As more software vendors provide their services on the Cloud, and more people consider SaaS offerings, it’s worth noting that SaaS vendors will soon face the same kinds of problems that corporate IT has faced day-to-day for many years now.

As a person who has worked in IT for many years, these crude drawings represent merely a fraction of what can go wrong, and they are intentionally high-level just to illustrate the challenges that SaaS providers will need to consider, as will their customers.

1. Data Loss

Data loss is the #1 problem that IT tries to avoid.  Data loss comes in a lot of forms: viruses that destroy data, hardware failure, accidental deletion of files, or malicious removal of data by some other means.  Data loss also comes in the form of data that should’ve been protected (such as sensitive company data) that ultimately gets exposed to someone that shouldn’t have access to it.

Methods of mitigation:

• Data backup/retention policies
• Anti-virus/anti-malware protection
• Security policies (detailed firewall documentation)

A lot of SaaS providers actually don’t provide many details on how they handle data backup and retention, or details on security.  As more people move to SaaS, though, the potential for data loss or accidental exposure will increase in turn, and this topic will become much more crucial.

2. Service Outages

Service outages are situations wherein end users cannot access whatever IT services you are providing.  Data loss can actually be a reason that a service outage occurs, but there can be other reasons as well: unexpected network hardware failure, power failures, or errant foundational network services (DHCP, DNS, or Web Server failures).

Methods of mitigation:

• Redundancy, redundancy, redundancy
• Network monitoring/alerting tools
• Service Level Agreements (SLAs) to set uptime expectations

Probably the #1 thing that SaaS providers will be asked to produce is an SLA that explains possible downtimes and how long services might be out in the event of a failure.

3. “Problem” Users

There are some people that just have inherently bad luck when it comes to anything computer-related.  Some people, however, are actually intentionally difficult on computer software and hardware.  These people may have different motivations: they don’t want to do their real work and they make excuses by breaking their PC, they honestly believe they know what they are doing (but don’t), or they are just difficult people who like to cause trouble.

Methods of mitigation:

• Helpdesk service SLAs – set expectations regarding how long a ticket will be open
• Building self-help tools that preface submitting a help request
• Building services that prevent intentional abuse

As SaaS providers considering offering integrated, live-person helpdesk (via instant messaging chat), having a front-line of support to intercept problem users will be key to keeping support costs down.  User support will make or break customer’s satisfaction with their SaaS solutions, and spending too much time on the “problem” users may lower overall response time, and ultimately kill profit margins.

Filed under: Cloud Computing, Enterprise Web, SaaS