Service level and hardware considerations for WordPress hosting

Disclosure: As of the time of this writing, although I am currently a GoDaddy employee, the views expressed in this post are my personal opinion.

As a follow-up to a recent deep-dive discussion on “understanding the web hosting options for your WordPress site (Download Presentation)” that Shayda and I led as a part of The Austin WordPress Meetup Group, I threw together some more detail around choosing the appropriate service level and hardware for hosting your WordPress site.

So many choices

There are a lot of choices for hosting in general, and WordPress hosting specifically. When thinking about hosting, I like to break it down into hardware and service. Note: not all hosts are created equally. Do you research and check with your host(s). The service level and hardware features listed here are best-case.

Service level

Service is how much involvement the host has with managing the your website hosting environment (aside from keeping underlying hardware / software running properly):

[table id=3 /]

Hosting hardware

Hardware is the hardware configuration you’re using to host your website:

[table id=2 /]

During the meetup, there was a question about unbiased hosting reviews. Check out Cloud Spectator (which I consider to be a neutral source), a company that tests performance of the top managed WordPress hosts monthly.

See something you’d change? Let me know in the comments.

10 thoughts on “Service level and hardware considerations for WordPress hosting”

  1. Mendel, if you like CloudSpectator, you might be interested in reading: http://reviewsignal.com/blog/2014/11/03/wordpress-hosting-performance-benchmarks-november-2014/

    Disclaimer: I did the testing and wrote it. But I tested the performance a lot more thoroughly.

    I also keep review data about most of the big companies (http://reviewsignal.com/webhosting/compare) and many of the WordPress ones are under Cloud tab as well. I know your reaction should be to mistrust them and that’s good. Hopefully my explanation and your judgement after actually looking at the results can convince you. The way it works is by collecting and analyzing hundreds of thousands of opinions people have shared on Twitter. It’s been entirely automated and has up to four years of data on some companies. The goal is creating a transparent way to get data and publish it so that it’s verifiable. If you’re really curious, click the `How It Works` link at the bottom which explains the algorithms and classification method used.

    1. Thanks Kevin. The biggest question I have related to your algorithm is for what range does it display data? Is it a rolling week? Month? Year? Good hosts get bad and bad hosts get good. Curious how to account for that.

      1. It’s all the data. There is currently no decay in the algorithm. It’s one of those things I keep putting off because I’m not sure what the fairest way to treat it is. I do have a chart (trend tab) which shows the past years worth at monthly intervals. I think sample sizes smaller than a month for most hosts are likely to be statistically not meaningful (unless it’s a mega host like GoDaddy).

        1. IMO, the fairest way is to let the consumer choose. Why not make the list filterable by week, mo, quarter, year, all-time? There are real-life parallels such as stock tickers. I love the work you’ve done, just think it makes sense to let the customer choose what threshold they’d like. It’s a tool for the customer after-all. 🙂

          1. Yes and no. I agree it would be nice to build some more customizable access to the data and better charting tools. But what matters absolutely the most is the default setting because most users don’t bother looking or changing from it. So whatever I decide is the default becomes the de-facto measure that people would use. That’s the decision that actually concerns me. I’m trying to build a tool to help consumers who are really being screwed over by just about every source out there on getting good information about web hosting. For those customers, having all the fancy tools to help visualize and chart differently doesn’t do much good, you would be amazed how many people just click through on whatever is at the top. That’s why all those fake review sites do it like that. They know.

            So yes, I agree adding more visualization tools would be nice. But I still need a default which is what I expect 90%+ of people to use that needs to be fair.

            Bonus: old graph of data: http://reviewsignal.com/blog/historicalratings/ it’s from 2012-2013. You can see the major EIG outage and how badly it affected their rankings across the board. You can also see GoDaddy’s super bowl commercial and an outage.

  2. I’ll make sure it’s on the list of stuff to add. I think the next big thing though is going to be round 3 of wordpress testing. A bunch of new companies want to get included and that data is getting old. But building some more powerful graphing tools would definitely be good.

    Any thoughts on the default setting?

  3. Pingback: Understanding The Web Hosting Options for Your WordPress Site - Austin WordPress Meetup

  4. Hey Mendel & Kevin,

    Since you mentioned CloudSpectator and ReviewSignals was also mentioned in the comments I thought I’d chime in and share my own thoughts:

    I recently launched a web hosting review site, too: https://hostingfacts.com – mostly because the current ones are fake (reviewsignal and cloudspectator seem to be legit).

    However, can you give me some feedback to my site? I’ve signed up with the top 27 shared hosting providers and started to track their uptime & page speed via Pingdom.

    Anything I should add/remove to make it look even more legitimate?

    Thanks!
    John

Leave a Comment

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

Posted in