Global CIO: When The Cloud Doesn't Win

Here is what we're hearing when CIOs say 'no' to SaaS and cloud computing.
InformationWeek Boardroom Journal - May 2010 InformationWeek Boardroom Journal Logo
Download the entire May 2010 issue of
InformationWeek Boardroom Journal
distributed in an all-digital format (registration required).

Chris Murphy, InformationWeek Anyone who writes off software as a service is being shortsighted. It's the real deal. But even big SaaS advocates say no to it sometimes. Here's what we're hearing in those cases.

Not now. This is a big one. Chiquita Brands CIO Manjit Singh compares the SaaS or on-premises app decision to leasing or buying a car. To extend that metaphor, the "not now" decision is when that 6-year-old car is paid off: It might not have every latest feature, but life with no car payments sure is sweet. (Granted, software's annual maintenance fee is pricier than an oil change.)

SunPower CIO Jorg Heinemann is no SaaS skeptic, building out enterprise-scale IT architecture with a SaaS-first mentality. But he keeps e-mail on premises. Heinemann looked at SaaS, but the return wasn't big enough to justify a change yet.

Not always cheaper. My colleague John Foley writes how San Jose CIO Steve Ferguson ran the numbers and found his creaky Outlook/Office 2003 costs $1.88 per user per month, compared with about $3 he'd expect to pay for Google Apps. Our recent outsourcing survey finds most people like SaaS and cloud computing: 37% say they perform better at lower cost. But 18% say they perform better at a higher cost. That may be fine, but has any SaaS vendor given you that Cadillac sales pitch lately--that it's not cheaper but better? Even slightly cheaper might not be enough, given the potential compliance and vendor viability risks.

Every app doesn't fit the SaaS model. There's a bunch of applications that CIOs just aren't thinking about converting to services, such as graphics-intensive apps like computer-aided design and many financial and transactional apps, which tend to be latency-sensitive and hold data that companies prefer to keep on premises. It might seem like every up-and-coming software category is a shoo-in for SaaS, but consider iRise, a growth company that provides application development visualization software--and has no plans for SaaS. Oracle sold $1.7 billion worth of good ol' software licenses last quarter. Microsoft's new Office 2010 includes a substantial Web element, but the core pitch for this multibillion-dollar franchise is still around the client software. In our survey, 39% of companies not using SaaS say there's no business requirement for it. On-premises software is far from a legacy concept.

Security. SaaS vendors get that their existence hinges on rock-solid security. But they're also a bigger target--banks need guards cuz da crooks know da money's there. Our survey finds 39% of companies not using SaaS cite security as the biggest reason (tied with no business requirement as the most-often cited). This one will never entirely go away.

Governance. It's different from security. Even people who trust a SaaS vendor's security might pass for compliance reasons. For example, Global Crossing, a user of Microsoft SharePoint and Exchange for collaboration, doesn't consider SaaS versions because, as a foreign-owned company that has contracts with the U.S. government, it faces a maze of data-handling rules best met on premises. Recruiting firm Manpower, a big customer of with strong faith in its security, keeps some of its most sensitive client data in-house because of internal governance rules, says CIO Denis Edwards. It will get interesting when companies try to keep the sensitive data on premises while tapping online services to interact with that data to create the end-user experience they want.

We're in--for now. This is the most intriguing reason of all. Manpower's Edwards says much of the cloud development he's looking at is for software he thinks might someday move back in-house. So new software-supported products start in a cloud environment to get to market faster and scale up quickly if they're a hit. But if it becomes a blockbuster app, Edwards thinks it'll make economic sense to run them in-house long term.

SaaS is winning a lot these days. But it won't win 'em all.

To find out more about Chris Murphy, please visit his page.