It has been a couple years now since page speed has been an official Google ranking factor and one year since the launch of PageSpeed Online, the Web-based version of Google’s tool for analyzing site performance.
So I thought I’d check in on the PageSpeed scores for a selection of major news sites.
Page speed is not a major ranking factor (according to Google it impacts roughly 1 in 100 searches) but it does have a direct and indirect impact not only on SEO but also on social media activity and the overall user experience. For these reasons it is something that is worth paying attention to.
For this year’s comparison I started with the group of sites I looked at last year (see Google Page Speed Ratings One Year Later) and added in the majority of sites from News Organizations on Google+: Which Pages Get the Most Engagement?
For the sake of simplicity I pulled PageSpeed Online scores for the home page of each news site. PageSpeed scores are on a 0-100 scale with a higher number indicating better performance.
Here are the fastest – and the slowest – news sites in 2012:
Home Page | Page Speed Score April 2012 |
|
---|---|---|
1 | Yahoo! News | 96 |
2 | Mail Online (US home page) | 94 |
3 | Business Insider | 93 |
4 | TechCrunch | 91 |
T-5 | Breaking News | 90 |
T-5 | CBS News | 90 |
T-5 | CNET | 90 |
T-5 | The Economist | 90 |
T-5 | Mashable | 90 |
T-5 | The Washington Post | 90 |
T-11 | The Huffington Post | 89 |
T-11 | Newser | 89 |
T-13 | Le Monde | 88 |
T-13 | The Wall Street Journal | 88 |
T-13 | ReadWriteWeb | 88 |
T-13 | The Next Web | 88 |
T-17 | BBC News | 87 |
T-17 | Reuters | 87 |
T-17 | The Atlantic | 87 |
20 | The Daily Beast | 85 |
T-21 | Bloomberg | 84 |
T-21 | The Christian Science Monitor | 84 |
T-23 | Bloomberg Businessweek | 83 |
T-23 | Forbes | 83 |
T-23 | The Guardian (US home page) | 83 |
T-23 | msnbc.com | 83 |
T-24 | Time | 82 |
T-24 | USA Today | 82 |
T-25 | Chicago Tribute | 81 |
T-25 | Financial Times (US home page) | 81 |
T-25 | Slate | 81 |
32 | CNN | 80 |
T-33 | Boston.com | 79 |
T-33 | Fox News | 79 |
T-33 | The Fiscal Times | 79 |
T-33 | The Times (UK) | 79 |
37 | CNNMoney | 78 |
T-38 | ABC News | 77 |
T-38 | Los Angeles Times | 77 |
40 | Sky News | 76 |
T-41 | New York Magazine | 75 |
T-41 | The New York Times | 75 |
T-43 | PBS NewsHour | 70 |
T-43 | The Telegraph | 70 |
45 | The Verge | 68 |
46 | NPR | 67 |
47 | Al Jazeera English | 62 |
Average | 83 |
Yahoo! News is the winner, coming in with a very high score of 96. The UK’s Mail Online and Business Insider round out the top three with similarly high scores.
The top 10 all received scores of 90 or higher with six sites tied for fifth place.
It is not surprising to see sites like Business Insider, TechCrunch and Mashable in the top group (and The Huffington Post close at #11) but it was good to see some traditional news organizations make it in there too. Their sites sometimes lag behind due to a legacy CMS or other site architecture issues.
I didn’t expect to see NPR near the bottom of the table as their home page appears to load rather quickly in my own browser. But I checked it a few times and as you’ll see below 67 is actually an improvement over last year.
Whenever I do a search or social media comparison post for news sites The New York Times tends to be near the top of the group. But site speed is one area in which the competition has a leg up.
Note that for The Daily Mail, The Guardian and The Financial Times the PageSpeed tool was redirected to the URL for the US version of their home page. So the US URLs were used in the comparison.
One-Year Gains and Losses
Since I did a similar exercise with a smaller group of news sites last April I can also provide some data on changes in site speed over the past year.
Keep in mind that the content and particularly the advertisements on a home page are frequently changing so the PageSpeed scores will regularly fluctuate too.
But capturing scores at periodic intervals does provide a general sense of whether the pages are performing better or worse than previously.
Among the smaller sampling of news sites here are the home pages that have seen the greatest YoY improvement:
Home Page | Page Speed Score April 2012 | Page Speed Score April 2011 | Change | |
---|---|---|---|---|
1 | Reuters | 87 | 20 | 335.0% |
2 | Chicago Tribune | 81 | 59 | 37.3% |
3 | The Washington Post | 90 | 68 | 32.4% |
4 | msnbc.com | 83 | 67 | 23.9% |
5 | The Times (UK) | 79 | 64 | 23.4% |
6 | CNN | 80 | 68 | 17.7% |
7 | The Huffington Post | 89 | 77 | 15.6% |
8 | Los Angeles Times | 77 | 67 | 14.9% |
9 | The Christian Science Monitor | 84 | 74 | 13.5% |
10 | The Economist | 90 | 80 | 12.5% |
11 | Fox News | 79 | 71 | 11.3% |
12 | Forbes | 83 | 75 | 10.7% |
13 | NPR | 67 | 61 | 9.8% |
14 | Yahoo! News | 96 | 88 | 9.1% |
15 | Time | 82 | 76 | 7.9% |
16 | The Guardian | 83 | 77 | 7.8% |
17 | USA Today | 82 | 77 | 6.5% |
18 | The New York Times | 75 | 71 | 5.6% |
19 | The Wall Street Journal | 88 | 84 | 4.8% |
20 | ABC News | 77 | 74 | 4.1% |
21 | CBS News | 90 | 87 | 3.5% |
22 | Mail Online | 94 | 92 | 2.2% |
23 | BBC News | 87 | 86 | 1.2% |
24 | The Telegraph | 70 | 86 | -18.6% |
Average | 83 | 73 | 13.7% |
Reuters should probably be put into a special category since its home page score of 20 last year was bizarrely low (I recall checking that one several times to confirm it). They may have been having some kind of technical issue at the time that has since been resolved.
