Comparing WordPress Hosting Options

WordPress is the leading Content Management System (CMS) in the world. It makes up over 64% of the websites on any CMS. In case you are wondering, that translates into 43% of all websites on the internet (W3 Techs). That’s a big deal.

Unlike some of its competitors, WordPress has the major benefit of being agnostic to where a site is hosted. Shopify, Wix, and Squarespace sites are all built on proprietary platforms hosted by the companies themselves. This means once you build a site on one of those platforms, you will remain there at least until you re-build your site elsewhere. In essence, you’re locked in.

This can be a very useful feature of the platform, but also brings with it a very important question for someone building on WordPress. Where do you actually want to host your WordPress site? There are certainly a number of options. Since WordPress is agnostic to the server it runs on, the options run the gamut from Cloud hosted servers all the way down to a computer hosted in your very own basement; a place which we’d not suggest, for the record.  Some maintain that is because computers get depressed in the dark. 😉

Most people are not in the business these days of maintaining their own hosting hardware. For WordPress, this even extends out to include the upkeep of much of the underlying software including the web server (i.e. Apache vs. nginx), the database (MySQL), and supporting web languages (PHP). That means most hosts offer what they might call “Managed WordPress Hosting” which basically means you can click a button to get WordPress installed. However, that really doesn’t tell the whole story about what you get. That makes the selection process even more complicated.

What Matters for Hosting

Speed

Speed is probably the biggest factor in the success of a website. If a page loads too slowly, visitors get frustrated and “bounce.” It also is a factor in Search Engine Optimization (SEO) because what search engine wants to send people to a a page that is going to frustrate them? Finally, the faster a page loads, the more momentum a visitor has for any excitement involved with an online purchase. This is why multiple studies have shown that even small increases in page speed can increase both conversions and order size significantly.

So, how do hosting choices impact website speed? It’s basically a question of available computing power compared to the demand of serving up the pages to visitors. The more pages being requested, the more computing power is needed. The demand can come from traffic to your site or maybe even to other sites if they are being hosted on the same computer. Hosting companies manage these resources in different ways to makes sure the get the right traffic and demands on the right computers so that they can provide the most speed without running more computers than they need.

The difficulty is that the demands aren’t always predictable, and spikes in traffic can have a big impact.  So, in addition to providing the right speed, hosts have to find ways to manage for those scenarios as well.

Uptime

It should be somewhat obvious that if the website isn’t working, there’s going to be a loss in money. It’s hard to get a true monitoring of how well a host does in keeping up its end of the bargain on uptime, though. Is the problem with an individual site? Is the problem with the server? Is the problem with a network issue between the visitor and the hosting server that has nothing to do with either the host or the website itself?

There are a number of small tests run by people with varying degrees of precision for determining the role the hosting service plays in uptime. Still, the numbers are useful if one is trying to make a reasoned decision about which host is less likely to have a problem.

Support and Features

This is the really difficult part of any comparison. Not every host provides the same kind of services in their offerings. We took a look at what we have found to be most important for our clients and did a review of what each managed WordPress host offered in their plans. The features were chosen because of how we historically have benefited from them in the past and how often we relied on them for keeping a site running or gracefully making changes along the way.

Comparing the Providers

We selected 12 major providers of managed WordPress hosting. Many are reviewed regularly by blogs all over the internet. We put together a quick chart for your reference along with our own scoring in several categories.

Here are the Hosts:

• A2 Hosting
• BlueHost
• Cloudways
• Dreamhost
• GoDaddy
• GreenGeeks
• Kinsta
• Liquid Web
• Pantheon
• SiteGround
• WP Engine (with and without the C2 performance options)
• WPX Hosting

Speed Results

Matthew Woodward put together a great set of tests to check out the speed of the 12 different providers. He even broke out two levels of service for the very popular WP Engine hosting platform. The article is definitely worth the read if you want to get into the nitty gritty details of how the tests were run as well as his take on the results. We differ slightly from his end assessment, but his methodology was sound and produced some very illustrative results.

As you can see from the summary table here, there is more than one way to measure server speed. Overall page load speed is good for assessing not just the server speed, but also of the page design, coding, and overall “leanness” of the webpage. It’s a great one to start with and he includes that stat from two separate tools to add some validity. He also includes Time to First Byte (TTFB), which is much more focused on the server itself. This tells you how quickly the server 1) receives the request, 2) processes the request, and then 3) begins transmitting data back to the user’s web browser. This measure is a great one to include since it really focuses on the infrastructure the host provides to serve web pages to your site visitors. If your server responds more slowly than your competitor’s to simply get the first byte out to the requesting web browser, it’ll be hard for it to catch up in order to load the page in full faster than for your competitor.

Finally, he does a load test. Basically, this is simulating a point when there is a big influx of traffic in a spike of activity. In other words, what happens when you go viral and suddenly the entire world is trying to buy your stuff? Will the server keep up? We feel this and the TTFB measurements were a little overshadowed in Woodward’s assessment and deserve a bigger influence on the outcome.

In the end, we combined all of these tests and averaged out how each host “placed” for each part of the speed assessment. from that we created an overall speed ranking that we fell is a pretty accurate sense of who is going to really provide speed no matter the traffic levels you might see. In truth, the two standouts here? Pantheon and WPX Hosting are pretty much twice as fast as their nearest competitors and really shine here. Something that surprised us, though, was the measurements made to compare WP Engine’s two plan levels. It seems the basic level plan outperformed the higher priced package. There could be any number of reasons why this happened, and the actual differences are small in absolute terms, but the overall impact put a big distance between them in our rankings.

Uptime Results

Determining a host’s uptime is often a challenge.  As we said earlier, it can be disputed between those running tests, the end users, and the hosts themselves as to what the real number is. That said, we looked around each of the provider’s sites to see what they had to say about their actual uptime has been historically. In some cases, we weren’t able to find an answer on the host’s website, so we searched for an answer on the internet until we could come up with a best guess.

Here, the big leader is Liquid Web. They boast a 99.9992% uptime. It may not seem like a big deal to add the extra “9” to the uptime number, but it has a big impact in the long run. The closest to them is a three-way tie between Pantheon, Cloudways, and SiteGround all with 99.99%. Please note: we allowed ties and ranked the results like sports competitions. We did this so we had a good way to summarize our overall rankings.

Support and Features Results