CIOs – Which Level of -aaS is Right for Your Company?

data scientist data center
Dell Technologies

As new as-a-Service (-aaS) offerings emerge, CIOs may find it more difficult to know which to consider. Choices between infrastructure-as-a-Service (IaaS), platform-as-a-Service (PaaS), or software-as-a-Service (SaaS), among others, are now vastly more complex as offerings and vendors continue to emerge. Add on the complexity of siloed data and the need to support AI and analytics use cases, and many CIOs find themselves in the challenging position of having to define their aaS strategy in a confusing and rapidly changing landscape of options. In particular, training AI models and running near-real time analytics typically requires a large and diverse data set. The creation of new data silos based on -aaS platform decisions has the potential to disrupt those AI projects, and so platform decisions must be made carefully and with an eye toward sustaining future innovation. 

A recent Dell Technologies research project revealed a connection between how an organization uses technology and the -aaS offerings that may best suit the organization. CIOs may be able to self-identify against one of three organization types (Resource Aware, IT Confident, and Advanced Enablers) and use that information to help them determine an -aaS strategy that is most likely to benefit their organization. 

The three organization types exhibited unique behaviors along the following areas: 

  • Frequency of change: How often are new technologies, applications, or other assets deployed. 
  • Skillset availability: Which IT skills are readily available in-house, which IT skills were Aware, and which skills are considered a differentiator for the organization.
  • Types of interactions: How complex are the projects and technologies being used on the teams, what is the need for cross-application analytics and data, and who must interact with the technology and manage the infrastructure.

Resource Aware organizations were likely to have a small IT staff in relation to the size of their organization. Those interviewed typically employed IT generalists who were charged with keeping productivity applications up and running. There was also a low frequency of change – this group did not experience daily or weekly application or machine deployments.  Using data for automated decision making and using AI or data science was not proactively mentioned by the organizations.  Where security and organizational strategy allows, Resource Aware may benefit most from SaaS offerings.  Software offerings at this level remove the burden of management from the IT team. 

One drawback to using many SaaS offerings is that they can be siloed. This means that data created or used in one application is frequently isolated from that of other company applications.  Without a way to bring the data back in-house or a data centralization plan, this could impact a company’s ability to use data for automated decision making in the long-term.

IT Confident organizations were most notable for retaining centralized infrastructure control by the IT team.  They perform frequent application, machine, or other deployments and infrastructure changes. While end users were not able to directly deploy to the infrastructure, these organizations did give employees the ability to request applications, machines or other infrastructure through catalogs or service requests in a semi-automated or automated fashion. In this group, an application developer or data scientist would request a deployment and then manipulate the applications, databases, or other workspaces within the IT-deployed machine.  Where organizational strategy allows, this group may be best suited to PaaS. PaaS provides IT with the control they need over the applications and data that is deployed within the offering. The underlying maintenance of middleware, operating systems, virtualization layers, server and other infrastructure can be offloaded to a trustworthy provider

When considering the drawbacks to this approach, consider how the provider will ensure compliance with security standards, service levels, and how your organization will approach hosting on-premise applications not suited to a virtualized platform. Also, be sure to get a good understanding of the long term and ongoing costs of hosted VMs and containers. Many companies are unpleasantly surprised by support fees. Again, siloed applications and data may create long-term challenges when pursing automation.   

Advanced Enabler organizations exhibited a need for frequent deployments within their infrastructure by a wide variety of end users. The control was no longer centralized through IT. These organizations had advanced software development practices and, in some cases, data science practices. Several mentioned that they use data to drive business decisions and business value. Some organizations in this group mentioned the need to support edge deployments, enabling application developers to deploy updates directly to point-of-sale and other devices. This group may be best served by IaaS. IaaS allows the organization to offload some of the hardware and virtualization maintenance without giving up the ability to granularly control the operating system and runtime environments.  Most IaaS providers also offer modern platform deployment abilities, supporting container and function runtime environments alongside infrastructure control. 

In cases where security, governance, or organizational strategy prohibits the consumption of hosted -aaS offerings, on-premises data centers as a service (DCaaS) offer a unique opportunity to reduce the IT burden and transition from capital expenditures to operational expenditures. DCaaS offers come with different models of support, with everything from fully managed modern platforms to simple hardware management where IT retains full control. 

Consider if your organization’s behavior and needs fit more closely with one group than the others. Are you most closely aligned with Resource Aware, IT Confident, or Advanced Enabler? Think about your desired end state, is the group still the same?  It is prudent to align your long-term plans and -aaS choices to the types of interactions you wish to support. 

 Learn More

Listen and learn more about ways companies are organizing their IT teams and the roles and responsibilities of devOps and developer users in The Next Horizon podcast from Dell Technologies.

Explore AI solutions from Dell Technologies and Intel.

Related:

Copyright © 2021 IDG Communications, Inc.