Review of WPEngine WordPress Hosting

wpengine

Hosting is probably the fastest growing segment of the WordPress industry.

There’s several large and well-funded options out there, offering hosting that’s tuned specifically for WordPress.

A few weeks ago we moved our sites to WPEngine.com, one of the most prominent WordPress hosts.

Here’s our warts-and-all review of hosting with WPEngine.

Our situation

In addition to OSTraining, I own a small company that builds tourism sites.

Although it’s not the most popular of the tourism sites, the one that most people would recognize is AppalachianTrail.com. We also host few sites I run for family and friends.

For the last few years, our company ran those sites using WordPress on a VPS with at traditional hosting company.

However, those sites were increasingly slow and prone to botnet attacks that would take them offline for several hours. Even in the quiet times, the sites were running significantly more slowly than they should have.

So, we looked around for some WordPress-specific hosting. We settled on WPEngine because of the pricing scheme. Whereas other WP hosting platforms charge about $30 per site, WPEngine offered one package to cover all 10 sites.

Migrating

The toughest part of our WPEngine experience so far was definitely the migration.

The WPEngine team admitted to us that, because of their unique server setup, their migration process needs to be done differently than with most hosts.

They do have some documentation on migrating here. They’ve greatly improved that documentation since we made our migrations.

In addition to the unusual migration, we encountered bugs in the main WPEngine dashboard. The dashboard allowed you to enter some incorrect settings (they assume you read the documentation correctly) and there were also some bugs that could only be fixed by the WPEngine team. We spent quite a lot of time looking at the squirrel on their error page:

media_1382539132619.png

We also spent quite a lot of time getting help from the WPEngine support team …

Support

WPEngine seem to be making enormous strides in improving their support.

When we first considered becoming a customer we emailed them via the contact form on their site several times and never got a response. My guess is that these were growing pains as their staff numbers climbed very quickly during a period of “hyper growth“.

Now support tickets are turned around quickly. They also have live chat which helped us through a lot of our migration problems.

The migrations were bumpy, but WPEngine support was great and continues to be so.

Pricing and Visits

One of the quirks of WPEngine is that they provide unlimited bandwidth but charge overages based on visits to your site.

This pricing model may lead to some confusion, particularly because their hit counts don’t line up with standard analytics software.

We track our sites with Google Analytics and Clicky.com. Here are the hit counts recorded for one of our sites over the last month:

  • WP Engine: 58,141 visits
  • Clicky.com: 22,607 visits
  • Google Analytics: 23,250 visits

Here’s a second site:

  • WP Engine: 16,052 visits
  • Clicky.com: 3,683 visits
  • Google Analytics: 3,915 visits

And here’s a third site:

  • WP Engine: 31,822 visits
  • Clicky.com: 8,045 visits
  • Google Analytics: 8,965 visits

WPEngine has an explanation of how they count visits, but as you can see, WPEngine counts betwen 2.5 and 4 times more visitors than standard analytics software.

After a month with WPEngine and seeing our first overages, my guess is this confusing definition of “hits” will lead to WPEngine being twice as expensive as we had planned for.

If you do a Google search for “WPEngine hits”, you’ll find that a good number of other people have suffered the same surprise. By moving to WPEngine, make sure you that understand how their definition of hits will impact your costs.

Speed, Security and Stability

As mentioned earlier, we moved to WPEngine hoping for an improvement in our speed, security and stability.

I complained about the way they measure visits in the last section. Well, on the good side, here’s what happened to the loading speed of those sites, as measured by Pingdom.com.

The first site went from loading in 1500 ms to 600 ms:

media_1382538606865.png

The second site went from 1500 ms to 500 ms:

media_1382538636739.png

The third site went from 700 ms to 300 ms:

media_1382538672930.png

So our sites loaded between 2 and 3 times more quickly at WPEngine than with a traditional host.

This speed increase came even though WPEngine asks users to remove all caching plugins.

Uptime for those sites over the last 30 days has been 99.97%, 99.95% and 99.11%.

Security has been great. Even with .htaccess and plugin protections, our sites used to get hammered by spammers. That has largely gone away during the first few weeks with WPEngine. Our sites have been significantly easier to manage since the move to WPEngne.

