How to Choose a Managed VPS Hosting Provider

On a pricing page, most VPSs look remarkably similar. The bigger differences often only become obvious after you've bought one.

CPU, RAM, storage and bandwidth are straightforward to compare. What is much harder to judge is everything that happens around the server.

If you're trying to work out what to look for in managed VPS hosting, the specifications are only part of the answer.

Will the provider take time to understand what you're actually hosting? Will somebody notice when something starts going wrong? If the problem turns out to be outside their immediate responsibility, will they still help you understand it? And when they recommend spending more money, will they be able to show you why?

Those questions tend to matter far more once you're relying on the hosting to keep an important website, application or service running.

Here are five things worth looking at when choosing a managed VPS hosting provider.

1. Does the hosting provider understand your workload?

An 8GB VPS is an 8GB VPS on a pricing table. That doesn't mean two customers with an 8GB server have anything like the same requirements.

A WooCommerce store, an agency hosting dozens of WordPress sites and a SaaS application running database-heavy background jobs all behave differently.

A good managed VPS provider should want to understand that before recommending infrastructure.

They should ask about things such as traffic, applications, databases, expected growth, security requirements and how important the service is to your business.

There may also be several technically valid ways to host the same workload. The job isn't simply to find something that will run it. It's to work out which option makes the most sense for the customer.

Legacy applications are a good example. Rebuilding everything onto the latest technology might sound ideal, but it isn't always practical. Sometimes the better answer is finding a safe and sensible way to continue running an older application while managing the risks around it.

If a hosting recommendation arrives before anybody has properly understood what you're running, you're effectively being asked to design the solution yourself.

Ask: What would you want to know about my application before recommending a hosting platform?

2. What does managed VPS hosting actually include?

"Managed hosting" is one of those terms that can mean very different things depending on who is selling it.

For one provider it might mean installing the operating system and keeping it patched. For another it can mean taking a much more active role in keeping the services running on that server healthy.

Neither definition should be assumed.

When comparing managed VPS hosting providers, find out exactly what is included, what remains your responsibility, and what happens when a problem falls somewhere between the two.

Every provider has limits.

A hosting company isn't normally going to rewrite your application code, for example. But there is a significant difference between:

"The server is working. Speak to your developer."

and:

"The server looks healthy, but we've traced the problem to this part of the application. Here's what we're seeing and what we'd suggest asking your developer to investigate."

The provider hasn't crossed the boundary into maintaining the customer's software, but they also haven't left the customer stranded.

In other cases they may be able to go further. Database performance problems, for example, might involve configuration changes, database tuning or identifying indexes that could improve performance.

The important distinction is that having a support boundary doesn't mean support has to stop being helpful when you reach it.

(For a sense of where that boundary sits in practice, see what you can expect from a managed hosting provider.)

Ask: If you discover that my problem is outside the part of the system you manage, what happens next?

3. What does good VPS monitoring actually look like?

Almost every managed hosting provider will tell you that its infrastructure is monitored.

That isn't really the interesting part.

The question is what happens when the monitoring finds something.

Good VPS monitoring should provide visibility into the health and performance of the server, services and underlying infrastructure. More importantly, it should help identify developing problems before they become bigger ones.

A server gradually running short of memory is a simple example.

The answer might genuinely be that it needs more RAM. But that shouldn't automatically be the first conclusion.

Someone should be able to look at what is consuming the memory, whether the usage is temporary or persistent, and whether there is an underlying application or database issue contributing to it.

If more resources are required, the provider should be able to explain why.

You shouldn't just be told that your server needs an upgrade. You should be shown the evidence and given enough information to understand the recommendation.

This is where proactive VPS monitoring becomes more than a collection of graphs and alerts. It helps the provider and customer make better decisions about the infrastructure, and it's a core part of what we mean by fully managed VPS hosting.

Ask: If monitoring shows that my server is running short of a resource, will you investigate why before recommending an upgrade?

4. What should managed VPS support look like when things go wrong?

Almost every hosting company looks good when everything is working.

Problems tell you much more about the relationship you're actually buying.

When comparing managed VPS support, don't look only at advertised response times. Find out what happens after somebody responds.

When something goes wrong, you want clear information about what has happened, what is being done and what happens next.

Sometimes the provider may have made a mistake. Hosting companies are run by people, and people make them.

What matters is how that mistake is handled.

Good incident communication isn't about producing a carefully worded update that avoids admitting anything. It's about giving the customer enough information to understand what happened and what is being done about it.

The same principle applies to everyday support.

If a provider recommends more resources, they should be prepared to explain the reason. If something can't be done, they should say so. And if the customer's preferred approach isn't sensible, they should be willing to explain that too.

You should never have to read between the lines to work out what your hosting provider is trying to tell you.

Ask: Can you show me how you communicate with customers when something goes wrong?

5. Questions to ask a managed hosting provider before you buy

One of the best times to judge a hosting provider is before you become a customer.

See what happens when you start asking technical questions.

Can you talk through what you're trying to build with somebody who understands the infrastructure, like the people who'd actually be looking after it?

Will they help you work out the right configuration?

Are they prepared to discuss different approaches?

Most importantly, does every conversation mysteriously end with the most expensive option?

Good technical advice should sometimes result in buying more resources, but sometimes it should result in buying less.

It might mean choosing a different type of hosting entirely. It might mean changing the architecture. It might mean keeping an existing system and making it safer rather than rebuilding it.

The objective should be to find the best way of meeting the customer's requirements, not simply to find the largest package they can be persuaded to buy.

That matters after the sale too.

Requirements change. Traffic grows. Applications evolve. Old systems sometimes need to stay running far longer than originally planned.

A good hosting relationship should be able to adapt to those changes rather than forcing every customer into the same predefined box.

Before choosing a managed hosting provider, ask:

• What would you want to know about my workload before recommending a server?
• What exactly is included in your managed VPS support?
• What happens when monitoring detects a problem?
• If you recommend more resources, can you show me why they're needed?
• What happens if the problem turns out to be outside your support boundary?
• Can I speak to someone technical before I buy?

The answers will tell you far more than another row on a VPS comparison table.

The server is only part of what you're buying

Specifications still matter.

So do price, performance, security, backups and the underlying infrastructure.

But they're only part of how to choose a managed VPS hosting provider.

You're also choosing the people you will depend on when the application starts behaving strangely, the database slows down, monitoring detects a problem or something simply doesn't work the way you expected.

Before choosing a provider, find out how those people behave.

Do they listen before recommending, or do they explain rather than simply tell? Do they investigate before asking you to spend more?

Do they admit when something has gone wrong?

And when they reach the limit of what they can reasonably do for you, do they leave you with a dead end, or help point you towards the next step?

That's the real test.

Those answers will probably tell you more about the hosting relationship you're buying than another comparison of CPU cores and gigabytes.

If you're considering Layershift, you're welcome to talk to us and ask us the same questions before making a decision.