Almost every site worth reading still publishes a feed. The link to it just got buried somewhere between the cookie banner and the footer, or removed from the page entirely while the feed itself kept running. If you have a specific site in mind and want its RSS feed URL, the trick is almost always one of five patterns. Once you know them you can find a feed in about thirty seconds.
Try the obvious URLs first
For most blogs and news sites the feed lives at one of a handful of paths bolted onto the homepage URL. In rough order of how often they work:
/feed/rss/feed.xml/atom.xml/index.xml/feed.json(for JSON Feed sites, mostly Micro.blog and a few Ghost installs)
Type the site URL in your browser, add one of those, hit enter. If you see a wall of XML, that is your feed URL. Copy the address bar.
/feed and /rss cover the bulk of the modern web because every WordPress install publishes them by default and a long list of platforms (Ghost, Hugo, Jekyll, Eleventy, Astro starters, Substack) ship the same convention. If /feed does not work, /rss usually does. If neither works, the next two paths catch most static sites.
When the URL guess does not work, view source
If the obvious paths return a 404, the feed is probably still there but at a non-standard URL. Almost every site that publishes a feed declares it in the page head with a tag that looks like this:
<link rel="alternate" type="application/rss+xml" title="Feed" href="https://example.com/path/to/feed.xml">
That is the RSS autodiscovery convention from the RSS Advisory Board. Browsers used to draw a little orange icon in the address bar when they saw it, then they stopped. The tag is still in the HTML on the vast majority of sites that publish a feed.
To find it, open the homepage, view the page source (Ctrl+U on Windows, Option+Cmd+U on Mac in most browsers), and search the source for rss+xml or atom+xml. The href on that tag is the feed URL. This works on sites where the feed link is not visible anywhere on the rendered page. I find feeds this way more often than any other method.
If view-source shows nothing matching rss+xml or atom+xml, the site genuinely does not publish a feed, and you are now in feed-generator territory (covered further down).
YouTube channel feeds
Every YouTube channel has a feed. YouTube does not link it from anywhere in the interface, which is why nobody knows.
The pattern is documented in YouTube's own push notifications guide:
https://www.youtube.com/feeds/videos.xml?channel_id=CHANNEL_ID
Replace CHANNEL_ID with the channel's actual ID, which starts with UC and is 24 characters long. To find a channel ID, open the channel page in YouTube, view source, and search for "channelId". The value next to it is what you want. (Custom handles like @mkbhd will not work as the ID. You need the UC... string.)
A working example: YouTube's own channel feed sits at https://www.youtube.com/feeds/videos.xml?channel_id=UCBR8-60-B28hp2BmDPdntcQ. Open it in a browser and you will see the last fifteen videos as Atom XML.
WordPress sites and what their feeds expose
Every WordPress site publishes a feed by default. The endpoints are predictable:
/feedor/feed/for the main site feed in RSS 2.0 format/feed/atom/for the same feed in Atom format/feed/rss2/for explicit RSS 2.0/comments/feed/for site-wide comments/category/news/feed/for one category/tag/javascript/feed/for one tag
The category and tag feeds are the underused ones. If you only want posts tagged interviews from a site you read, you can subscribe to the tag feed and skip everything else. WordPress generates these on demand, so the slug you see in the URL of a category page works as the feed slug too.
Google News (search-as-feed)
Google News publishes RSS for any search query. The base feed lives at https://news.google.com/rss and you can append a query to turn any topic into a feed. The pattern:
https://news.google.com/rss/search?q=YOUR+QUERY
A few notes that save time:
- Use
+between words, or URL-encode spaces.q=apple+earningsworks. - Add
&hl=en-US&gl=US&ceid=US:ento pin the results to US English. Other locales are documented at the same URL with the relevant codes. - The feed is rate limited if you hammer it. Once or twice an hour from a normal feed reader is fine.
This is the closest thing to a free, working "alert" feed for arbitrary topics on the modern web. Set up a Google News query feed for a competitor, a beat, or a person, and you will catch news the day it lands.
