Sergey Nivens - Fotolia


Support for SaaS applications demands a hands-on approach

While they reduce some management burdens for enterprise IT teams, SaaS apps don't exactly run on autopilot, either. Be sure to put a proper support model in place.

Some enterprises wrongly assume that, once an application is in the cloud, they can take their hands off the wheel. But cloud still requires IT support to ensure high performance and prevent sticker shock -- and that goes for SaaS applications, too.

SaaS, on the surface, appears to require the least administrative work of all the cloud computing models: A provider hosts the application on its own infrastructure, and enterprises just consume the end product. In reality, though, SaaS can sometimes be the most troublesome from an IT support perspective.

Network support

While a SaaS application is hosted on a vendor's infrastructure, an enterprise still needs to ensure its users can access that software. First, the networking team must review role-based access controls to make sure that users have appropriate levels of internet access to the SaaS app.

Then, that team needs to check whether there's enough bandwidth to handle large-scale access. Internal network speed is typically faster than an internet connection, so when an enterprise redirects internal traffic to an external source, it will likely see performance issues. This shouldn't necessarily prevent the migration of an on-premises app to a SaaS model, but it is something enterprises need to take into account in terms of overall cost.

Application support

When an IT team migrates an app to the cloud, it no longer has to support the underlying infrastructure on which that application runs. However, this doesn't free the team of application support.

In fact, compared to those hosted on premises, SaaS applications can be more difficult for IT teams to support and customize because of a lack of back-end access. If there is a problem, IT staff may have to go back and forth with a SaaS vendor to locate and address the issue -- a process that can cost time and money to reach a resolution.

A cloud vendor provides the SaaS application but isn't there to show you how to use it.

An often-heard argument is to use SaaS applications as they are and avoid customizations. However, customization can improve an app through the use of unique features or data points; it's often surprising how customized many enterprise apps are compared to a SaaS application that comes right out of the box.

The extent to which an enterprise can customize a cloud-based application depends on the size of the app. While a large-scale, SaaS-based email system, such as Microsoft Outlook, is limited in terms of customization options, some smaller SaaS vendors might be willing to work with users to customize their software.

In general, remember that a cloud vendor provides the SaaS application but isn't there to show you how to use it. Have on-site staff and resources to provide users with that support, but be sure to strike a good balance. Dedicated personnel for every SaaS app in use could become a cost nightmare.

enterprise app deployment
Common app deployment patterns in the enterprise

Prepare IT staff

As enterprises prepare IT staff for SaaS, one of the first things to look at is which applications will need the most, and least, amount of support. This gives teams a picture of where to train and focus resources and where to outsource support. For SaaS applications that require little support, it is often more cost-effective to use consulting services, rather than full on-site staff.

For more critical applications that need a lot of support, retrain existing IT staff. As they become better acquainted with the new SaaS environment -- and are not burdened with as much infrastructure maintenance -- have the same person look after multiple applications. Again, this can drive huge cost savings, but be careful not to overload any one engineer.

Also, an enterprise that moves to SaaS should invest in new monitoring tools, as its existing tools are likely meant for internal applications. Most cloud vendors offer native performance and monitoring tools, but they can be sorely lacking when compared to third-party tools, such as Datadog and SolarWinds. With new tools, staff can better manage cloud-based applications and be more cost-effective.

Dig Deeper on Cloud app development and management

Data Center