A baseball pitching scout points a radar gun toward the pitcher's mound. The pitcher lifts his leg, kicks, then hurls a pitch toward home plate. 73 miles per hour, according to radar. That's quite slow for a major league player.
A website developer opens a new Chrome browser tab to Google's PageSpeed Insights. The developer enters a website address. PageSpeed runs a series of tests. A yellow box surrounds the displayed results, indicating speed concerns. The tests return a speed rating of 69 (out of 100) for mobile devices, and a 74 for desktop browsers.
PageSpeed suggests that the website "should fix" two items:
- Prioritize visible content (on mobile devices)
- Enable compression
The developers should also "consider fixing" at least four other issues, PageSpeed suggests. For each recommended speed improvement, PageSpeed provides a "Show how to fix" link to learn more.
Given Google's famous focus on speed, you might be surprised to learn the website measured above: http://maps.google.com. The Google Maps team "should fix" a few things, according to Google's Page Speed.
And the pitcher that throws just 73 miles per hour? R.A. Dickey, a knuckleball thrower. For non-fans, the knuckleball is a pitch with minimal spin that is notoriously challenging to throw with control, hit, and catch. The speed of the pitch is inherent to what makes the pitch effective. Change the speed, and you change the nature of the pitch.
Google Maps likely receives a low score from PageSpeed because, like a knuckleball-throwing pitcher, the complexity of the task is inherent to the service. Google Maps dynamically detects your location, then depicts an aerial view of the surrounding area, all in little more than a second. Maps would load faster if it defaulted to an aerial view of a specific location, such as Google headquarters, but doing so would change the nature of the service.
That said, PageSpeed Insights identifies improvements most website managers should make (Figure A).
Google's PageSpeed Insights measures how fast your website loads and suggests improvements.
I recently used PageSpeed Insights to measure speeds for two different Wordpress sites. Based on identified items, I experimented with several plug-ins. Some had no impact, while others "broke" various aspects of the website. The following four plug-ins improved the PageSpeed scores.
1. PageSpeed suggestion: "Enable compression"
PageSpeed tests whether gzip compressed files are served. Where possible, a compressed file is conveyed, expanded, and then the contents displayed in the visitor's browser. The speed increase occurs because the internet connection often represents a significant "bottleneck" in the system, whereas compressing / uncompressing a file occurs quickly. I found that installing and enabling the Gzip Ninja Speed Compression plug-in boosted PageSpeed significantly.
2. PageSpeed suggestion: "Leverage browser caching"
When enabled, the browser displays locally stored items instead of downloading items from the web host. For previously visited sites, this speeds load times. But the browser needs a way to identify which items to use — and how long to use them before downloading new items. The Leverage Browser Caching plug-in improved PageSpeed results a couple of points.
3. PageSpeed suggestion: "Optimize images"
Lossless image compression preserves image quality, while reducing file size. Lossy image compression, in contrast, reduces file size at a cost of information loss, which may also degrade quality. For non-image experts, I recommend lossless compression. The EWWW Image Optimizer adds an "Optimize now!" option next to images in the Wordpress Media Library. The optimizer losslessly reduces image sizes, often producing file sizes 3% to 25% smaller.
The order that items load affects PageSpeed rankings. Depending on the templates and code on your site, some items take longer to process. Sometimes, changing the sequence speeds load times and increases PageSpeed scores. The Speed Booster Pack applies several techniques that affect the rendering sequence in an attempt to improve load times. I disabled some options to ensure that special features, such as image sliders, worked.
WP Super Cache and W3 Total Cache support many complex caching configurations, including the ability to leverage third-party content delivery networks (CDNs). CDNs improve speed by delivering content from a server close to the visitor's location. Caches serve stored content rather than creating it "on demand."
Your web host and domain name server (DNS) also play a significant role in your site's speed. If your site is on a shared server, your PageSpeed results vary wildly. For example, when another site on the server receives massive traffic, your site may be slowed.
Other measuring tools
In addition to PageSpeed, you might also run a couple of other website speed tests. The Pingdom Website Speed Test returns a "performance grade" along with load time data (in seconds). GTMetrix returns YSlow results, which provides analysis and recommendations from Yahoo!'s Exceptional Performance team.
The bottom line
PageSpeed, like pitch speed, is important. In most cases, faster is demonstrably better. With the majority of sites, a few customizations will produce significant PageSpeed improvements.
Complex sites, like Google Maps, have a more challenging task. Like R.A. Dickey, they need to deliver dynamic data with control over difficult-to-predict networks. Standard PageSpeed recommendations won't help improve such sites, just as a coach telling Dickey to "throw faster" won't make him a more effective pitcher.
How do you measure the speed of your website? Have you used PageSpeed Insights to improve your site's speed? If so, what was the impact? Share your experience in the discussion thread below.
Andy Wolber helps people understand and leverage technology for social impact. He resides in Ann Arbor, MI with his wife, Liz, and daughter, Katie.