When I launched ClichéWebsite, I promised myself that I’d only work on things that I enjoyed.
Never would I be forced into writing things that bored me purely in the pursuit of traffic and revenue.
With that in mind, I certainly would never have anticipated that I’d be writing a post called ‘Managed Hosting vs Shared Hosting’.
As someone that isn’t interested in computers and is generally scared of anything remotely technical, I just didn’t see this coming.
So why am I putting this experiment together?
I’ve created websites for a long time.
They’ve never looked like anything special.
I’ve always prioritised performance and function over form — results over style has always been my focus.
One thing I always believed in was ‘cheaping out’ wherever I could.
However, over the last few years, I’ve developed a deeper understanding of value.
I still like cheap, but I also like profit.
Sometimes you need to spend more to make more.
With that in mind, I wanted to test whether there was any value in investing in more expensive web hosting for my sites.
I’ve seen a lot of people giving high praise to certain hosting companies, but with huge affiliate bounties on offer, who can you really trust?
As someone that genuinely has the best interests of my audience at heart, I also wanted to see if there was an opportunity that people could be missing out on with their own sites — as well as trying to improve my own.
I decided to take the plunge with ClichéWebsite as I jumped into my first ever managed hosting package.
If you’re reading this and you have no idea what ‘managed hosting’ means, here’s a very quick overview:
Shared hosting is very cheap. The reason it’s very cheap is that you’re sharing server space with many other websites, so it doesn’t cost the hosting company much to support your site. If you have low-price web hosting, it’s almost certainly on a shared plan.
Managed hosting, or dedicated hosting — technically they’re different, but they’re similar enough to band as one for this example — means you have your own server space specifically for your site.
Managed hosting solutions usually cost a lot more than shared hosting, typically more than $30-per-month for a reasonable package compared to less than $10-per-month for shared hosting.
Now, whilst I’m always someone that will take the cheap route given the chance, I also want to take competitive advantages if I can get one.
The online world is extremely saturated, but an edge that I definitely know is possible is having a fast website.
I’d estimate that 90% of small websites — and a reasonable percentage of larger ones — are missing an opportunity here.
I didn’t know if managed or dedicated hosting would be the solution I needed to get a lightning-quick site, but this seemed like an extremely low-risk investment to me.
After all, if it didn’t work out, I could return to the cheaper shared hosting, right?
It was roughly a $50 expense if things didn’t work out and refunds are always a possibility too.
If I saw no benefits from my experiment, at least I’d have tried it and learned from my experience. Win-win.
I used three different tools to measure the speed of ClichéWebsite:
I didn’t want to test my homepage as I knew it was relatively small, so I chose the biggest post on my site; my 9,500+ word case study.
I only cared about one main thing:
How long did my site take to load?
And, as a small subsection of that, I also wanted to know my Time To First Byte (TTFB) — basically, how long did my site take to start loading?
The site received a 100/100 score for desktop speed. Oh stop it you, you’re making me blush.
The TTFB was an incredibly low 0.12s.
I got a ‘Speed Index’ time of 1.9s, which is the time it took for the page to be fully visible and usable.
All well and good, but other tools can sometimes show other results so I wanted to triple check.
Pingdom gave me a performance grade of 76 which is more than acceptable considering they gave me an ‘F’ for not using a CDN — which I do use.
I was really happy with the load time of 843ms, or just over 0.8s, tested from London, UK, near-ish to where I’m based.
Both Pingdom and GTMetrix (which I also ran the site through to triple check results) measured the TTFB at around the same as Pagespeed Insights did; 0.1s or thereabouts.
So, basically, my speed is fucking amazing — please now shower me with praise and maybe even kiss my feet should you wish to do so.
Despite the results…
I still wasn’t convinced.
I was now paying $30-per-month for my managed hosting compared to the $7-ish I’d be paying with the host I usually use.
With almost a $300 annual difference, I needed more convincing that the choice of host was delivering the improvements, rather than my extremely lightweight new website.
Fortunately, I had a couple of existing websites that I could test on both shared and dedicated hosting to see what the differences were.
So, the first thing I did was move my main affiliate blog to managed hosting.
“Risking a large chunk of your income for an experiment, great idea Luke!”
I knew my affiliate blog was a much better test because I’d tried every WordPress speed optimisation under the sun over the years of running it.
Those optimisations definitely do their bit to help, but I wanted my site to be fast, not just quicker than it was before.
