Considerations for a successful SaaS implementation

The SaaS landscape

An organization considering implementing or increasing the use of SaaS as part of its technology roadmap should understand the SaaS landscape. At a high level, SaaS can be grouped into two categories: horizontal and vertical. They are categorized based on the business domains in which the software will be effective.

Horizontal SaaS

Horizontal SaaS applications offer functions and workflows that are not generally specific to the domain in which the organization operates. Some of the major types of applications in the horizontal category include:

• Customer Relationship Management (CRM). Examples include Salesforce, Hubspot and Microsoft Dynamics 365.
• Enterprise Resource Planning (ERP). Examples include SAP, Workday and Oracle.
• Project Management. Tools include Asana, Trello and Monday.com.
• Human Resources. Examples include Workday and Workforce.
• Unified Communication. Examples include Teams, Slack and Zoom.
• IT Service Management. Examples include Sailpoint Identity Management and ServiceNow.

When implementing horizontal packages, especially those that provide standard functions such as HR or accounting, it is important to recognize that the details of standard workflows might differ among organizations. Change management and training will be required for a smooth transition.

When an organization is considering a SaaS implementation for project management and communication tools that will be used by most members of the organization, standardizing one tool and getting user buy-in is critical. An example of this is an organization where employees continued using Slack because of personal preference, even after MS Teams was implemented.

Vertical SaaS

The vertical category includes applications that apply to specific verticals or organizations within specific domains. They are niche products that address specific needs of the industries to which they are geared. Examples include:

• Healthcare. One example is electronic health records management (EHR) software, such as EPIC and Cerner.
• Utilities. One example is geospatial systems, such as ArcGIS and Small World.

When an organization is considering implementing a vertical SaaS, thoroughly documenting their requirements is critical. This helps the organization choose a package that closely meets their needs and has superior implementation support.

Considerations for implementing the SaaS delivery model

When an organization is considering implementing SaaS, it is important to review the pros and cons of the delivery model. First, consider how choosing SaaS fits into and can potentially advance the organization's business goals and the technology strategy that supports those goals.

Some organizations, especially those with unique requirements, might choose to build custom applications in-house. Other organizations might choose to buy applications, implementing a SaaS model. Other strategies include a mix of buy and build, often choosing horizontal SaaS for the functions that maintain the business, such as HR, and building their more niche products.

After determining how the SaaS model fits into the overall strategy, review the general pros and cons of the SaaS model, then address specific considerations as they apply to the organizations' domain.

Pros of SaaS

The pros of implementing the SaaS model include:

• Lower total cost of ownership. This includes the subscription fee for the product, as well as the implementation and maintenance costs. The cost is usually lower than custom builds.
• Faster implementation. This involves coordinated updates, managed installations, SLAs and out-of-box integrations. The speed of the implementation process is especially important when issues with the current application or the total lack of an application impedes business workflows.
• Ease of scalability. This is also important, especially when an organization has made an acquisition and needs to quickly manage functions such as HR or work management for a much larger workforce.

Cons of SaaS

The cons of implementing the SaaS model include:

• Internet dependent. A reliable internet connection is required for SaaS. If the connection is lost, the application will be unavailable. This means that a backup plan must be in place for safety-critical SaaS applications.
• Security, privacy and compliance considerations. These must be carefully managed, especially in safety or health-critical domains.
• Less control over customizations. This is especially important for niche products that may require customization.
• Dependence on a specific vendor. Once a vendor's application is installed, it is very difficult to change vendors.

Specific considerations in choosing the SaaS model

With the pros and cons of SaaS in mind, take these elements into consideration before choosing a SaaS package:

• Understand your requirements. This includes both functional and non-functional requirements. Functional workflows are important, especially if the organization is considering a horizontal SaaS. For example, if you are considering an HR package, you will get all the usual functions, but the steps to execute them might not align with your current workflows. Consider your strategy for customizations vs. configurations. It is best to take advantage of the out-of-box configurations and limit customizations. This will save on both implementation and ongoing maintenance costs. Non-functional requirements, such as peak load times for retail applications, must also be documented and considered.
• Data requirements. Understanding your organization's data is a critical part of the SaaS decision. This is especially important when considering an upgrade to a niche product. If your data is currently complex and difficult to maintain, data conversion and transformation will require careful planning and additional time in the schedule. For example, an organization's data was exceedingly complex due to the geographic area it served. Being aware of that upfront makes a big difference in selection and planning.
• Integration requirements. It is important to consider how the SaaS model will work within your end-to-end application landscape. Consider the number and complexity of your upstream and downstream applications, as well as the data and compliance requirements of each. For example, in some industries, certain applications within an end-to-end workflow might be subject to more stringent regulations than others. In this case, it is important to consider how the SaaS model can be integrated into such a landscape. One organization decided it was necessary to separate the regulated and unregulated workstreams.
• Security and compliance requirements. One of the most important aspects of the SaaS model is multi-tenancy. Multi-tenancy is a cloud-based architectural approach where multiple organizations' subscriptions are hosted on the same infrastructure. Although this increases scalability and reduces costs, it is critical to ensure the approach aligns with your organization's security and compliance frameworks.

Choosing the right SaaS package

Once an organization has carefully considered the SaaS model, it's time to select the SaaS package. There are many application packages delivered through the SaaS model that offer similar functionality, so it is important to evaluate several options before committing to one vendor.

Consider factors such as: 

• Key features 
• Scalability
• Comprehensive security
• Learning curve
• Integration capabilities
• Vendor support packages
• Upgrades and maintenance plans

Ensure the provider's approach fits your current needs and can scale with your future growth. Thoroughly research and evaluate the deliverables of available SaaS tools. Selecting the right SaaS vendor should include market research. You will find that there are multiple products available for almost every type of application. Look at customer reviews and, if possible, talk with current customers. Schedule demonstrations and a proof-of-concept.

Finally, it is important to ensure the statement of work not only covers all services needed from the vendor and/or integration, but also gives the client the final say in accepting the system. 

Gerie Owen is a QA engineering manager at Roobrik. She is a conference presenter and author on technology and testing topics, and a certified Scrum master.