Common feeds people ask about
A short list, only because these come up over and over:
- TechCrunch:
https://techcrunch.com/feed/ - Hacker News front page:
https://news.ycombinator.com/rss - AP News, by topic:
https://apnews.com/index.rssand the per-topic variants linked from their site - Substack newsletters:
https://yourname.substack.com/feed - Medium publications and users:
https://medium.com/feed/@usernameorhttps://medium.com/feed/publication-name
Spotify is the awkward one. The Spotify app does not publish RSS for podcasts the way Apple Podcasts does. If you want a Spotify-distributed podcast as RSS, search the show name on a podcast index like Podcast Index or iTunes' search API. The podcast itself almost always has an RSS feed (that is how it got into Spotify in the first place); Spotify just hides it. A podcast that is genuinely Spotify-exclusive will not have one.
Reddit, X, and the platforms that almost have feeds
Reddit still publishes feeds, though they are easy to miss. Append .rss to almost any URL on reddit.com (or old.reddit.com, which is friendlier to fetchers): a subreddit, a user profile, a search, your own saved page. So https://www.reddit.com/r/programming/.rss and https://www.reddit.com/user/spez/.rss both work. Heavy fetching gets rate-limited, so use a sensible polling interval.
X (Twitter) does not publish public RSS for accounts. It used to. It does not now. There were third-party services that re-scraped X into RSS feeds for years; most of them stopped working when X cracked down on scraping. If somebody told you to use one and the URL no longer loads, that is why. There is no clean RSS path for X today. The workable substitute is to lean on X bookmarks plus a tool that reads them out (the Keep guide to exporting X bookmarks covers the working ones).
LinkedIn, Instagram, TikTok and Threads do not publish feeds. Do not waste an afternoon looking.
Or skip all of this and let Keep find the feed
If you save articles into Keep, you do not need to find the feed URL yourself. Paste the site's homepage into the new-source flow and Keep checks every common path, parses the page head for the autodiscovery tag, and falls back to a JSON Feed or h-feed if either is published. Whatever it finds, it adds as a feed source and starts saving new posts as they appear.
Keep supports RSS 2.0, Atom, JSON Feed, and h-feed in the same flow, so a Micro.blog feed and a WordPress feed both end up in the same library next to your articles, X bookmarks, and Kindle highlights. The Keep web feeds docs cover the setup and the backfill options if you want to grab the last 90 days of posts when you subscribe.
This is the part that tips it for me personally. I keep meaning to subscribe to half the writers I read and never get around to it because finding the URL is a tax. Pasting the homepage and letting the tool do the work removes the tax. Keep stores the full text of each post in your library, so the feed becomes a permanent personal archive, not just an inbox.
Plenty of dedicated readers do feed-discovery too. Feedly, Inoreader, NetNewsWire and a few others all parse the page head for autodiscovery tags. If you already use a reader, paste the homepage there and you will probably get the same result. The point is you do not have to view source unless you want to.
When a site does not publish a feed at all
Some sites genuinely do not publish a feed and never will. Marketing sites, single-page apps with no blog, sites where the publisher has actively turned RSS off. For those, a feed-generator service can scrape the page on a schedule and produce a feed from it.
Two worth knowing about:
- RSS.app generates feeds from a URL or social account. It runs on a freemium model with a 7-day free trial and paid tiers after that. It claims to handle around 45 million articles a month, so it is not a hobby project.
- Kill the Newsletter is the one I recommend most often, but only for newsletters. It gives you a unique email address; anything sent to it becomes an Atom feed entry. Free, open source, run by Leandro Facchinetti. Useful for paid newsletters you want to read in a feed reader instead of email.
A generated feed is always a second-best option compared to a real one. It depends on a third party staying online and on the source page staying scrapeable. If the publisher ever ships a real feed, switch to it.
How to create an RSS feed for your own site
If you run a site without a feed and want one, the answer depends on the platform:
- WordPress, Ghost, Substack, Tumblr, Blogger, Squarespace and most other blogging platforms publish a feed by default. You do not need to do anything except find its URL using the patterns above and add it to your site head as an autodiscovery tag if it is not already there.
- Static site generators (Hugo, Jekyll, Eleventy, Astro, Next.js) all have a feed plugin or template you can wire up in an afternoon. Search the docs for "RSS" and you will find it.
- Hand-rolled sites can publish a feed by writing the XML themselves. The RSS 2.0 spec from the RSS Advisory Board is short and readable. The Atom RFC is longer but better designed; pick whichever fits your stack. Most readers support both.
Add the autodiscovery <link rel="alternate" type="application/rss+xml"> tag to your site head once the feed exists. That is the part that makes feed readers (and tools like Keep) find it automatically.
If you read more about feeds than you publish, the plain-English explainer of RSS covers what the format is for and why it stopped being a fringe technology.