Thinking Strategically About IT As A Service

February 27, 2012 Dan Burger

Real change doesn’t come from the change machine, where a dollar will get you four quarters but it doesn’t change what you can buy with the money. The IBM midrange, on the other hand, could be in for some real change. I’m talking about the change that comes with consuming IT as a service instead of buying it as a product. Are we any closer to that happening than we were five years ago? Could be. Here’s why.

Although it’s tempting to say it all comes down to money and saving it whenever and wherever you can, that’s not the driving force for this change that I am talking about. It has more to do with eliminating complexity and staying current with information technology, which has a degree of difficulty just slightly less challenging than teaching a cat to hold its breath under water.

Let’s start with an example of how things traditionally get accomplished inside the average business when there is a request to add functionality to a business process. Typically, whoever wants this change in process goes to the IT department. There, it is determined the requested change will require the purchase of a new server and other resources. In order to proceed, a capital request needs approval by upper management’s IT committee. So you get in line with 13 other projects that are also awaiting funding approval.

Compare that scenario with the option to get the needed functionality by going to a company offering it as a hosted or managed service and at a price that provides a bypass of the capital expense process and in many cases can be operational in a week or less.

Let’s say a hosted software package costs $100 per user and there will be 10 users. That’s $1,000 a month and $12,000 per year. In five years, the cost is $60,000. Could it be done cheaper buying a server, building the application in-house, or buying software and implementing it? Maybe, but it might take six months, or maybe a year or even 18 months, to get the internal project done.

Last week I spoke with John Siniscal, whose experience in the IBM midmarket includes a view from the perspective of two application tool vendors. At LANSA, he was president for more than 10 years before retiring in 2007. Currently he’s a consultant working with looksoftware. Siniscal believes strongly in the IT as a service model and is expecting it to grow considerably in the next five to 10 years.

One of several examples of ISV involvement that he discussed with me involved an ISV with packaged software for keeping track of retail inventory. Ten years ago that company was selling its software to more than 100 customers. However, sales growth had peaked and a change was needed to survive. An application modernization project to rid itself of the green-screen interface was instigated and the business model was redirected from selling software licenses to selling hosted services. It turned the business around, Siniscal says, and proved that hosted services could be successful in the IBM i market.

“The decision to purchase a hosted service can be influenced by internal bureaucracy,” Siniscal says. “It’s much easier to get approval to spend $1,000 a month as a recurring expense that is an operating expense than it is to go to the well and get approval for a $50,000 capital purchase. In a lot of cases, people don’t look at the bottom line as much as they consider the ease of getting operational. Compare that with going through the usual approval process, the process of sourcing hardware and software, and then going through an implementation cycle. If a hosted service is priced right, it has a lot of appeal.”

Debating the economic comparisons between managed services and purchasing licenses is a favorite pastime for many IT department heads. That discussion frequently rocks back and forth on the pros and cons of budget lines labeled operational expenses (Op Ex) and capital expenses (Cap Ex). Buying hardware and software goes in the books as capital expenses and it tends to have peaks and valleys in conjunction with hardware and software upgrades; and you have to depreciate the investment over time, so your bottom looks better than your cash balance when you do a big IT installation. Contracting for managed services whether hardware or software or both is an expenditure that is fairly constant, although it takes into account predictable periods of peak demand when the average monthly service charges are raised to match the increased demand.

Usually the internal IT contingent would prefer to buy equipment and manage IT internally as opposed to managing vendors who are managing the hardware and software. But these decisions are increasingly being decided by the finance department and the C-level executives without a lot of input from the IT perspective. The CFO viewpoint often wields the most influence. (For more on that particular topic, see “CEOs and CFOs Pull Rank On CIOs In IT Purchasing Decisions” in the July 18, 2011, issue of The Four Hundred.)

The reasons for choosing to put an IT project on the Op Ex or Cap Ex side of the budget vary widely depending on whether a business is private or public and whether the executives want it to be profitable or operate at a loss. Sometimes keeping items off the balance sheet is a strategy and sometimes owning and depreciating items provides the best benefit. Even loading up debt can be a financial gain in certain leveraged buyouts. It’s not out of the realm of possibility for decisions to be made with the intent to depreciate the cost over five years, while retaining the asset for 10 years, which makes the financial executive look good while handcuffing the company with outdated equipment during years five through ten.

“It can be a control issue,” says Jim Kandrac, president and founder of United Computer Group, a company that sells and leases IT equipment and also hosts data backup and recovery. There can be internal turf wars. If the IT department sees value and takes its job seriously, the due diligence gets done. If not, a fantastic idea may get tossed in the “circular files” under the desk, because they are worried about their jobs.

“The bottom line is that someone has to be in charge and on top of things, which means taking into account the costs, the benefits, and the hassle factor.”

“We look at Op Ex versus Cap Ex as just one possible piece of the overall benefits of a DR in the cloud model,” says Simon O’Sullivan, executive vice president at Maxava, a business continuity specialist in the IBM i market.

“In the HA/DR world, if a customer chooses to buy a second server, the HA software and implementation up front is a capital cost, and a purchase probably works out less expensive for the customer in the long run over about 36 months. If the customer has admin staff already onsite, it just makes sense to train them and to monitor and manage the environment themselves and own it, especially if it’s a bigger site and the company has the skills required in-house,” O’Sullivan says. “This is also true in a stable operating environment where there is little change and the customer is committed to the existing hardware platform long-term. On the other hand, if a site does not have the time or staff to implement their own HA/DR, then a cloud solution (and Op Ex model) will be easier and cheaper. If they get a shared piece of a virtualized machine and shared data center services along with full facilities management that includes testing, for instance, then for some it’s a much better model.”

If you believe like I do that the value of IT is wrapped up in the applications, then the decision is weighted by evidence that the apps can either run best while being managed internally because of the skills and business knowledge of the staff or that the value lies with an external source that knows the applications and is in a better position to deliver the extras the best practices that can’t be done internally.

“Many iSeries environments are understaffed, overworked, and maybe under-knowledged and stuck with older technology,” says Patrick Schutz, iSeries product manager at Abacus Solutions, which offers secondary IBM i environments for development, upgrade testing, disaster recovery, high availability, and archiving. “They have hurdles to get over and have limited resources, which prevents their staff from concentrating on applications. The outsourcing aspect frees staff to concentrate on the higher levels of the business.”

As Schutz sees it, nine out of 10 times buying hardware and software and managing it internally (the capital expense) is more expensive than choosing IT as a service (the operating expense). It is, however, a matter of scale.

Larger environments get more value for less investment than they’d have when handling everything internally, Schutz says. In the smaller the environments, the savings may not be there, but they get more for the same amount invested or sometimes they pay more and get more. “The result is that they get best practices that they didn’t have before.”