All of these tests were done with exactly the same website settings, other than the plugins that are removed by default by the managed hosting company I use.
Therefore, the only differences between the ‘before’ and ‘after’ tests are caching and image compression plugins (that were removed from the site, not added) and, of course, the choice of host.
So, here are my ‘before’ tests:
I can’t quite believe how bad those results are.
They weren’t always this bad, I swear.
The server response time was the bane of my life with this site.
It’s something I had longed to cure but no changes I made — either on-page or behind-the-scenes — could ever seem to make it budge.
I hired numerous supposed experts to speed it up, but these results are the best they could muster up.
Pingdom, like PageSpeed Insights, gave a damning verdict of my site’s performance.
They measured the load time at 3.95s.
Based on these results, it was an easy decision:
The site had to be moved.
Any improvement would be down to the switch to managed hosting.
Here are my ‘after’ tests, taken a few days after switching:
The speed score had risen from 24/100 to 96/100.
The TTFB was reduced from 3.36s to 0.2s.
Pingdom backed up the results of PageSpeed Insights, judging the total load time of the site at just over 0.7s — around an 80% improvement.
I’d done it.
I’d finally found the cure to my affiliate site’s speed issues.
I was now very confident that managed hosting was giving me the speed boost necessary to finally have a truly fast website.
The final thing to do was move my friend’s e-commerce site — that also used shared hosting — to see if a site with a completely different setup to mine would also see a positive difference.
Also built on WordPress, and using WooCommerce, this site had always been painfully slow throughout its 5+ years of existence.
Just like before, the only differences between the ‘before’ and ‘after’ results is the 2-3 speed-related plugins that automatically get removed after moving to the new host I chose, and the new hosting.
Just look at how slow it was — especially for an e-commerce store:
The TTFB was a massive 5.84s.
That means any potential customer would have to sit around for almost 6 seconds just to see their website start to load.
Pingdom wasn’t quite as critical, measuring in at 2.8s, but people don’t want to wait this long for websites to load.
This is especially true when it comes to people looking to buy products.
It was obvious that there was huge potential for improvement here.
We switched the site to managed hosting and I will attribute any change to that in the following ‘after’ tests:
Not only was the site receiving better speed scores, but you can also see from the image loading at the bottom (intentionally blurred) that the page was rendering much more efficiently too.
The TTFB was reduced from 2825.5ms to 7.5ms.
That’s over 2.8 seconds down to less than 0.1 seconds. A huge improvement.
Total load time was scored at 0.5s by Pingdom.
Whilst it’s not been long enough to comment on conversion rate or user behaviour statistics just yet (though I may return to add those at a later date), we can see in Google Search Console that Googlebot is recognising the incredible shift in page speed:
As speed is an important part of ranking in Google, I expect this will have a huge part to play in the future success of the site.
The average ‘Time Spent Downloading a Page’ is now just 18% of what it was on shared hosting.
You will not find a single optimisation for speed that comes anywhere close to that level of improvement, and that’s almost a promise. I’m not technically-gifted enough to say whether it’s actually a fact, but I’ve never seen anything like that before.
For my friend’s e-commerce site — that generates a minimum of five-figures in revenue every month, and often a lot more depending on seasonality — the cost of switching to dedicated hosting is going to be paid back very, very quickly.
But what about a site like ClichéWebsite, a brand new site that no-one has heard of? A site that doesn’t even sell anything, and only includes a few affiliate links to products the author (that’s me) truly loves?
What about your small website or blog?
Obviously, there is going to be no catch-all answer here as this totally depends on your circumstances, but I’d wholeheartedly recommend going with managed hosting if you want to do any of the following:
The cost of managed hosting varies depending on your traffic levels, but I’m currently spending $60-per-month to host two sites with Kinsta.
I’d check Kinsta out here, and yes that is an affiliate link, but I’ve intentionally not named them until now because I wanted the experiment to take priority over a potential sales pitch.
On my affiliate blog, the cost will have paid for itself already:
You can see that in the short period of time since moving to managed hosting, the traffic to that site has been very similar. However, engagement is up massively; people are browsing for longer and more often, viewing more pages, and bouncing (leaving) less.
An 8.34% increase in pageviews from just a 0.91% increase in users isn’t too shabby at all. I attribute that to the new found speed as there have been no other changes I can point to.
I’m sure that’s going to have a knock-on effect for the rest of the site’s results.
As for this site…I don’t know.
But I like the fact it runs extremely quickly.
I hope you do too.