Planning Your Cloud ERP Deployment

Your next big decision when considering a new ERP system? How it will be deployed. There are several approaches to choose from, including traditional on-premise, cloud deployment, or a hybrid combination of both. You will want to consider financial differences and look at each ERP deployment strategy’s strengths and attractions, as well as any unique limitations or challenges. It’s vitally important to consider all options and choose the deployment approach that will best serve your business needs today – and through the foreseeable future.

Before you check out your ERP deployment options, see how – and why – the cloud has become a vital environment for business success in the digital economy.


Why all the interest in cloud deployment?

We all watch with keen interest as computer technology continues to evolve at an ever-increasing pace. The earliest business systems and the first ERP systems were hosted on large mainframe and mid-range computers with “dumb” terminals for user input – after they’d replaced punch cards and key-to-disk input, that is! The next “state-of-the-art” technology was client/server architecture wherein the dumb terminals were replaced by PCs (now called clients) that could handle a portion of the workload. This reduced the amount of data that had to be shuttled back and forth with the server computer.

Around the millennium, two significant developments changed the world of computing and ERP: the Internet and the cloud. No longer was it necessary to buy and support hardware and software to run your business. All or major pieces of the technology could now be “rented” or outsourced in a bundle that included all maintenance and much of the technical operations. The ubiquitous Internet provided the communications infrastructure needed to make the cloud practical and available anywhere in the world.

But businesses did not immediately migrate their ERP systems to a cloud deployment. It took some time for the technology and applications to mature – and for companies to realize that the computer does not have to be on-site in order to get reliable access and security for business-critical applications. Part of that evolution involved developers learning how to:

  • Take full advantage of cloud deployment
  • Write or re-write applications appropriately
  • Rebuild their infrastructure – both technology and human resources – to support cloud-based systems

And, importantly, now cloud ERP systems deliver advanced technologies like artificial intelligence (AI) and machine learning to enhance productivity and service; tailored user experiences to drive adoption; and wide-ranging functionality and embedded analytics to provide a complete view of the business – ultimately supporting business innovation and growth.


The SaaS pricing model

In the past, most software was installed on company premises and the only licensing option was to purchase a perpetual license – where the application is licensed for an upfront amount plus a yearly maintenance contract for upgrades and bug fixes. Software licenses are most often priced on a per-user basis. Annual maintenance is commonly charged at 18% to 20% of the then current software list price. This means that the software license is essentially “repurchased” every five to six years.

With on-premise ERP, all the hardware and software is purchased or leased and installed at a company’s location(s). The company is responsible for maintenance, support, and possible upgrades or expansion of the hardware, systems, and application software – as well as the facility space, utilities, insurance, failover resources, and off-site backup storage.

Cloud-based ERP systems, on the other hand, are not typically installed on-site and are supported by the supplier as part of a monthly or annual fee. They are licensed on what is called a software-as-a-service (SaaS) basis. SaaS licenses can be priced per user, per application, or per application set (all of ERP, for example), based on the size of your company or other variations.

An interesting aspect of cloud deployment licensing is its scalability. If licensed by the user “seat,” you can add or decrease users and pay a higher or lower price thereafter based on the new user count. If your transaction volume, storage capacity, or computing power requirements change, the supplier is responsible for upgrading its facilities to accommodate the change, which means you won’t have to buy and install more servers or more disk storage.

The closest analogy might be cable TV. You pay for what you require, and that price includes use and operation of all the physical facilities, personnel, maintenance, and any other expenses related to the cable service at your location. If you need and add more channels, you will just pay for what you requested and not worry about how they manage to provide the additional channels.


Public vs. private vs. hybrid cloud vs. two-tier

There are four possible ways to implement a true cloud ERP system:

Types of ERP cloud deployment


1. Public cloud or SaaS ERP

The public cloud is the primary licensing model for software-as-a-service (SaaS) ERP solutions. The system provider will have its own data center – or may lease space on a public cloud to host its applications and systems. All hardware, systems, and support services are provided through the public cloud. This makes implementation quicker and easier for the user company because, with all the hardware and software pieces already in place, they can start right in on the data transfer and user training.

