For a long time Google have advised the web community that site speed is taken into account as a ranking factor. The move to displaying a publicly visible “slow” label may be the push site owners need to take action.
The slow label is currently being tested, it’s only visible to a certain proportion of Google users. As we saw with the recently introduced mobile friendly label, this could indicate the feature is on the horizon. As the slow label is in it’s testing phase, no official advice has been given on how to improve performance. However, Google have been offering advice and building tools for a while now.
There’s lots of good advice out there. What you can put into practice depends upon many factors including your websites specific setup, technical ability and of course your budget. Not everything will be easily achievable or feasible and that’s okay. Like anything, the law of diminishing returns comes into play. There are likely a number of quick wins you can identify from the benchmarking stage which will have a high impact upon performance.
Over the coming weeks we’ll be taking a look at some of the available tools and techniques site owners can implement to speed up their website for users. However, before implementing any changes, it’s important to first benchmark your site. Benchmarking gives you an idea of how your website is currently performing and what can be improved upon.
Benchmarking provides us with a quantifiable measurement which can be used to track and compare site performance over time.
Google themselves offer a number of useful tools for measuring and tracking performance. If you have the Google Analytics tracking code running on your website, you have likely already recorded some real world performance statistics. The Site Speed report in Google Analytics gives you some interesting numbers including overall average page load time as well as average load times broken down to a per page level. PageSpeed Insights, another Google tool, scores pages out of 100 and offers performance suggestions.
To know whether a change has had a positive or negative impact, you’ll want to benchmark your website before and after making the change. While Google’s PageSpeed Insights score can give an indication of overall performance, you’ll have to wait a fair while for Google Analytics to show real world average page load times. This is where tools such as GTmetrix and Pingdom’s Website Speed Test come in. You can run these services on demand and collect feedback as you’re making changes.
Google Analytics is a common website tracking service which you might already have running on your website. As well as tracking visitors and the pages they visit, Google Analytics can also track total page load time and the time taken for users to be able to see and interact with the page.
Site Speed reports can be found under the Behaviour heading and are a useful real world performance metric for measuring changes made to your website over a longer period of time. By default, performance tracking includes only 1% of website visitors and so you might find for small websites, the information is not complete enough to work from.
Google PageSpeed Insights hits your website twice, testing both the mobile and desktop versions of your site.
Google PageSpeed Insights scores your website out of 100. The results are clearly presented, a list of suggestions combined with detailed information on how to go about fixing performance issues make Google PageSpeed Insights an important tool, particularly if your aim is to avoid a potential slow label.
A score above 85 is considered good. It should be noted that the tool is under active development and so the rules are subject to change.
Interestingly, the mobile site is also tested for a number of common mobile usability issues, such as ensuring links have a large hit area and that fonts are of a legible size.
You should consider using additional tools alongside Google PageSpeed Insights as having access to a waterfall-style overview and exact page load times/page size can be useful when making tweaks and measuring their impact.
GTmetrix is a valuable tool for benchmarking and for finding potential bottlenecks in performance. GTmetrix combines both Google’s Page Speed and YSlow, a similar performance testing tool.
YSlow includes a number of tests over Google’s Page Speed with an emphasis on server side performance techniques which can sometimes be misleading. For example, YSlow does not always recognise your chosen CDN provider which can skew your overall results.
I’m including Pingdom’s Website Speed Test as it’s probably one of the fastest page speed testing tools. It has a simple interface and it gives you instant feedback, from the moment you hit the Test Now button, you’re given the page loading waterfall in real time. Once complete, you’re presented with a nice overview including a performance grade out of 100, number of requests, load time and total page size. Additional, detailed information can be found by digging through the various tabs.
The Performance Grade tab lists all the factors taken into account when calculating their performance grade. The list is ordered by where the most improvement can be made. Each item has a dropdown offering detailed information, including a link off to Google’s PageSpeed Insights documentation for further reading.
The Page Analysis tab includes a number of useful tidbits including a breakdown of size per content type. Again, helping you pinpoint areas where you can focus your optimisations.
Pingdom keeps a history of previous tests and charts the page load time, page size, request count and overall score over time which makes it a useful tool to use between changes while optimising your websites performance.