The world's move into the cloud continues. In fact, Gartner predicts that the global software-as-a-service (SaaS) market alone with grow from $85.1 billion in 2019 to $113.1 billion by 2021. As more organizations of all sizes start entering the cloud, they need to answer the constant question of "on premise vs cloud". The on prem vs cloud analysis applies to all new systems a company is bringing online, but also to existing systems already running onsite. Should they remain as on-premise installations or are there advantages to moving them to the cloud?
Some of the nuance to the analysis may vary based on whether an existing or new system being considered. However, the basic questions remain the same.
Regardless of the size of the IT department's team and budget, there are always limitations on what can be a priority and where money can be spent most efficiently. Cloud solutions offer smaller and medium-sized organizations that can't afford the full-time tech expertise clear opportunities to access the most sophisticated business technologies by shifting the support, maintenance and upgrade responsibilities to the SaaS vendor.
However, even enterprise organizations – heavy users of cloud computing – are trending towards leveraging internal IT resources for developing in-house tools that generate quantifiable business value in their company's business mission. One way they support this priority is by using SaaS applications to handle administrative processes like time and attendance and workforce management. A hotel's expertise isn't in managing tax compliance in tip reporting. A hotel chain wants to focus internal IT resources on how to deliver better guest experience.
The question to ask is whether the system touches on processes and skills within the company's core competencies or if it's outside its area of expertise. The farther outside core competencies the organization wants to prioritize, the more it makes sense to push its deployment and ongoing support to the experts who developed the system and work in it every day.
The cost models for on premise vs cloud are very different, which impacts their relative total cost of ownership (TCO).
Companies pay licensing fees for on-premise software, which are typically governed by strict, multi-year contracts. Cloud-based software is typically sold on a subscription basis with more flexibility regarding adding/removing users or modules. The subscription model gives companies more control and flexibility over their monthly costs for the software.
The cloud-based system agreement generally also includes automatic upgrades as new versions of the software rollout, although major software advances may not be covered. However, the SaaS model will likely include a broader range of software upgrades than what's included in an on-premise contract.
The main driver that causes on premise TCO to spike is the need for companies to buy and manage the hardware and space needed to run the system. The more complex and data-driven the system, the more server space and processing power it needs to function. If the company needs to provide access to the system to external locations, the processing and server requirements go up to accommodate the remote access. The hardware issue aligns with the issue of where a company wants to dedicate its internal IT resources. Does it want to manage servers, their upgrades, security and operations; adding the costs to maintain that personnel expertise to the system's TCO?
The hardware cost issue is why even those enterprises that want to develop business-value system internally look to run those systems on external hardware it can access through the same "as-a-Service" model as software. They can access commoditized hardware, like server space and processing power, on the same flexible terms as a software. This leaves budget for specialized hardware needs, like time clocks or onsite security tools.
Run the numbers to compare the ongoing costs and TCO over a multi-year period between a cloud and on-premise option to determine which makes the most sense economically.
Companies get nervous about cloud or hosted solutions because they feel like they don't have control over their data: How secure it is against bad-intent intruders and how easy is it for the company to maintain access to it?
With an on-premise solution, the company knows it will always have access to and control over its data. The company also has control over their data's security, which may or may not be a good thing (see section above about internal IT expertise).
In some highly regulated industries, there may be legal requirements, such as those regarding protecting privacy of people's sensitive information, that place a higher burden on the company to secure it.
A good SaaS company is aware of data security and access concerns. Getting the specifics regarding encryption and protection features of a SaaS system is critical. The threshold question to ask, however, is what iss the nature of the data being stored? The more sensitive and the higher obligation to protect data that a company has, the more it may prefer to keep that data on-premise. The same analysis holds true for business mission critical data.
The on premise vs cloud debate will continue. Before running the analysis for a specific system, it's worthwhile to develop internal criteria regarding business goals and priorities. Having these internal criteria clarifies adding consistency to the analysis as it gets applied to multiple systems.