With this cloud ERP deployment option, your software provider will also take care of all the installation, maintenance, and support – including any and all software updates and upgrades such as adding computing power or storage. As well, systems, applications, and resources can “autoscaled,” or increased or decreased automatically to meet changing, short-term needs. This eliminates having to pay for computing resources that may sit unused most of the time as is the case with on-premise ERP systems.

Public cloud-based SaaS ERP has little or no upfront cost (a “capital expense”) but has a somewhat higher monthly cost (an “operating expense”), compared with a typical on-premise installation. When viewed over the normal lifecycle cost period of five to seven years, total cost of ownership (TCO) is similar to, if not less than, an on-premise installation and offers potentially better service, support, and security.

Public cloud ERP also offers the fastest path to innovation, making it ideal for businesses that want to aggressively pursue their digital transformation strategy. This deployment option allows companies to easily reimagine, optimize, and adapt their business processes as needed – and to take advantage of standardized best practices which modern ERP vendors should support.

2. Private cloud ERP

Although similar to the public cloud option, private cloud hardware, system software, and support may be owned, managed, and operated by the company, a third party, or some combination of both for the dedicated use of a single organization. With a private cloud deployment, the user company is usually required to pay for the ERP software license.  

The third-party ownership option is popular with IT departments that wish to outsource the hardware, database, and much of the networking tasks – giving them some of the benefits of a public cloud. This deployment option is also favored by companies that want to transition to the cloud in phases, whether quickly or gradually, or as an interim step to the public cloud. This is especially true for large global manufacturers and other businesses with complex, fragmented, or highly customized system landscapes. 

Private cloud deployment typically has more upfront investment (capital expense) but may work out to a lifecycle cost somewhere between the public cloud and on premise. Some vendors are changing this calculus by offering bundled implementation packages that reduce upfront costs and include all the tools and services, infrastructure, and network requirements through subscription-based pricing. Companies can take advantage of the lower TCO from cloud economics, a modern cloud-based architecture, as well as full ERP functionality that includes partner add-ons, extensions, and enhancements.    

3. Hybrid cloud ERP

Elements of a private cloud, public cloud, and on-prem ERP deployment can be combined to create a hybrid cloud, which provides the flexibility to choose the optimal deployment for each application. Hybrid cloud ERP can be used as a stepping stone to the public cloud, or to satisfy industry regulatory issues and special security requirements that may dictate the need for on-prem applications in certain situations. There may also be other restrictions or preferences that make on-prem desirable for certain applications. The complexity of a business and its current environment as well as desiring a slower speed of change factor into the decision to deploy in a hybrid scenario.

A hybrid implementation allows applications and data to move between the options based on workload changes. It delivers cloud benefits from that part of the system that is on the cloud. However, it requires more local IT involvement to support the on-prem pieces, as well as the coordination between the two – or more – ERP system environments.

4. Two-Tier ERP

Really a variation of the hybrid approach implemented for the same reasons, two-tier ERP deployment – sometimes called hub-and-spoke deployment – employs a central system with smaller satellite systems supporting remote facilities. Think of corporate ERP as the hub, with individual ERP systems at subsidiary plants, warehouses, or offices all feeding data back to the hub. This is not a new idea; it emerged during the 1990s’ distributed processing phase, with companies choosing to implement smaller, simpler, and less costly systems at remote locations while maintaining the larger, more capable, corporate system at company headquarters. Any or all of the systems in a two-tier network can be on-prem or cloud-based, purchased or SaaS-licensed.

The overall cost for a two-tier ERP deployment – with less costly systems at the nodes instead of the same corporate system everywhere – will yield a lower cost for the initial purchase. However, integration and support can result in a higher overall, continuing cost because the interfaces must be built and maintained. And, year after year, it will take more IT support to coordinate with multiple suppliers, as well as manage uncoordinated upgrade schedules and interface changes.


What does the term “fake cloud” mean?

