Webmaster tools site performance - reliable measurements

Question

I am measuring a few sites and use both webmaster tools and analytics to track various aspects. One of the aspects I try to measure and track is performance.

I am getting a much more reliable measurements in analytics, mostly thanks to this article. However, I am also curious about the stats that appear on the webmaster tools site performance section.

One thing I noticed is that the numbers don't exactly match. The second thing is that Webmaster Tools updates much less frequently. That's probably to be expected, but most of the sites I have will show latest information from a couple of days ago, but one particular site still displays data as updated on Apr 12, 2012 (today is May 05, 2012). In terms of accesses, it's not a particularly small site compared to others. So it's unlikely the slow updates are because it has too little or no traffic.

My questions, in case someone knows:

• How come one website only shows performance stats from nearly a month ago and doesn't get updated?
• What is the difference in measurements (method, accuracy, source of data etc) between analytics and webmaster tools?
• Is there any 'best-practice' to make sure google (and me) get the most reliable measurements?

UPDATE: To clarify, I am not asking what's the difference between webmaster tools and analytics, what are the best tools to test my site performance, or what's the best practice for SEO in general. I am quite aware of those. I am just wondering about the particular issues I experienced above and the specific aspects affecting performance measurements and update frequency in those tools.

Answer 1

Put simply, the sampling methods are different. The page speed in Google Analytics is measured via Javascript, while in Google Webmaster Tools it is measured from users who have the Google Toolbar installed. I believe the Google search engine uses the latter method (there is no data sharing between Analytics and search), but they may mix in their own algorithms too.

You will likely find the numbers in GWT slower than GA: this is mainly because the Google Toolbar is only available for Internet Explorer, which is typically slower than other browsers. One could also reason that IE users are typically less computer-savvy and so may have older computers or PCs clogged up with other apps/toolbars slowing down the computer in general.

Both methods are accurate in what they actually record - i.e. if a page takes 5 seconds to load, then both methods would record 5 seconds as the loading time - however as above the results in GWT may not be accurate to your site's users as a whole.

Regarding the updates, GWT just seems to update when it feels like it. Sometimes it will update every few days, sometimes it won't update for a month. Google recently announced they are removing site speed from GWT.

Answer 2

The Google Webmaster Tools is not a performance measurement tool, it's just a tool for better Google accessibility to your website. Control what's occurring about your website links, content,.... The Google Analytics also is not what's you want. It's a website tracker for visits stats, analyze them and make reports based on different standards for you and not any other thing.

There is no way to manual update data on Google Webmaster Tools, Google will update them periodically.

I think you need to use other performance measurement tools like Page Speed, YSlow, Pingdom, WebPageTest,...

Answer 3

There is no way to refresh stats on Webmaster Tools. You can test submitted sitemaps though, which is helpful.

Webmaster Tools is how Google see's your website as a crawler (bot). Analytics is the way Google see's your site as a user. Analytics is also Javascript based, so users who have JS turned off are removed from any stats. Webmaster Tools is all based on Google-bot crawling data.

Webmaster Tools will show you bad links pointing to you're site, and alert you to query parameters that are causing it to crawl multiple versions of the same page (duplicate content). Analytics will tell you information about the people who somehow ended up at those broken links, like how they got there.

Also, Analytics and Webmaster Tools can share data, so the most important information for day-to-day actionable data can be (and should be) pulled into Analytics reports.

This isn't necessarily "best practices", but it's my practices. I always use both Google Webmaster Tools, and Analytics when I'm setting up a new site. Check Webmaster Tools at the beginning to make sure sitemaps, RSS, any other feeds are working properly, make sure you're site is using rel=canonical properly with any parameters you might have. After a couple of weeks having no SEO issues, you can pretty much forget about Webmaster Tools. All of the other relevant data it has is pushed into Analytics.