Systems Integrators: Danger Ahead

Posted by Alan Armstrong, VP Business Development

One of my first full-time job interviews after college was with Andersen Consulting. You know them now as Accenture (and in retrospect, it was just prescient to have rebranded Andersen Consulting before the … incidents).

In any case, my interviews were back in the early 1990’s, when Groupware was all the rage. I was not offered a job, and, again, in retrospect, I am so glad that my career took another path at such an early stage. Some of my friends at the time spent several years toiling away at Andersen. For the most part they became jockeys of Lotus Notes.  At the time Andersen Consulting had made a massive “Change Management” business that usually involved a Lotus Notes implementation. A more skeptical comment might be that the “Change Management” was simply a cover for a huge Notes implementation, but I am sure that some business value was also created in many cases. When it came to implementing Notes, however, so much of the cost of these changes involved install, configuration, and maintenance of Notes its associated infrastructure.

There are many other examples that could be named, where massive businesses have been built simply to customize and configure enterprise software for individual companies. SAP, PeopleSoft, Novell, Tivoli, and CA; each has spawned its own ecology of Systems Integrators. In all cases the SIs claimed, and still claim, to “add value” to their implementation.

SaaS has become a large potential threat to this breed of company. An example from my own market: Email Archiving. The first generation of archiving solutions brought a great opportunity for Systems Integrators to install, customize and configure policies for these in-house solutions. However, as time passed, it became increasingly clear that the ongoing management of an email archive is very difficult – much harder than is anticipated by most IT departments. 

Most IT departments think of archiving as storage, and storage is, in theory, cheap and getting cheaper. But there is a lot more to archiving than storage; archiving requires data to be collected, processed, indexed, searched, and disposed according to complex retention policies. This all gets even more difficult when certain subsets of information must be put on “legal  hold” so that it is retained beyond standard retention policies to be used as potential evidence in a legal proceeding.

To manage an archive, a system must perform all of these tasks and simultaneously manage the long-term integrity, disaster recovery, and high-availability of the data. Clearly there is a lot more here than just disks and tapes.

With all of this complexity, why would a company run their own email archive? The usual objection is that companies want control over their data, and companies are concerned about security. Large companies, it is argued, can achieve the economies of scale, so they can host an application just as efficiently as an outsourced provider.   By that same argument, SMBs, because they are small, cannot achieve the economies of scale, and so they far prefer to outsource, or consume applications on demand, like a utility. No one would dispute this. But I believe the enterprise too will turn to SaaS for many kinds of applications.

The critical issue is not customer size, though initially SaaS and other hosted applications have had issues with scaling to the enterprise. (On the other hand, does anyone question whether Google Apps could handle a company of 100,000 employees?) Over time, the critical issue will become whether the application itself is, as Geoffery Moore describes, Core or Context. In Moore’s terms, core activities are those that differentiate a company, the activities that create and sustain the company’s competitive advantages. Apple, for example, is differentiated by design.

Context activities are those activities that are necessary for the company to survive, but they do not create or sustain the company’s competitive advantage. We have a receptionist in the office, and while she does important work, she does not make our company what it is in the competitive market place. To use an example closer to home, your company uses email, but do you use it in such a unique way that it adds to your competitive advantage?

Moore says that the truly great companies will essentially outsource all context activities, and focus only on the core. Apple will not outsource design, but does it really need to host its own email system? Does your company need to host its own email system?

Which brings us back to SaaS. For too long the enterprise has spent precious up-front capital and person-energy deploying all sorts of context activities and IT projects. Some argue that large companies, because of their size, can achieve economies of scale in IT, and so it makes sense to host their own applications.

But Moore says that this is a distraction, and I agree. In 5-10 years, it will appear quaint to be hosting your own email / communications server, and many other applications. SaaS is not just a delivery model, and it’s not just a consumption model. It’s a model that allows companies to invest in their core differentiators, and leave the rest to specialists.

If this is true, what is the future of the Systems Integrator? Stay tuned for my views on this in a future post.