Juzer Burhani

Google Reader decided to simplify the way you add feeds. Previously, you had to click on "Add subscription" and either type the address of the site or enter some keywords. After entering some keywords like [Google blog] or [new scientist], Google Reader displayed a list of results mostly obtained using a standard web search. Basically, Google Reader took the list of web search results and filtered the web pages that didn't have feeds. This approach worked well in many cases, but not when the search results include pages from Wikipedia or YouTube, sites that have irrelevant feeds.

Now when you enter a navigational query in the "add subscription" box, Google Reader will directly subscribe to the top result. For example, if you type [new scientist], Google Reader will automatically subscribe you to New Scientist's feed, but that's not the case when enter a more general query like [scientist].


This works similarly to Google's Browse by Name, a feature available in Google Toolbar and Firefox that sends you directly to the top result for navigational queries. The problem is that Google Reader doesn't have a good method for ranking results and biases Google Web Search's ranking order with information about the popularity of a feed and other data. Here are some situations when Google Reader automatically subscribes to irrelevant feeds:

* search for [the economist] and Google Reader subscribes you to Wikipedia's feed
* search for [google os] and Google Reader subscribes you to Engadget
* search for [fake steve jobs] and Google Reader tries to auto-subscribe you to... Wikipedia's feed of recent changes

Until Google Reader fixes these poor results, I suggest to search for feeds from the directory page (Discover > Browse > Search for feeds). This way, you'll always get the list of search results.