It’s satisfying to test your site using page speed performance tools like Google PageSpeed and WebPageTest.
They help you find the critical obstacles to a faster site. And they let you verify the improvements you make to your website load time.
But these are synthetic tests. They are not based on real user measurement.
What about the people who visit your site? What is their experience?
Does it vary by geography or by browser? Are some pages or sections of your site slower than others?
The Site Speed section of Google Analytics makes it easy for you to answer all these questions. And with Custom Reports you can investigate financial implications by comparing Site Speed metrics to your Conversion Rates.
Ilya Gregorik and Justin Cutroni from Google recently broadcast a webinar on how to Make the Web Fast by measuring performance with Google Analytics Site Speed reports. You can also get Ilya’s slides on Making the Web Faster.
They made a lot of practical suggestions which I will summarize in this article.
Does site speed matter to my business?
It’s legitimate to ask this question before spending any time on performance improvement. If there are no financial gains from site speed improvements, then it’s hardly a good investment of your energy and resources.
In the chart below, Ilya and Justin make the financial case. Multiple studies from companies like Akamai and Tagman have shown that conversion rates decline as page load times get longer.
You have more at stake than dissatisfied visitors who experience the pain of waiting for a page to load and the torturous uncertainty from not knowing how long the load will take. These visitors vote with their wallet and take their business elsewhere.
First, collect enough data.
Fortunately Google Analytics is already collecting lots of data to help you understand how visitors experience your website’s performance.
However, without a modification to your tracking code, GA is not collecting enough information to give you the performance insights that you can act upon.
Out of the box, Google Analytics defaults to a 5% sampling rate of pageviews.
So, the first order of the day is to modify your tracking code. You simply need to add a line directly above the _trackPageview line.
Google’s maximum sample is 10,000 pageviews per day. If you site fits in that traffic volume, then you can set the sampling rate at 100%. If the volume is higher, then you’ll have to adjust the sampling rate to less than 100.
Second, go beyond standard Site Speed reports. Look at conversion rates and financial outcomes.
The standard reports in the Content >> Site Speed section of Google Analytics show you a great deal about the speed of both your site and individual pages.
You can look at Technical metrics which show how your network and server are performing for each page on your site. Since averages can be deceiving, the Performance section looks at the distribution of load times. And finally, Map Overlay lets you examine page load times in different parts of the world.
What the standard reports in Site Speed don’t tell you is how Site Speed relates to Conversion Rates and financial outcomes.
Justin then shows how to easily create a Custom Report that connects your site speed metrics with conversion rates and financial outcomes. The three tabs in his report examine financial outcomes in the context of Geographic Data, Browser Data, and Landing Page Performance.
You can grab Cutroni’s SPEED: Business Impact of Speed Custom Report and explore for yourself.
So for example, you might look at conversions by country and see that in some countries in Asia you are getting lower conversion rates and that those are also the countries where you are seeing slow response times. This would be evidence to dig deeper and see what you can do to fix the problem. Perhaps another server located closer to those countries? Or a Content Development Network so your content files load more quickly regardless of the visitor’s location?
Ilya and Justin do a great job in this webinar. They explain the context for website performance. They teach the specifics of using GA to understand how your visitors around the world and across various browsers experience your site.
Even Ilya seemed to be truly amazed at the detail of network, server, and browser information that Google Analytics collects. This information on real user metrics has simply never been available to mainstream website owners.
Links from the webinar: