Which is more expensive: hosting your tools on-premises (on-prem) or in the cloud? Ask a dozen people, and you’ll get a dozen different answers. While businesses tend to treat it as a simple question, it’s actually a rather complicated one.
If we’re talking about the monthly subscription cost of cloud vs. software licensing, cloud looks more expensive. If we’re talking about the cost of migrating from on-prem to cloud, it’s again easy to conclude that the migration itself makes cloud a pricier option.
But when we look at long-term value, that’s where on-prem starts to look less like the conservative choice and more like a consistent drain on your profits.
Why? Because, iceberg-like, the price tag of on-prem is mostly hidden, and bigger than you might think. Systems downtime can cost three times as much as a year-long cloud subscription in a matter of minutes or hours. IT time and resources can be cut in half by a move off-prem. And that’s not even factoring in operational expenditures and the cost of overprovisioning resources (which impacts the majority of on-prem companies).
In fact, right-sizing your servers by moving to the cloud brings in an average annual cost savings around 30 percent, according to a recent study of 35,000 servers. The cost of unused licensed software in the US and UK is a whopping $34 billion per year. And IT pros report an average of 20 percent overall cost savings after ditching on-prem, according to a study by Office 365.
So, the real question here isn’t what’s cheaper – it’s whether you’re taking the long or the short view. Are you comparing only the visible, up-front costs? Or looking at the big picture, factoring in the total cost of ownership – everything from IT time to server replacements? When you look past the tip of that iceberg, you’ll find a long list of ways that cloud saves money in the long run.
Reduce – or even eliminate – the cost of major incidents
The average amount a company spends on downtime is $5,600 per minute, according to 2014 research by Gartner. And since 2014, that estimate has only gone up, with more recent reports putting the figure somewhere around $9,000.
Of course, that’s just an average – and some companies have a lot more to lose, like Facebook, whose 14-hour outage in 2019 lost them an estimated $90 million.
With on-prem, downtime falls squarely on the shoulders of your IT team, and it can cost you big in the form of line items ranging from revenue to internal productivity, SLA penalties, overtime, or on-call emergency pay.
This is one of the biggest opportunities for cost savings in the cloud. Instead of putting uptime on your team’s plate and hoping your servers and systems can handle a major incident, you outsource those responsibilities to your cloud vendor. Atlassian, for example, guarantees 99.95 percent uptime, and if an incident does happen, we’ve got the resources in place to resolve it, quickly and without additional cost to you.
Free up your IT team (because we all know time is money)
Make a list of all the things IT has to do to manage your on-prem servers and that list will get long fast. Performance upgrades. Scheduled upgrades. Security patches. Server replacements. VPN installations for remote access. Incident management. Change management. Manual integrations.
When you shift to the cloud, all those tasks fall squarely on the shoulders of your vendor. They’re responsible for upgrading security and maintaining servers, replacing old technology with new, and regularly upgrading software to address feature requests and bugs.
This means your IT teams – who tend to be situated on the high end of the pay scale, by the way – are free to focus on creative or urgent tasks instead of tedious ones, and is why 74 percent of organizations say cloud gives their team a competitive advantage.
Reduce operational and physical costs
On-prem also comes with a lot of hidden operational and physical costs that simply aren’t a factor with cloud. This includes:
- Servers: With an average lifespan of 3-5 years, servers need to be regularly repaired and physically replaced.
- Server support: Load balancers, climate control, server racks, replacement parts…in addition to the servers themselves, on-prem comes with some supporting hardware, parts, and physical assets that need to be purchased, maintained, and replaced at regular intervals.
- Software renewal/licensing (and over-licensing): Over-licensing costs US and UK companies as much as $34 billion per year, according to one study. To avoid this common pitfall, companies either need to keep rigorous track of who needs which software, or they need to move to the cloud, where the number of users can often be automatically tracked, updated, and viewed by admins.
- Electric bills: If 80 percent of servers are overprovisioned, that means 80 percent of on-prem companies are using more energy than they need and paying higher energy bills than they otherwise would.
