System iNetwork

November 27, 2007

Evidence Grows That App Modernization Tops App Replacement

The assertion that modernizing your applications is a better investment than replacing them outright recently gained a little more credibility in the mainframe world. I suspect that's likely true for at least some parts of the System i market as well.

A recent study by Forrester Consulting conducted for Software AG, maker of Adabas (a data-management solution) and Natural (an application and deployment environment), both mainframe tools, found that legacy modernization would produce a five-year return on investment (ROI) of 331 percent, with payback achieved in less than four months in a composite case study of several representative enterprises. Titled "The Total Economic Impact of Maintaining Adabas and Natural," the study went on to report other findings about Adabas and Natural specifically that don't concern us but also noted a few other things that at least seem to apply to the System i.

I'll be the first to admit that the mainframe world and the System i market only touch at certain points and that it may be overdoing it to read too much into studies and surveys about Big Iron shops. However, at the same time, looking over the history of just the past 10 years in our market shows an accelerating convergence between these two universes that were once starkly separate. The two platforms lately have come to have a number of similarities. High-end System i models have mainframe horsepower, and enterprises that were once considered too large to be part of the System i market are finding it just as economical, if not more so, to operate a cluster of i5s instead of one or two System z-type machines. So the platforms' markets are overlapping more than before, a fact acknowledged by IBM's own recent reorganization to key its sales efforts to potential customer size rather than specific platforms.

What's more, there's a mutual theme of "relevancy" that's plaguing both the System i and its larger counterparts. Large businesses have been tinkering with applications running on the System z for years, and smaller businesses have been maintaining applications running on the System i for a similar period. Now both kinds of enterprises are starting to face pressure to consider whether "the way we've always done it" is "the way we should continue to operate," and application modernization is one of the touchstones of this debate. (As it should be, really. What workers don't have to pause now and again to sharpen their tools? Information workers are no different.)

Enterprises interviewed for the study include a European branch of an automobile manufacturer, a U.S. health-insurance provider, an agency of the Australian government, and a France-based financial-services firm. Such organizations could easily be System i rather than System z shops. Participants were considering modernization because they feared their existing systems were becoming outmoded.

In addition to the hearty endorsement of "update and extend" application-development strategies as opposed to the "rip and replace" idea for most organizations, the study found that many users at these enterprises were basically happy with such elements of current systems as transaction speed, reliability, and ease of use (at least, with regards to their personal familiarity with using the software). Most users would be satisfied simply with improvements to UIs (such as browser accessibility) and better integration between existing applications. They were wary of risks involved with having to deal with learning to navigate new systems that probably wouldn't have all the bugs worked out of them at first.

This sounds pretty familiar, doesn't it? You probably have a fair amount of anecdotal evidence that many of your end users have similar thoughts.

I wonder if, in some regards, the modernization debate isn't just another expression of a philosophical divide, namely the difference between the idealists and the pragmatists, that can make it hard for us to come to terms with other aspects of society. Some folks at the top who don't necessarily use every day all the applications that make their businesses run are sometimes the first to react to the emotional appeal of the "let's start over with something new" approach, while many of the worker bees are thinking, "if the software I have to use to get work done just did this and this as well as what it already does, it'd be good enough."

At some companies, application modernization means there are advocates for tossing the System i onto the scrap heap and opting for reengineering the business on a new platform.
Perhaps now that the modernization debate has gone on a while, this study is more than just an interesting factoid. It's maybe a clue that conventional IT wisdom has already started to turn against the "rip and replace" philosophy. If that's so, rather than a threat to the System i, isn't legacy app modernization really an opportunity to maintain the platform?

Common sense tells us that evolution is generally easier to achieve than revolution. There are plenty of packaged tools that can help System i developers with little experience in modernization jump-start the process. Factor in the expenses of creating or buying new systems and software, training people to use them, and living through the other hassles of adapting people and processes to them, and it's perhaps a bit less surprising that the Forrester study found such a dramatically better ROI for legacy app modernization over replacement. Those are also powerful arguments that it's usually more cost effective to make the best of what you have rather than trying to reinvent the wheel, and that idea favors keeping System i machines in System i shops.

Posted by at November 27, 2007 1:19 PM