Google PageSpeed Insights: The perfect page speed test?

Jan Hornung Last updated 23.01.2020
6 Min.
Bottleneck

The web server is purring, the hosting fits perfectly and yet WordPress is lame ? Time for an onpage analysis. Today I present you two tools, which reliably identify bottlenecks of your WordPress site and tell you where optimization potential is hidden and where you can start to speed up your site again. Conclusion: Google PageSpeed Insights is not the last word.

One thing first: The tests presented here, the Google PageSpeed Insights test and Yahoo's YSlow, do not measure the speed of your site ! We experience again and again that the results of these tests are misinterpreted. Instead of measuring the speed of your site , these tests compare a theoretical optimum, i.e. a predefined rule set, with the actual state of your site .

This means that sticking to achieving the highest possible Google or Yahoo score should be done with caution. This is because an absolute top score is not possible in all cases. Both tests only indicate what is theoretically possible, but not whether the recommended measures make sense for your project or not.

Example: If you have embedded external resources, e.g. a Google Maps map on your site , then Google indicates that your site should use browser caching. And this even if you have already activated this. Google does not distinguish between internal and external resources, because the service states quite dryly and without context that there are resources here that are not covered by the browser cache. However, Google does not take into account that this is not possible.

Google's Page Speed Test - a useful first starting point

If you want to find optimization potentials for your website, there is hardly any way around the Google tool. PageSpeed Insights tool. Not only is it a good measurement tool in itself, it is also built into many other page speed tests and thus one of the most important points of reference for onpage optimization.

In all likelihood, you've seen results like this before:

Result of the Page Speed Test for RAIDBOXES: Google gives here quite still potential for improvement, which is in this case however practically not convertible.
Result of the page speed test for raidboxes.io. Google still has room for improvement here, but in this case it is practically impossible to implement.

Google outputs two values: One is the absolute rating of the page potential. For this, Google gives a value between 1 and 100. A score of 100 means that the site - in Google's opinion - is optimised by all rules of the art. On the other hand, PageSpeed Insights marks optimization potentials with colors: Red stands for necessary optimizations, orange for recommended and green for already optimized parts of your site .

But beware: The Google scores are to be taken with a grain of salt. Because Google does not give a real speed value, but a speed potential measured by the Google criteria. For the interpretation of the data, this means: Google gives indications of measures that theoretically optimize page performance. However, Google does not include whether these measures can also be implemented.

For RAIDBOXES for example, Google PageSpeed Insights states that browser caching should be optimized. But that this is practically not possible, you realize only when you look closer at the point. Our own websites is thus a good example of the fact that one must not equate a Google rating in need of improvement with a relatively slow website. In part, you simply have to ignore Google's statements.

Google's Page Speed Test is a great tool for planning onpage optimization

Overall, the Google test allows conclusions to be drawn about potential in two central areas of your page performance: the page load time and the load time of the above-the-fold area, i.e. the area that is visible to the visitor without scrolling. This underlines that Google PageSpeed Insights is not a real performance test, but a test to identify optimization potential. It is therefore ideally suited for planning an onpage optimization.

However, to answer the question of how relevant and feasible certain optimization suggestions from Google are, one needs further data. In this context, the easiest relative value is certainly the technical feasibility. For this, of course, you need to know your site accordingly well. So if you can't estimate whether and how an optimization is feasible, you can't conclusively classify Google's score.

An alternative to Google's performance test, on the other hand, also allows rudimentary insights into data that determine how fast a site actually loads. In addition, YSlow, Yahoo's PageSpeed Insights counterpart, offers an alternative test set that allows you to capture and improve other performance-relevant aspects of your site .

The alternative: YSlow

YSlow also rates your site according to a predefined rule set. This set can either be used in a predefined form, for example by Yahoo! or the user defines it himself. YSlow is a browser extension and can analyze any site at the click of a button after installation.

Result of the page speed test YSlow for raidboxes.io
Result of the Yahoo Page Speed Test for raidboxes.io. Yahoo also sees potential for improvement here. The score is clearly different from Google's note. In the top right corner you can see the rule set used, in the top left corner the overall score and the score. The individual suggestions for improvement are listed below.

Similar to Google's Page Speed Test, YSlow offers very detailed instructions for some optimization steps, which come from the Yahoo Developersite . The evaluation system works similarly to Google: A school grade and traffic light system indicate which optimizations are absolutely necessary, recommended or already fulfilled. A numerical score is also given, indicating the fit of the page structure with the underlying rule set.

Yahoo's Page Speed test offers a little more than PageSpeed Insights

YSlow provides very concrete optimization suggestions, similar to PageSpeed Insights. In addition, you get a whole range of details, such as statistics for the cached and uncached site , the number of HTTP requests and the file size of the individual data packets.

Statistics function of the Page Speed Test YSlow
Statistics function of YSlow. In addition to the number of HTTP requests, the size of the data types is also broken down and compared for the cached and uncached site .

Yahoo's Page Speed test doesn't take a real time measurement either

Just like Google, YSlow does not provide any insight into the actual loading time of site , but refers to its performance potential. YSlow can also only find potential bottlenecks on site , but cannot put them in direct relation to the actual loading time.

Furthermore, it is questionable which potential the Yahoo rulesets unfold at Google, which SEO potential YSlow has compared to PageSpeed Insights . YSlow provides more information than Google PageSpeed Insights, but can quickly overwhelm beginners.

However, YSlow also provides initial data towards a comprehensive understanding of its own site , such as the size of individual data packets or the number of HTTP requests. In addition, YSlow offers an alternative rule set that complements Google's and can be defined by the user.

Conclusion: Google for the first analysis, YSlow for fine tuning

Depending on your own level of knowledge and which phase of onpage optimization you are in, it makes sense to use either Google PageSpeed Insights alone, or in combination with YSlow.

If you want to optimize your site quickly and SEO-efficiently, Google PageSpeed Insights is particularly suitable. The Page Speed Test offers simple analysis data with clear instructions without being primitive.

YSlow is more complex, both in terms of handling and in terms of the data that the page speed test provides. Nevertheless, it does not overwhelm the user and allows insights that Google denies you.

Google is therefore recommended for an initial measurement of site . Once the first optimization cycle has been completed, you can use YSlow to uncover further potential, such as an accumulation of 404 errors. Google PageSpeed Insights is therefore not the perfect test for onpage optimization. Not only can YSlow provide even more data, but also the preparation of an onpage optimization without a real time measurement is only possible to a limited extent.

Because as already mentioned several times, both tools ignore the actually decisive factor: the time and thus a relatable speed of your site . This means that after optimization, the usefulness of such tests ends. Because they can only reflect whether their suggestions for performance improvement were carried out or not. However, they cannot indicate what influence these improvements had on the page performance.

A whole range of tools allow you to gain comprehensive insight into the loading time of your site . In the next article in this series, we will introduce and explain some of these tools.

How important are the two tools presented in your everyday work? Have you already gathered extensive experience with YSlow? Do you know of any other good alternatives? We are looking forward to your comments!

RAIDBOXER from the very start and our Head of Support. He loves talking about PageSpeed and website performance at BarCamps and WordCamps. The best way to bribe him is with an espresso – or a Bavarian pretzel.

Comments on this article

Post a comment

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