- Real estate/space: Physical servers call for physical space, which means a move to the cloud can open up existing space for other uses, or remove data center real estate from your budget entirely.
- Maintenance: Server maintenance often calls for temporary staff or contractors, which is a line item you can ditch when you delegate that responsibility to your cloud vendor.
- Asset management time/audits: The more assets your IT team has (including physical servers, load balancers, and parts as well as non-physical assets such as software licenses and databases), the more your asset management practice has to track. This means more time, resources, and mental overhead.
Reduce environmental costs
Most of us would love to be more environmentally friendly for no other reason than it’s the right thing to do. But the additional good news is that when it comes to on-prem vs. cloud, the environmentally friendly option (cloud) is also the more affordable one.
The reason for this is, of course, that energy costs money. And using more of it than you need necessarily costs more. So when we say cloud is up to 98 percent more eco-friendly than on-prem, we’re also saying it’s cheaper.
Offload the cost of scaling
The vast majority of on-prem resources (80 percent) are overprovisioned, meaning companies are paying for far more computing power than they need. In those cases, a move to a cloud service that automatically scales resources up, down, in, and out saves these companies as much as 30 percent annually, according to research by TSO Logic.
The problem here is that, with on-prem hosting, your IT team makes an educated guess about how much computing power you’ll need. If they guess too high, you’re paying for resources – servers, load balancers, power – you don’t need.
On the other hand, if the team guesses too low, a lengthy, costly manual scaling process is in your future. You’ll need to add more servers or more computing power to meet demand – and that addition will require both money and manpower. Not to mention the weeks, if not months, of slow or unavailable services in the meantime, and the impact they could have on profits and customer loyalty. An incorrect guess in either direction, then, can have a major impact on your bottom line.
The solution here is to choose a cloud service with automatic scaling options. When usage spikes, your computing power grows to meet that demand. When usage slows, it scales down to save you money.
Calculating the cost of cloud
The simplest way to calculate the return on any investment (including a move to the cloud) is this:
So, for example, if you invest $50,000 in a migration from on-prem to cloud, and you save or gain $50,000 per year after the migration, your equation would look like this over three years:
($150,000 – $50,000 = $100,000) ÷ ($50,000) = 2
In this example, your ROI over three years would be 2x, or 200 percent. In the first year, with that equation, you’d simply break even. In years two and three, though, you’d start to see real gains.
The tricky part of this equation is calculating the two numbers you need for your ROI. To understand your initial investment in migrating from on-prem to cloud, you’ll need to add up the cost of professional services, internal resources, software licenses, data migration, cloud subscription, and any required re-training on cloud tools (if they differ from your on-prem tools).
Then, to calculate your gains, you’ll need to add up savings on hardware, software licenses, energy, real estate/server rooms/data centers, maintenance (including both employee time and external contractors), asset management time, incident management time, change management time, security upgrades, feature upgrades, and IT team or reduced headcount.
More difficult to calculate before you make the switch – but still important – are the cost of downtime (even a reduction of one hour per year can save companies hundreds of thousands), performance gains, and time saved by non-technical teams who have faster access to new features that increase productivity, collaboration, and security.
Case study: igloo software
After a major outage that cost them three times as much as a Jira cloud subscription, Igloo Software decided to make the shift from on-prem to cloud. And that shift saved them big – not only on the cost of future major incidents, but also on admin time and scheduled downtime.
As their Senior Tools Admin James Seddon explains,
“When we managed our own Jira server, every upgrade required at least two hours of downtime, and we had to schedule it after 8 PM, which meant a late night for me, the admin. Upgrades to Bamboo and Bitbucket, which we did separately, would also each take at least two hours.”
Another time-saver (and, therefore, cost-saver) Seddon highlights is that users can configure features on their own—no admin assistance required. Since the switch, support tickets are down by a whopping 50 percent because users are empowered to do so much more of their own admin work.
If you’re considering cloud migration, we’re here to help. Visit our Cloud Migration Center for more resources, migration guides, and a free extended cloud trial to help you assess our cloud products.