Google Page Speed Ratings One Year Later: News Sites Range from Up 27% to Down 73%

Google Page Speed logoThe launch of Google’s Page Speed Online (which allows users to get Page Speed scores without installing a browser extension) reminded me that it has been a year since I looked at the page load times of major news sites. So I thought I’d go back and see if there have been any performance changes in the past year.

Last year in 25 Major News Sites Ranked by Page Speed I ran the Page Speed tool on the home page of a selection of online news sites including newspapers, magazines, TV sites, wire services and other news organizations.

Today I checked each of those home pages again and ranked the sites by improvement. As a reminder Page Speed scores range from 0-100 with a higher score indicating better performance.

 SitePage Speed Score
April 2011
Page Speed Score
April 2010
Change
1economist.com806327.0%
2bbc.co.uk/news867219.4%
3newsweek.com827115.5%
4time.com766615.2%
5dailymail.co.uk928015.0%
6usatoday.com776714.9%
7foxnews.com716214.5%
8telegraph.co.uk867613.2%
9forbes.com756711.9%
10latimes.com67619.8%
11washingtonpost.com68629.7%
12cbsnews.com87808.8%
13news.yahoo.com88836.0%
14nytimes.com71692.9%
15abcnews.go.com74722.8%
16huffingtonpost.com77761.3%
17guardian.co.uk77770.0%
18csmonitor.com74740.0%
19msnbc.msn.com6769-2.9%
20online.wsj.com8488-4.5%
21aolnews.com7377-5.2%
22CNN.com6872-5.6%
23timesonline.co.uk6470-8.6%
24chicagotribune.com5966-10.6%
25npr.org6174-17.6%
26reuters.com2074-73.0%
Average73.271.82.3%

While the average improvement across all of the sites is only 2.3% several sites saw significant improvements, particularly The Economist at 27%. However eight sites had lower Page Speed scores than last year with seven of them showing decreases from 2.9% to 17.6%.

The Reuters home page was the poorest performer, dropping from 74 to 20 for a massive decrease of 73%. I checked it several times so the score is correct according to Google. Maybe there was just something odd on the Reuters home page on this particular day, or perhaps there is a larger problem that can be addressed (see below). I also checked a couple of articles on Reuters; while those pages fared better than the home page they still scored poorly (36 and 40).

Regardless of the cause, this shows why it is good to frequently monitor site performance because you never know when an issue will crop up. On news and content sites sudden, significant changes in page load time are often related to advertisements or third-party implementations.

So which sites are now the best performers? Here are the news sites ranked by current home page Site Speed score:

 SitePage Speed Score
April 2011
Page Speed Score
April 2010
Change
1dailymail.co.uk928015.0%
2news.yahoo.com88836.0%
3cbsnews.com87808.8%
4telegraph.co.uk867613.2%
bbc.co.uk/news867219.4%
6online.wsj.com8488-4.5%
7newsweek.com827115.5%
8economist.com806327.0%
9guardian.co.uk77770.0%
huffingtonpost.com77761.3%
usatoday.com776714.9%
12time.com766615.2%
13forbes.com756711.9%
14csmonitor.com74740.0%
abcnews.go.com74722.8%
16aolnews.com7377-5.2%
17nytimes.com71692.9%
foxnews.com716214.5%
19CNN.com6872-5.6%
washingtonpost.com68629.7%
21msnbc.msn.com6769-2.9%
latimes.com67619.8%
23timesonline.co.uk6470-8.6%
24npr.org6174-17.6%
25chicagotribune.com5966-10.6%
26reuters.com2074-73.0%

The UK’s Daily Mail was tied for third last year; this year it is the only site to break 90. (The 92 rating is actually for http://www.dailymail.co.uk/ushome/index.html which is the URL that the Google tool was redirected to). One year later Yahoo News and CBS News are still in the top three. Telegraph.co.uk went from tied for 7th in 2010 to being tied for 4th this year.

It is important to note that home page load times will vary depending on what is on the page on a particular day, so each site’s scores fluctuate a bit from day to day. Nonetheless these single-day snapshots allow for some basic analysis and comparison, both competitively and YoY.

Back to poor Reuters and its 20 Page Speed score, here are the improvements that Google suggested:

Reuters Google Page Speed suggestions

Comments

  1. says

    I remember when page load time first came out everyone was frantic to clean up their website. I think the factor on the algorithm is much less than we first expected. I personally think Google did this to light a match under the web community to at least make an attempt to clean up their websites so they move quicker in the search space.

  2. says

    You’re probably right Maciej. As I covered in my original post last year it’s certainly a small ranking factor and there are other things that should be given greater priority in a tech road map. But since improving site performance is good not only for SEO but also usability, why not do what you can to make it better? At a minimum you can check the scores across various page types on a site and see if there are some major issues and/or easy fixes that crop up.

  3. says

    Thanks for testing Page Speed Online. The scores between April 2010 and April 2011 should be different since:

    – we have added a lot of rules in the past year
    – we have changed some of the scoring as we’ve gained more experience on the impact of rules on performance
    – finally, ads are highly variable.

  4. says

    Thanks Richard, I appreciate the information. I noticed the Page Speed Online scores didn’t necessarily sync with the scores shown via the Firebug plug-in, so I wondered if the rules and/or scoring had changed.

    By the way I’m definitely a fan of the new online interface. I lot of the publishers we work with are paying more attention to site performance now, and this makes it easier for them to dig in to issues.

  5. says

    I have just changed all my social media badges/buttons to load asynchronously and that has made a huge difference in page speed. On some of my sites I have more than halved the load time so I’m hoping for improved user experience and a serps boost would be nice!

Leave a Reply

Your email address will not be published. Required fields are marked *


− 3 = one