Conclusion

Overall, we got what we came for with WPEngine. The sites are appreciably faster and more stable than they were with our VPS.

Yes, the migration was bumpy but the support was great and we’ve seen visible signs of improvement in just the last few weeks. They’ve launched a new, better dashboard and the documentation is improving.

FInally, despite the overrages, the cost of hosting with WPEngine will be about the same cheaper than our old VPS.

Instructor

  • Steve is the founder of OSTraining. Originally from the UK, he now lives in Sarasota in the USA. Steve's work straddles the line between teaching and web development.

0 0 votes
Blog Rating
Subscribe
Notify of
17 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Dan Knauss

It sounds like it will be more expensive if you have growth in your visitors, but you are getting far better performance — except for the egregious downtime on the third site. What caused that?

steve

Hey Dan. Yes, the pay-based-on-visits will mean that WPEngine gets more expensive with growth or traffic spikes (at least if they keep up the funny way of counting visits)
Sorry, there wasn’t any major downtime on the third site. Did I make a typo?

Dan Knauss

The third site was down a total of 6.5 hours if uptime was 99.11% out of a 30 day/720 hour period. I’ve never seen an SLA for less than 99.9% uptime/40-60 mins downtime per month.

steve

Good point. Yes, there was some downtime about a week into our stay with WPEngine. On the speed charts, you can see a spike for all 3 at the end of September. It was a few hours of on-and-off downtime one night.
Since then, in October, the sites have all been at 99.9%.

Sabih Ahmed

Hey Steve,
Apparently, you did a good review of WPEngine and what they offer standout in WordPress market. I have been closely monitoring WPEngine and similar companies that offer hosting for WordPress and it would be fair enough to name WPEngine as the best among all.
They have basic package of $29.00.month which is a good fit for SMB’s and bloggers, but yes, the price will increase as you get more of traffic. It is a pay-based-on-visits model as quoted by you, but if you look at services like Cloud hosting – they have better usage option which could be cheap as well. I hosted a couple of my websites on WPEngine and then I moved to cloud. Because on cloud I pay for what I use, and I it can be scaled on the basis of consumption and traffic I get altogether.
One question, did you actually try if WPEngine works on Multisite? I doubt they don’t and which is a big concern for businesses and bloggers. What do you think about it?

steve

Hi Sabih
No, we didn’t use multisite with WPEngine. Actually, we’re on record as not being the biggest multisite fans: [url=http://www.ostraining.com/blog/general/multisites-bad-idea/]http://www.ostraining.com/b…[/url]

Sabih Ahmed

Hello Steve,
Read that blog and it was very nicely written. Taking some notes from the article, I think the best part of multi-site usage is to have administrative control and mostly Devs. and agencies who have a big client base use it. Because they can’t update all the sites one by one and they need solution like multi-site.
I agree that there are complexities and it is with every new technology introduced. So I personally favors multi-site; and its my own opinion and experience.
One thing I always quote – do it if you have any idea how to fix it. So the better is to let experts handle the code work and you (user) simply focus on managing multiple things from a single dashboard.

brettbum

Just curious how this is working out in practice for you. I’m a WPEngine customer also (a new one as well).
I have not yet been billed for an overage. My last bill was the standard plan price, however when I look in my user portal, it shows an overage 5x of my actual bill for last month?
Ergo, I racked up a monster overage, but they didn’t charge me for it.
That’s sort of good and sort of scary. I don’t know if they will fix my billing problem and hit me with a monster combined overage later, or if maybe (as I suspect) their overage calculation is just plane bonkers and they will have to rework this strange hit tracking process.
Great article btw!

steve

Thanks Brett.
I was in the same situation as you and thought they were letting us go without paying the overrage. But, the charge was added to the next bill.

brettbum

Hey Steve, I’m curious did some of the numbers in your overage report look distorted?
I believe a small part of my overage came from a hotlinked image. That looks real.
However, a much larger part of the overage seems to come from the double, triple, maybe 9x counting of a visit. Seems like each time a unique (per Google’s definition) visitor ‘visited’ (per WPEngine’s definition my report shows extra visits (per WPE) for each css style sheet that happens to load supporting that page. (this was from WPE for a month when Google Analytics showed that I had about 15.1k visits, 13k unique visitors, 38k pageviews)
Did you see anything like this? (from WPE)
URL
Visits
/wp-content/plugins/page-list/css/page-list.css?ver=4.2

