Building a Cloud Cost Model for Microsoft Dynamics: How to Compare On-Premise with the Cloud

Article originally published on

As applications and supporting infrastructure age, there comes a time to decide how to move forward. The cloud is at the forefront of each one of those decisions today. With general public cloud offerings like Office 365, comparing costs of existing applications to their newer cloud counterparts is straightforward.  But for other applications like a Microsoft Dynamics ERP or CRM solution, the model is more complex due to the range of factors: the conversion of existing licensing, the alignment of support, data security and customizations, the complexity of the solution architecture, and more.

Different cloud models offered as Infrastructure as a Service (IaaS) add to the challenge of comparing the costs. Some providers charge based on incremental server sizes (e.g., small, medium, and large). These offerings may also have a transmission charge for data that traverses their internet connections. With these IaaS models, you may be extending your network topology and adding requirements to integrate the cloud applications with on-premise data. These transmission charges can also result into an organization moving more applications into their cloud ecosystem.

This example illustrates the importance of understanding all of the components of a cloud offering and selecting a model that will support both the current and future plans of an organization.

The Components of a Cloud Deployment

A private cloud is most often recommended for Microsoft Dynamics ERP for security and performance reasons, as well as for highly customized deployments of Microsoft Dynamics CRM. This type of deployment starts with a base price for the application's environment and infrastructure and then adds in the appropriate system resources and software licenses. There are four areas to assess in a cost comparison for a private cloud deployment of Dynamics:


Many customers choose to maintain their current software licensing and opt for an Infrastructure as a Service (IaaS) model. This approach is attractive for organizations that are current on their enhancement plan or already have licensing through an enterprise agreement. Others may choose to transition their licensing to a Software as a Service (subscription) model, which is ideal for organizations that have lapsed on their enhancement plan, have significant changes to make to their licensing, or are new to Dynamics applications. In addition to the Dynamics licenses, there are requirements for SQL Server Client Access Licenses (CALs), Windows operating system licenses, Microsoft Office licenses and any other third-party applications you may choose to deploy within the environment.


It's important when calculating hardware costs to take into account the environments that will be required above and beyond production. A cloud implementation should consist of servers for production, a dedicated test environment and the necessary resources for a disaster recovery (DR) deployment. In the event of a Dynamics AX deployment, you might require several more environments to support production, multiple dedicated development environments, and testing and quality assurance.


Depending on your organization's technology deployment strategies, your facility costs can vary greatly. Over my years of consulting, I've seen many random incidents, including a roof fire over the server room, water spills, the transport of production servers to avoid a hurricane and much more. Your business critical applications should be deployed in an environment that provides redundancies and security protocols for things like power, cooling, fire prevention and human and natural disasters. Datacenter organizations leverage economies of scale to provide these services at affordable prices.


If you decide to go with Microsoft Dynamics on premise, you will need support to deploy the new hardware as well as individuals with the specific Dynamics solution experience. You may also require the assistance of a database administrator to optimize the deployment. When considering a full-time team equivalent to support an on-premise deployment, organizations often run into two challenges: 1) these multiple skill sets don't often exist in a single individual, and 2) ongoing support for a typical application requires 25-40% of a team member's time after the initial deployment. Weighing the available expertise against the workload of IT projects is a key factor when comparing outsourcing versus a full-time employee.

Breaking Down the Costs:  Microsoft Dynamics AX in the Cloud

Let's apply the four areas to a real-world example. ABC Company wants to move their Microsoft Dynamics AX solution to the cloud from a classic on-premise deployment. Driving the decision is the need to offload support and maintenance of this specialized application from their IT department to create bandwidth for other strategic projects. They have 25 users and manage these users out of two locations.


In addition to the AX server licenses and 25 AX Enterprise user licenses (list price: 25 x $3,500), this environment also requires four Windows Server licenses and a SQL Server license. Depending on the version that you are deploying, you may be looking at either a SQL CAL or you may be licensing the server based on the number of processors or cores. Our sample calculation assumes deployment of SQL 2012 (list price 25 x $200). Including 16% maintenance, our year one software costs calculate to $106,500 for on-premise licensing with a continuing annual maintenance cost of $14,000.


The production AX environment for this example customer would consist of two Application Object Servers (AOS), a Remote Desktop Services (RDS) server and a database server. As with any AX deployment, there are multiple environments that are required for pre-production. This includes individual development environments for AX developers, along with test and quality assurance servers. There are additional components such as Active Directory domain controllers that are required as well, but for the purposes of a cost sample, only the core AX functions are included. Our calculations for year one hardware costs equal $103,200 with no additional costs in following years.


This environment could easily be deployed into a virtual environment; however, many of these servers will require significant memory availability, as well as a significant amount of memory in the physical hosts. All production and pre-production environments could be deployed within a 10U space, which would require a half rack in a datacenter plus any needed power and internet connectivity. If deploying on premise, your cost per rack unit will vary. Our calculation for data center, power and cooling and bandwidth costs equal $12,364 annually.


Based on this size deployment, an organization can plan on consuming 10% of a full-time employees' availability to support this deployment. The challenge, as mentioned earlier, is recruiting and keeping team members with the specific skill sets required. Our calculations assume a fully loaded employee at $70,000, plus $7,000 yearly in infrastructure support services.

Total Costs

As you would expect of an on-premise deployment, year one is cost heavy with costs reducing over years two and three. Our three-year total cost of ownership for this on-premise example totals over $346K, with $237K in year one costs.

By contrast, this same deployment in a private cloud equals $9,100 per month, or $109K per year. These fees would be less in cases where the customer retains their licensing and maintains their annual maintenance fee. By comparison, over three years, the cloud solution is less than the on-premise deployment, totaling $327K.

The cost examples utilize the Microsoft price list, accounts for the deployment of both production and test environments, as well as disaster recovery capabilities to an alternate location. Oftentimes when calculating a total cost of ownership, an organization does not include these components. Easy scalability over unplanned capital expenditures is a powerful factor for some organizations that are growing rapidly. "Always available" support is a powerful factor for others.

Beyond the decisions around the cloud model and provider, it's important to put a value to the less tangible benefits and to factor in all of the components of a deployment. The more detailed the picture you can paint, the more informed the decision you can make for your organization's path forward with these powerhouse solutions.

View the sample cost spreadsheet

Next Post