a different URL than the one in the sitemap. 15% listed URLs that redirect.
- Metadata gaps are widespread. 41% of sites had pages with no meta description, 34% reused the same description across pages and 29% reused titles. 32% had pages with no H1.
- Only 11 of 41 sites (27%) were clean. The typical site had 2 different medium or high-severity issue types.
- Some sites can't be audited from their sitemap at all. Of 48 sites we could reach, 6 had no sitemap at
/sitemap.xmlor listed in robots.txt. One sitemap URL redirected to a page that returns 404.
A sitemap should list only the final, indexable version of each page.The rule most agency sitemaps broke
Why it matters
Search engines use a sitemap as a list of pages worth crawling. Every redirect, noindexed page or non-canonical URL on that list is a fetch that leads nowhere useful, and a mixed signal about which version of a page should rank. None of these problems is visible to a visitor, so they tend to accumulate unnoticed between redesigns.
Missing and duplicated titles and descriptions matter for a different reason: they are what search engines usually start from when writing a result. When they're missing or repeated, Google is more likely to write its own.
What we found, issue by issue
Share of the 41 sites with at least one affected page. Response times are excluded (see Method).
| Issue | Severity | Sites affected | Share of sites |
|---|---|---|---|
| Title over 60 characters | Guideline | 35 of 41 | |
| Description over 160 characters | Guideline | 30 of 41 | |
| Images missing alt text | Guideline | 26 of 41 | |
| Title under 30 characters | Guideline | 24 of 41 | |
| Multiple H1s | Guideline | 19 of 41 | |
| Missing meta description | Medium | 17 of 41 | |
| Low text volume (under 200 words) | Guideline | 17 of 41 | |
| Canonical points elsewhere | Medium | 15 of 41 | |
| Duplicate meta description | Medium | 14 of 41 | |
| Description under 70 characters | Guideline | 14 of 41 | |
| Missing H1 | Medium | 13 of 41 | |
| Duplicate title | Medium | 12 of 41 | |
| Redirecting URL in sitemap | Medium | 6 of 41 | |
| Blocked by robots.txt | High | 2 of 41 | |
| Noindex page in sitemap | High | 2 of 41 | |
| Missing title | High | 1 of 41 |
How to fix the common ones
- List final URLs only. Generate the sitemap from the same source as your canonical tags, so protocol, host and trailing slash always match.
- Keep noindexed pages out. Thank-you pages, tag archives and staging leftovers are the usual culprits. Most SEO plugins can exclude them.
- Write a description for every template. Start with the templates that produce the most pages: blog posts, services and case studies.
- Make titles unique. Duplicates usually come from a template that ignores the page name; fix the template, not the pages.
- Publish the sitemap where crawlers look. At
/sitemap.xml, or listed in robots.txt with aSitemap:line.
Method
Sample. 50 small web design and website-maintenance agencies, mostly in the United States, identified from public directories and search in September 2026. 2 sites blocked automated requests and 6 had no discoverable sitemap, leaving 42 audits. One was excluded because the site rate-limited our crawler, so its results weren't representative. The findings cover the remaining 41 sites.
Crawl. Every URL listed in each site's sitemap (median 61 URLs per site, 26,960 in total), fetched once with a polite, identified crawler that follows robots.txt. The same Check My Sitemap checks used in customer reports.
What we report. The share of sites with at least one affected URL. We don't pool URLs across sites, because a single large site would dominate. Server response times are excluded: they were measured from Ireland while crawling and say more about distance than about the sites. Temporary server errors are excluded too.
Limitations. A small, self-selected sample of small agencies; not representative of all agencies or all websites. Results reflect each site on the day it was crawled. No agency is named.
Free to cite with a link to this page. Data: Check My Sitemap, 28-29 September 2026.
Run the same checks on your site
The first 20 pages are free. Agencies can check client sites with prepaid credit and white-label reports.