Using a page speed optimization audit tool like GTmetrix or WebPageTest is a great first step towards assessing how your website loads for visitors. However, you may be making one of the common page speed testing mistakes that often cost website owners engagement, and ultimately, revenue.
In fact, we often get emails from frustrated website owners who aren’t getting consistent results. Often, their test results are good, but visitors are reporting slow load times (or they see a high bounce rate).
Many times, they are making one of the many common mistakes that we’ll detail below. The good news is that 9 times out of 10, it’s a simple oversight. Once you begin to view web performance in terms of how to provide the best experience to your visitors, you’ll find your results begin to match up with what other experience.
Mistake #1: Only Testing Desktop
This is one of the easiest page speed test mistakes to make, which is why we’re starting with it.
While Google’s PageSpeed Insights give you both a score for how your website performs on both desktop and mobile, most of the other audit services default to testing your site on a desktop.
That’s only half of the story!
As we covered in our Average Page Load Times for 2020 article, mobile web usage has surpassed desktop usage, and this will only get wider.
For this reason, it’s important that you test how your site loads on mobile devices with relevant connection speed. WebPageTest actually allows you to select specific Android and iOS devices so that you can see how you’re site loads on each platform.
While you’re testing on mobile make sure you also test your site on slower/older devices! Not every visitor (or even the majority) will be coming to you from the fastest flagship devices. It falls on you to make sure your site is performant for the average user using a lower-end device. You may be surprised at what you find.
Mistake #2: Only Testing the Default Location
Similar to making the mistake of only testing the default desktop device, it’s important to consider your testing location. WebPageTest defaults to Dulles, VA, while other services use their servers in Dallas, Chicago, etc.
While it may not be possible to test from exactly where you want, it’s important to get as close as possible. If your website promotes a small business local to a suburb in Chicago, testing your site from San Fransisco doesn’t really do you any good. Perhaps you run a business that serves many international clients. In that case, you need to be testing from the various countries that your visitors will be accessing from.
If you’re unsure of where your traffic is coming from, use a tool like Google Analytics and track it for a while. The ‘Audience’ stats there will show you were your visitors reside. After you have a better idea, then you can optimize your testing.
Sidenote – We just recently added a couple more testing locations, so you can now test your site from:
- Oregon, United States (US West)
- Virginia, United States (US East)
- Tokyo, Japan
- Sydney, Australia
- Mumbai, India
- Frankfurt, Germany (Europe Central)
- London, England (Europe West)
Mistake #3: Providing an Inaccessible URL
As you’re looking at the results of your audit, be sure you’re actually looking at the results of the page, not of an error page. This can occur when trying to test a page behind a firewall, or a website you’re building on a local machine (example: localhost:3000).
Most often, our readers see this issue when trying to test a site behind a log-in screen. That doesn’t mean you can’t test a password protected page though! You just need to implement a little scripting. Don’t worry. It’s easier than you think. In fact, we detailed the entire WebPageTest scripting process in an article recently.
Mistake #4: Only Testing Once
Last on our list of page speed testing mistakes – testing once and forgetting about it! It’s incredibly important to continuously be monitoring the performance of your site. We built MachMetrics specifically for this need.
You may think there’s no need to test your site frequently if you’re not making changes to it, but it’s a vital part of speed optimization. Below is an example of a site that had an unoptimized image added to it by an owner who forgot to check its impact on performance.
MachMetrics allows you to set alerts so you’ll know immediately if your site’s load time exceeds a set time. Because of this, the owner of this site caught his mistake, and all was fixed.
But what if the owner hadn’t been using a service that continuously monitors his site’s performance? His site could have been loading slowly for weeks, increasing his bounce rate, lowering his revenue, and alienating customers.
If you’re using a CMS like WordPress, there are many other things that can affect your load time without your knowledge. An automatic update to a plugin may bog your site down. Your hosting provider could have prolonged issues that you are unaware of.
And don’t get me started on third party scripts (such as ads or analytics pixels), which can single-handedly bloat your site beyond recognition. Those change more often than you think, and you have ZERO control over what they’re loading.
The point is, there are many constantly moving parts to hosting and displaying a website to the world. It’s incredibly important to monitor your site and test it daily!
Don’t Make These Page Speed Testing Mistakes
Now that you’re aware of the 4-page speed testing mistakes we see most often, you’re that much more prepared to get the most out of your site! As we’ve discussed before, consider that your competitors (even if you’re just competing for page views) are likely making at least one of these mistakes.
By making sure to check off each of these points as you do your own page speed optimization, you ensure that you have a leg up on their site.
Even if you’re not looking at it from a competitive point of view, you’re doing your visitors a huge favor by looking at your site from their point of view. Think of your target visitor. Are they coming from an office desktop computer or a mobile device? What part of the country (or world) are they coming from? Are they likely making a purchase, or just browsing for something interesting to read?
Keeping these points in mind will put you way ahead of the speed optimization game!