Fake cloud, also known as “faux cloud” or “cloud washing,” refers to a legacy ERP system ported to the cloud and perhaps “wrapped” with some additional software to adapt the system to this environment. But these applications are not written for cloud deployment so they cannot really benefit from what cloud has to offer. These are exactly the same legacy ERP applications installed on outsourced hardware. The “wrapper” might present modern, Web-like screens to the users, but the input information must be translated into the legacy system’s input requirements and translated back out to the wrapped screens for display – not a very efficient approach. To the user, it looks like the cloud, but it will not perform like a cloud application and will not be able to take advantage of cloud connectivity, advanced functionality, or optimized operational performance.

True cloud ERP vendors design their solutions from the ground up, specifically for the cloud. Cloud-wrapped legacy applications were not designed for the cloud, so performance issues may occur. Customizations and integrations can also be troublesome – and these solutions must still be updated and maintained, often by the user company’s own IT resources.

Since legacy applications ported to the cloud are essentially the same as on-prem applications, pricing is seldom based on usage requirements, so over-buying is possible. Furthermore, the SaaS model is not commonly applied, which means the user company retains all support and upgrade responsibility in-house.


When to choose an on-premise ERP system over cloud

More and more companies are moving to cloud ERP, however, it is not for every company. The number one reason to stay with an on-premise ERP solution is the need for compliance – whether it’s customer, industry, or government requirements regarding regulations and standards. Stricter requirements sometimes require on-premise implementation in more regulated industries. 

Unreliable Internet service is cited by some companies as a reason for not moving to the cloud. For mission-critical ERP applications, being up and available 99% of the time is crucial. However, with modern networks, servers, and processes, downtime is typically no longer a concern and seldom precludes deploying ERP in the cloud. 

Data management may be another reason to keep your ERP system on your premises. With a cloud deployment, you may or may not be able to easily move your data – depending on your service provider’s policies. Make sure they support the services you need. 

Another reason is loss of control (for example, over security, data, or upgrades). With cloud-based ERP, your company is off-loading many IT responsibilities to a third party so it’s important to ensure that the third party is reliable and has a stellar track record. However, some companies still choose to keep everything “in house.”  

Some of the challenges of on-premise ERP deployments include the need for manual system upgrades and the lack of built-in installation, maintenance, and support services. Thankfully, for companies that require this type of implementation, some vendors offer services that deliver some of the benefits of cloud-based software. 


Tips for the selection process

Choose the ERP software first through a careful evaluation process, then consider the deployment options based on the software’s deployment capabilities and your company’s needs and potential ERP return on investment. Some ERP software is cloud or SaaS only, while other vendors offer cloud, on-prem, and hybrid. Deployment options available may be a criterion for inclusion on the short list, but they should not be the sole determinant for system selection. .

  • Avoid fake cloud legacy ERP applications for the reasons stated above. 
  • Fast growing companies and those that expect upcoming changes in the number of users (either up or down) should probably focus on true cloud ERP systems for their scalability and pay-for-what-you-use pricing. 
  • Be sure all the members of the selection team understand the features and benefits of public versus private versus hybrid cloud ERP. 



Companies may initially be drawn to cloud ERP deployment for financial reasons – like little or no capital outlay and reduced total lifecycle cost – but are excited by the technical and operational advantages the cloud has to offer, including:

  • ERP that is always up to date with the latest upgrades (at no extra cost or effort)
  • Virtually unlimited scalability
  • Pay-only-for-what-you-need/use basis
  • Faster implementation
  • Better security and access controls, and more

And, full cloud deployment is not the only option. Sometimes it makes more sense to keep some of the applications on premise and use the cloud for the rest. Fortunately, a variety of configuration and deployment choices are available so you can choose the deployment that makes the most sense financially and operationally.

Moving to a new ERP system is a significant change for system users, IT, and the entire organization. It makes sense to explore all options and choose the system configuration that offers the best performance at the best cost. Then, plan for how your internal resources and structure will have to change to get the most from your investment, no matter the configuration and deployment. 

Ready to begin? Start with our ROI calculation worksheet.


This post originally appeared on the SAP Insights blog