SEO for Rails

Posted by Vincent Daubry on February 23, 2015

SEO for Rails

My latest project is a web app that give developers their ranking on github : Gitub awards. I looked for advices on how to improve SEO, once again, so I decided to gather a little check list here, to save me some time next time.
These search engine optimization advices won’t be specific for a Rails app but i’ll focus on the tools in ruby that are available to help.

Step 1 : setup Google webmaster tools

It takes only minutes to add your website on google webmaster tools and you’ll be able to see how many pages are indexed, check that your sitemap was found, etc.

Just go to : google webmaster tools

And follow the instructions.

Step 2 : generate a sitemap

Generating a sitemap is as easy as generating a XML file, if you have only a few pages to index, just use Nokogiri. It would look like this :

base_url = ''
sitemap = do |xml|
  xml.urlset('xmlns' => '') {
    your_objects.find_each do |obj|
      xml.url {
        xml.loc base_url + an_url
        xml.lastmod a_modification_date
        xml.changefreq "monthly"
        xml.priority "0.7"

Once the sitemap is generated, just put it either in your public folder.

If your sitemap is more complex (thousands of pages for example) you can use the sitemap_generator gem. You’ll get some nice features out of the box : slice your sitemap by 50.000 links, gzip them, and create the sitemap index. It can also upload it to S3

Follow the gem readme, but here is a sample code :

require 'sitemap_generator'

SitemapGenerator::Sitemap.default_host = ''
SitemapGenerator::Sitemap.create do
  add '/home', :changefreq => 'daily', :priority => 0.9
  add '/contact_us', :changefreq => 'weekly'
SitemapGenerator::Sitemap.ping_search_engines # Not needed if you use the rake tasks

What if you have millions of pages to index ?

That’s exactly the situation i faced with Github awards : i have a page for each Github user, so potentially that’s more than 10 Millions pages to include in the sitemap. So should i generate a huge sitemap with all these pages ?

You should add to the sitemap pages that are not easily reachable by the crawler through link walking. That’s exactly the situation for Github awards : the vast majority of user pages are only accessible through the search, there is no way for a search engine crawler to find them on its own.

Another thing to consider is the relevancy of these pages for a search engine, i.e: what is the amount of unique content on these pages ? For example on a e-commerce website each product may have a lot of unique content : user comments, even product description might be unique, so potentially it’s valuable to index all these pages. For github awards i’m only generating stats for each user, there is basically nothing to index for google.

Some discussion on even suggest that this could hurt your ranking : “we’ve found it detrimental to give Google hundreds of pages to crawl on a sitemap that we don’t feel are important” (source)

So in this case i decided not to include user pages to the sitemap.

Step 3 : submit your sitemap

Now that your sitemap is generated, go to Google webmaster tools. In the site dashboard wheck that your sitemap was found by google crawler. If not, submit it.

Tips : If your have a very large sitemap, you should split it into several files, with a maximum of 50k urls each. Create a sitemap index describing all the sitemap files, this is an xml file looking like this :

<?xml version="1.0" encoding="UTF-8"?>
   <sitemapindex xmlns="">

You can learn more about sitemap index here

Step 4 : add meta tags

  • Title : Displayed in search engine results, important for SEO. Recommended format is : ‘Primary Keyword - Secondary Keyword - Brand Name’

  • Description : Not used for rankings, displayed in search results. Optimal length is about 150 characters. Use a description that will make the user click on this link

  • Keywords : Not important for SEO, not displayed in search results, better to not include them.

  • nofollow : Tell crawlers to index that page but don’t follow any links on that page. Can be combined with noindex in the same tag : “nofollow, noindex”

  • noindex : Tell crawlers not to index that page

  • canonical : Set the original url for the content (de-duplicate root domain and www)

To include meta tags in your views use the content_for helper :

In application.html.erb :

  <%= yield :head %>

In your view :

<%= content_for :head do %>
  <title> Github ranking | github-awards</title>
  <meta name="description" content="Discover your ranking on Github ! Find out what is your rank by language, in your city and in your country"/>
  <link rel="canonical" href=<%= welcome_url %> />
<% end %>

That’s it ! Any more ideas on how to improve a Ruby On Rails ranking on search engines ? If so Post a comment.