CMS SEO Requirements: A Checklist Before You Choose
A CMS decides which search signals you can control at all. Before choosing a system or migrating, check seven things: whether content, links and metadata are in the HTML without running JavaScript, whether robots.txt and meta robots are editable, whether each page can have its own canonical URL, title and description, whether URLs that don't exist return 404 or 410, whether redirects are server-side 301s without chains, whether the sitemap carries an accurate last-modified date, and whether hreflang is reciprocal. Google's documentation covers every one of these points, and most can be checked on a test copy before signing a contract.
Why Choosing a CMS Is an SEO Decision
A content management system generates the HTML, status codes, redirects and sitemap, which is most of what a search engine crawler sees on first contact with a site. If the system won't let you set something, no amount of content optimization will help. Google itself notes this in its introduction to robots.txt: with some systems, site owners don't need to, or can't, edit robots.txt directly and only get a search settings page in the dashboard.
That is why the checklist is worth going through before choosing a system, not after a migration. The question is not whether a CMS is "SEO-friendly," but whether it lets you set a specific element for a specific page and whether the result ends up in the HTML the server sends. The points below apply to every kind of system: traditional, headless, static site generator or hosted site builder.
Rendering: Content in HTML, Not Only in JavaScript
Google processes JavaScript, but in its JavaScript SEO basics it says server-side rendering or pre-rendering is still a great idea, because it makes a site faster for users and crawlers, and not all bots can run JavaScript. The same documentation notes that for a non-200 status code, for example on an error page, rendering might be skipped.
Google describes dynamic rendering, serving crawlers a different version than people, as a workaround rather than a long-term solution. Instead it recommends server-side rendering, static rendering or hydration.
The test is simple: open the page source, not the browser's developer tools view, and check whether it contains the main content, links in plain <a href> elements, the title, the description and the canonical URL. The same documentation reminds readers that unique, descriptive titles and descriptions help users pick the right result, so the CMS must let you set them separately for every page.
robots.txt and Meta Robots Under the Owner's Control
The robots.txt file tells crawlers which URLs they may fetch. Google stresses that it is not a way to keep pages out of search: to keep a page out of Google, use noindex or password-protect it. A CMS therefore needs two separate things. The first is editing robots.txt, or at least a full view of what the system publishes in it. The second is the ability to set noindex for a single page or a whole page type, such as internal search results.
In a migration this point has another dimension. In its guide to moving a site with URL changes, Google notes that some owners block all crawling during development, and advises preparing the final robots.txt in advance, along with a list of URLs that need noindex removed when the move starts. The same guide says content that is not moving to the new site should return a 404 or 410.
A Canonical URL Set for Every Page
Google picks one URL among pages with the same or very similar content. In its documentation on specifying a canonical URL, it lists the signals in order of strength: redirects and rel="canonical" annotations are strong signals, while inclusion in a sitemap is a weak one. The signals stack, so it works best when they all point to the same URL.
That translates into requirements for the system. Every page should have a self-referencing canonical, written as a full URL, since Google recommends absolute rather than relative paths. Editors must be able to override it for a given page. A canonical should not point to a redirected page or one that returns an error, because the signals then contradict each other. Google also advises against using noindex to choose a canonical within a single site, because it blocks the page from Search entirely.
Status Codes: Real 404s and 410s
According to Google's documentation on HTTP status codes, Google doesn't index URLs that return a 4xx code, and already indexed URLs that start returning one are removed from the index. It treats 404 and 410 the same way. If a page returns 200 but the content looks like an error message or an empty page, Search Console reports a soft 404.
So the requirement for a CMS is that a URL that doesn't exist returns a real 404 or 410, not an error page with a 200 status and not a redirect to the homepage. The same applies to deleted posts, discontinued products and empty category pages. What this code means and when it needs fixing is covered in our glossary entry on the 404 error.
301 Redirects Without Chains
Google's documentation on redirects splits them into permanent and temporary. Permanent ones, 301 and 308, signal that the target should become canonical, and the new URL shows in search results. Temporary ones, 302, 303 and 307, are not such a signal, and the source URL usually stays in results. According to Google, server-side redirects have the highest chance of being interpreted correctly, and JavaScript redirects are worth using only when server-side or meta refresh redirects aren't possible.
Google's crawlers follow up to 10 redirect hops, but in the site move guide Google advises redirecting straight to the final destination, and if that's not possible, keeping the chain ideally no more than 3 and fewer than 5 redirects. Redirects should stay in place for as long as possible, generally at least a year.
A CMS therefore needs a server-side redirect manager with a choice of status code and bulk import of a full redirect map. One less obvious feature matters too: when a URL that is already a redirect target changes again, the old rules have to be repointed to the new target. Otherwise every change adds another link to the chain.
Sitemap and Hreflang from the Same Data
A sitemap should list the URLs you want shown in search results, which are the canonical URLs. In its documentation on building a sitemap, Google says it ignores <priority> and <changefreq> values and uses <lastmod> if it is consistently and verifiably accurate. The date should reflect the last significant change to the content, not every rebuild of the site.
A multilingual site adds hreflang. In its guidance on localized versions of a page, Google requires each version to list itself and all the others, using fully qualified URLs. If two pages don't both point to each other, the annotations are ignored. Our glossary entry on hreflang covers the attribute itself.
The most common source of errors is a plugin that adds hreflang for a version that doesn't exist or points to a redirected URL. That is why it works best when the canonical URL, the sitemap, hreflang and internal links all come from one source of URL data rather than four independent mechanisms.
How We Check a CMS and a Migration
A test copy before the decision. Before recommending a system, we check the HTML source of several page types: the homepage, a category, a post and an offer page. We look at whether the content, links, title, description and canonical URL are in the server response, and whether a non-existent URL returns a 404.
One redirect map. Every old URL gets one new destination, with no intermediate steps, and the rules are generated from the map file rather than from manual fixes in a dashboard. With every later change we repoint old rules to the new target.
A probe after every deployment. On this site, after each deployment that changes redirect rules, we run a probe. It takes the list of rules from the redirect file itself, so it never falls behind the changes, and it checks both ends of every rule: whether the source returns a 301 and whether the target returns a 200 rather than another redirect or a 404. It starts with positive controls, checking that the homepage returns 200 and a known rule returns 301. If those fail, the result is marked as a blind measurement, not as "zero errors." We don't treat the first run as conclusive, because a new configuration spreads across servers gradually, URL by URL. We count a mismatch as an error only when it repeats in a second run.
Submitting changed URLs. After a deployment we notify search engines only of new, changed and deleted URLs. How that works is covered in our article on IndexNow.
Summary
The SEO requirements for a CMS come down to control over the signals Google documents: content, links and metadata in the HTML, an editable robots.txt and noindex, a canonical URL for every page, real 404 and 410 codes, server-side 301 or 308 redirects without chains, a sitemap with an accurate last-modified date, and reciprocal hreflang. A system that won't let you set one of these elements will get in the way with every change, and most of all during a migration.
The cheapest way to check is on a test copy before the decision is made. We handle system selection, migration and implementation as part of our web development service. If you are planning a CMS change or a migration and want to run this checklist on your site, request a quote.