Improving on Robots Exclusion Protocol

Tuesday, June 03, 2008 at 9:00 AM



Web publishers often ask us how they can maximize their visibility on the web. Much of this has to do with search engine optimization -- making sure a publisher's content shows up on all the search engines.

However, there are some cases in which publishers need to communicate more information to search engines -- like the fact that they don't want certain content to appear in search results. And for that they use something called the Robots Exclusion Protocol (REP), which lets publishers control how search engines access their site: whether it's controlling the visibility of their content across their site (via robots.txt) or down to a much more granular level for individual pages (via META tags).

Since it was introduced in the early '90s, REP has become the de facto standard by which web publishers specify which parts of their site they want public and which parts they want to keep private. Today, millions of publishers use REP as an easy and efficient way to communicate with search engines. Its strength lies in its flexibility to evolve in parallel with the web, its universal implementation across major search engines and all major robots, and in the way it works for any publisher, no matter how large or small.

While REP is observed by virtually all search engines, we've never come together to detail how we each interpret different tags. Over the last couple of years, we have worked with Microsoft and Yahoo! to bring forward standards such as Sitemaps and offer additional tools for webmasters. Since the original announcement, we have, and will continue to, deliver further improvements based on what we are hearing from the community.

Today, in that same spirit of making the lives of webmasters simpler, we're releasing detailed documentation about how we implement REP. This will provide a common implementation for webmasters and make it easier for any publisher to know how their REP directives will be handled by three major search providers -- making REP more intuitive and friendly to even more publishers on the web.

So, without further ado...

Common REP Directives
The following list are all the major REP features currently implemented by Google, Microsoft, and Yahoo!. With each feature, you'll see what it does and how you should communicate it.

Each of these directives can be specified to be applicable for all crawlers or for specific crawlers by targeting them to specific user-agents, which is how any crawler identifies itself. Apart from the identification by user-agent, each of our crawlers also supports Reverse DNS based authentication to allow you to verify the identity of the crawler.

1. Robots.txt Directives
DIRECTIVE IMPACT USE CASES
Disallow Tells a crawler not to index your site -- your site's robots.txt file still needs to be crawled to find this directive, however disallowed pages will not be crawled 'No Crawl' page from a site. This directive in the default syntax prevents specific path(s) of a site from being crawled.
Allow Tells a crawler the specific pages on your site you want indexed so you can use this in combination with Disallow This is useful in particular in conjunction with Disallow clauses, where a large section of a site is disallowed except for a small section within it
$ Wildcard Support Tells a crawler to match everything from the end of a URL -- large number of directories without specifying specific pages 'No Crawl' files with specific patterns, for example, files with certain filetypes that always have a certain extension, say pdf
* Wildcard Support Tells a crawler to match a sequence of characters 'No Crawl' URLs with certain patterns, for example, disallow URLs with session ids or other extraneous parameters
Sitemaps Location Tells a crawler where it can find your Sitemaps Point to other locations where feeds exist to help crawlers find URLs on a site

2. HTML META Directives
DIRECTIVE IMPACT USE CASES
NOINDEX META Tag Tells a crawler not to index a given page Don't index the page. This allows pages that are crawled to be kept out of the index.
NOFOLLOW META Tag Tells a crawler not to follow a link to other content on a given page Prevent publicly writeable areas to be abused by spammers looking for link credit. By using NOFOLLOW you let the robot know that you are discounting all outgoing links from this page.
NOSNIPPET META Tag Tells a crawler not to display snippets in the search results for a given page Present no snippet for the page on Search Results
NOARCHIVE META Tag Tells a search engine not to show a "cached" link for a given page Do not make available to users a copy of the page from the Search Engine cache
NOODP META Tag Tells a crawler not to use a title and snippet from the Open Directory Project for a given page Do not use the ODP (Open Directory Project) title and snippet for this page


These directives are applicable for all forms of content. They can be placed in either the HTML of a page or in the HTTP header for non-HTML content, e.g., PDF, video, etc. using an X-Robots-Tag. You can read more about it here:X-Robots-Tag Post or in our series of posts about using robots and Meta Tags.

Other REP Directives
The directives listed above are used by Microsoft, Google and Yahoo!, but may not be implemented by all other search engines. In addition, the following directives are supported by Google, but are not supported by all three as are those above:

UNAVAILABLE_AFTER Meta Tag - Tells a crawler when a page should "expire", i.e., after which date it should not show up in search results.

NOIMAGEINDEX Meta Tag - Tells a crawler not to index images for a given page in search results.

NOTRANSLATE Meta Tag - Tells a crawler not to translate the content on a page into different languages for search results.


Going forward, we plan to continue to work together to ensure that as new uses of REP arise, we're able to make it as easy as possible for webmasters to use them. So stay tuned for more!

Learn more
You can find out more about robots.txt in our documentation and at Google's Webmaster help center, which contains lots of helpful information, including:We've also done several posts in our webmaster blog about robots.txt that you may find useful, such as:There is also a useful list of the bots used by the major search engines.

To see what our colleagues have to say, you can also check out the blog posts published by Yahoo! and Microsoft.

Comments:

Congratulations! You've discovered botcoins, a new currency for webmasters! You can use botcoins for your daily searches, even without typing anything! (100 botcoins = 1 free search). To redeem your botcoins, simply find a Google Search Team member in person, hand over your botcoin certificate, and request your query. The Googler will then make the search for you. Note: Googler may ignore you or choose to search for a different phrase or topic if they are in a bad mood. And, currently no mining opportunities are provided for botcoins.