15,834
/wp-content/plugins/widgetize-pages-light/include/otw_components/otw_grid_manager/css/otw-grid.css?ver=1.1

15,922
/plugins/widgetize-pages-light/include/otw_components/otw_shortcode/css/otw_shortcode.css?ver=1.1

15,886
/wp-content/plugins/widgetize-pages-light/include/otw_components/otw_shortcode/css/social_foundicons.css?ver=1.1

15,912
/wp-content/themes/magazine/style.css?ver=2.0.1

15,889
/wp-content/plugins/responsive-slider/css/responsive-slider.css?ver=0.1

15,867
/wp-content/plugins/widgetize-pages-light/include/otw_components/otw_shortcode/css/general_foundicons.css?ver=1.1

15,848
/wp-content/plugins/genesis-responsive-slider/style.css?ver=0.9.2

15,790
/wp-content/plugins/digg-digg/css/diggdigg-style.css?ver=5.3.6

15,826
In general, the overage report, does not seem to tie numbers out with itself very well.

brettbum

They are still working with me on the account. I’ve since re-setup a cloudflare account on the domain in question. That was done mostly to block botnets, spammers, and trying to benefit from some hotlinking protection. But site is still running very high, even though we’ve dropped things in half with cloudflare.
Still struggling to zero in on what is driving visits…. Is it more hotlinking, is it crawler bots hitting the site, is it something else (javascript, css, some other thing drinving their visits or visitors metric, which is not your typical thing. More like raw hits.
Also trying to find where the raw access logs are stored, to see if that illustrates anything.
The cloudflare stats have proven somewhat helpful, but they aren’t necessarily the stats that wpengine uses to bill me, so that may or may not be a help in the end.

steve

Thanks Brett. They said the same thing to me: using Cloudflare might help reduce the traffic starts.
Looks like it has helped with you – dropping reported traffic in half?
Still, it is a pain to go through all of that, particularly if you have lot of sites with WPEngine.

brettbum

true it is a lot of work (dns changes). Ironically, I had my site on Cloudflare before moving to wpengine and took it off, thinking (inocorrectly) that wpengine would cover the caching and the speed that cloudflare provides. Didn’t think that they didn’t also block bad requests from spammers and botnets and things. I had the cloudflare hotlinking turned on too, but hadn forgotten about that service as it was just a little extra with cloudflare, didn’t expect it might be the difference in $150 a month in expenses.
I run all of my other sites through cloudflare as well. I’ve found that Cloudflare is the single best way to protect against your site getting hacked, plus it reduces comment spam significantly as well.
For me, the time saved around those two activities justified the investment a long time ago.

Nishil Prasad

a easy and perfect migration tutorials
[url=http://www.cmsthisway.com/wp-engine-migration-tutorials/]http://www.cmsthisway.com/w…[/url]

richards1052

I moved to WPEngine about a year ago. At first the speed of page serving was incredible. But I’ve noted in the past few months that it’s degraded and is slower than before I switched. Let me know if your page speed maintains itself over time.
I’ve also found that some of their techs range from incompetent to unhelpful (though there are a few good ones as well). Customer service can be hit or miss with them. If you hit a good tech they’re immensely helpful. Hit a bad one & they’re useless.

steve

Hi richards1052. That’s an interesting point. They were very fast until the end of 2013, but we’ve noticed a slowdown since the start of 2014. I’ll keep an eye on whether that continues and update this post if it does.

Lawrence Tam

I had the same experience and tracked my hostgator dedicated host compared to going to WPengine personal site w/o CDN or cloudflare.
i experienced about 2x the hits with bots and with overage it’s not that bad compared to the costs i was paying with hostgator dedicated.
it’s worth the costs and my 4 speed tests confirm why im sticking with wpening

17
0
Would love your thoughts, please comment.x
()
x