Among the rest of the group all but one site received a higher score this year. Two sites had 30%+ gains, two had 20%+ gains and the top 12 all improved by 10% or more.
Only The Telegraph saw a drop in home page PageSpeed; 18.6% is a considerable decline.
I’ve checked that one a few times but it continues to receive a 70. They were having an odd 404 issue with their home page yesterday which was resolved before I pulled these scores, but perhaps there are still lingering issues.
abe says
Adam, the news website isn’t MSNBC. That’s a TV network in New York. The news website is msnbc.com, in Redmond, Wash. Different companies.
Adam Sherk says
Thanks Abe. I didn’t include the .com in the tables above (will add it now) but it is the home page of msnbc.com (msnbc.msn.com) that was evaluated.
Pax Dickinson says
Thanks for this, I’ve been spending a LOT of time on improving our page speed and web performance at Business Insider, it’s pretty awesome to be recognized for it.
Now I just gotta get to #1. 🙂
Adam Sherk says
Nice work Pax, looks like it’s paying off.
Jack says
Good post but a little misleading because the high page speed score does NOT means a faster website.
Take Mail Online (Score: 94) and NPR (Score: 67) for example, if you record their load time using WebPageTest.com then you’ll see that NPR is a lot faster than Daily Mail. Here is a video of their page load comparison: http://www.webpagetest.org/video/view.php?id=120531_a3ea2c141f4ab0bce8ccf3818fec2b310b05ee75
Companies such as Gomez monitor news website portals which I think is a more accurate measure of news paper website page speeds thanks Google Page Speed.
See http://www.gomez.com/us-media-news-last-mile
http://www.gomez.com/uk-media-news-last-mile
http://www.gomez.com/france-media-news-last-mile etc.
Adam Sherk says
Thanks for the resources Jack.
Jarno van Driel says
I have to disagree with fastest/slowest data.
We recently published http://www.brixsgroep.nl Which reduces your data list to dust.
At this moment this site scores 100/100 in Pagespeed, 96/100 in Yslow and 100/100 at Pindom.
If you’re interested in how we did it, feel free to contact me.
Adam Sherk says
Glad your site is doing well Jarno. This comparison was specifically about news sites at a particular moment in time.
Jarno van Driel says
Sorry for not paying enough attention for the title of your post. Please feel free to remove my previous post, as I stand corrected.
Adam Sherk says
No problem Jarno, and good to hear that you’ve done well with your own site performance optimization.
Annmar says
you should had mentioned in year comparison if those sites changed their layout or not. And isn’t page load speed affected by the time of the day, I mean if you checked CNN at let’s say 1 pm and npr at 6 pm , then CNN would definitely be quicker?
Adam Sherk says
Annmar – thanks for your thoughts. Regarding the 2012 vs. 2011 figures, this is just a basic comparison based on a couple moments in time. I’m not in a position to know what technical or design changes each site may have made in the past year. I’m just showing rough differences in performance from one year to the next, regardless of the reason.
Regarding the time of day, all the sites were checked at roughly the same time of day each time.
Matthew Sweet says
Hi Adam,
Great article, I have managed to crack the 90 mark a week ago. It had been sitting miserably on 89 for a few months, but I finally managed to tweak my cache expires headers for images and css/js files and managed to increase a few points. Now it’s at 94 for the home page. I also figured out how to enable an origin-push cloudfront distribution on Amazon which significantly increased the overall site speed. So it’s nice to know that the little guy with an Amazon micro instance running Nginx & Varnish plus the excellent W3 Total Cache plugin and a little bit of research can get some trophy numbers up there with the big guys.
I followed Ewan Leith’s instructions:
http://www.ewanleith.com/blog/900/10-million-hits-a-day-with-wordpress-using-a-15-server
Adam Sherk says
Nice going Matthew, that’s impressive.
Vivek Moyal says
What is the average site speed for a normal website? These are the big guns only.
Adam Sherk says
Hi Vivek, this was a comparison specifically of major news sites so that’s the only data I’ve compiled. Since websites come in a lot of different shapes and sizes defining a “normal” site is a difficult thing to do. But you can do a relative comparison by running your site and some competitors (or similar types of sites) through PageSpeed Insights or another one like Pingdom Tools.
Tom says
It makes a lot of sense that Yahoo News would be on the top of the list. Long before Google came out with Page Speed, Yahoo was publishing information on factors that impact an HTML page’s speed. I think the first tool that ever gave an extensive analysis of page speed was YSlow which tested the different criteria that Yahoo developers identified as key points to focus on.
The first version of YSlow came out in late 2007. Two years before Google Page Speed and I’m pretty sure the information was publicly available before the plug in came out.
Yahoo essentially wrote the book on page speed.