
Step 1: Is the Source Still Publishing?
Start with the obvious. Open the website and compare its newest article with the newest item in your reader. Then open the feed itself in the feed viewer, which fetches it fresh. Three outcomes are possible:
- The website has new posts and the feed has them too. The feed is fine; the problem is in your reader. Skip to step 4.
- The website has new posts, but the feed does not. The feed is broken or stale at the source. Continue with step 2.
- Neither has anything new. Nothing is wrong: the publisher simply has not posted. Our source profiles show the publishing rhythm of feeds in the directory, so you can see whether a silence is unusual.
Step 2: Did the Feed Address Change?
Sites move to a new domain, switch to HTTPS or replace their content system, and the old feed address quietly stops updating. Sometimes it redirects; sometimes it keeps serving the last version forever, because the old file was never removed.
Paste the website address, not the feed address, into the Feed Finder. It reads the autodiscovery tags of the current site. If it finds a different feed URL than the one you follow, unsubscribe and subscribe to the new one. The health check shows every redirect in the chain, which tells you whether the address has moved.
Step 3: Does the Server Answer Correctly?
Run the RSS feed health check on the feed URL and look at these lines:
| What you see | What it means | What to do |
|---|---|---|
| HTTP 404 or 410 | The feed no longer exists at this address | Find the new address with the Feed Finder |
| HTTP 403 or 429 | The server refuses or rate-limits automated requests | Try a reader that fetches from your own device, or ask the publisher to allow feed readers |
| HTTP 500 to 504 | The server or its CDN has a problem | Wait a few hours; if it persists, tell the publisher |
| A web page instead of a feed | The address now points to HTML, often a login or consent page | Use the Feed Finder; some sites need a different feed URL |
| Newest item is months old | The feed is served but no longer updated | The site likely moved its feed; look for a new one |
If the feed loads but your reader still says it cannot, check it in the validator. A single unescaped ampersand or a stray character before <?xml makes some readers reject the whole feed, while others silently skip it.
Step 4: Is Your Reader Using an Old Copy?
Readers do not fetch feeds continuously. They check on a schedule, and cloud readers often check unpopular feeds less often: a feed with three subscribers may be refreshed every few hours, a feed with ten thousand every few minutes. Most readers have a refresh or update now button for a single feed; use it before assuming something is broken.
Caches can also get in the way:
- A CDN that caches too long. If the publisher's CDN keeps the feed for hours, every reader sees the old version. The
Cache-Controlline in the health check shows how long caches may keep it. - A wrong conditional answer. Some servers answer
304 Not Modifiedeven after the feed changed, because the ETag or Last-Modified header does not change with the content. Readers then keep their old copy. This is a publisher-side bug; the fix is to base those headers on the newest item. - A browser extension or app cache. Remove the feed and add it again to force a clean fetch.
Step 5: Do the New Items Look New?
Readers decide what is new based on the item's ID and date. Two publisher mistakes make new posts invisible:
- Future or wrong dates. An item dated in the past, for example because the server's time zone is wrong, can land far down your list where you never see it. An item dated in the future may be hidden until that moment.
- Reused IDs. If a feed gives every item the same
guid, or reuses old ones, the reader thinks it already has the item and ignores it.
Both show up in the health check under Item IDs and Dates. The opposite problem, old posts that suddenly appear as new, has its own page: why feeds show duplicate posts.
If You Publish the Feed
When readers tell you your feed is not updating, check these in order: the feed URL in your <link rel="alternate"> tag still works; the feed is regenerated when you publish (static site generators and caching plugins sometimes skip it); your CDN or security plugin does not block or cache feed requests too long; and every new item gets a new, permanent guid and a correct date. The publisher best practices cover each of these.
Questions
How long does it take for a new post to appear in my reader?
It depends on two schedules: how often the publisher regenerates the feed, and how often your reader checks it. With a busy feed and a paid cloud reader that is usually minutes; with a small feed in a free reader it can be several hours. Feeds that support WebSub can arrive within seconds in readers that use it.
My feed works in one reader but not in another. Why?
Readers differ in how strict they are and where they fetch from. A cloud reader fetches from a data centre, which some sites block, while a desktop reader fetches from your own connection. Readers also differ in how they handle invalid XML. Run the feed through the validator and the health check; a 403 or 429 in the health check points to blocking, an error in the validator to a broken feed.