All Posts in Technology

|In: Architecture|Last Updated:

Dear Dr. Yeahbut …

Every time I turn around I read that IT needs application portfolio rationalization (APR).

I’m the newly appointed CIO in a mid-size (~$5B revenue) financial services enterprise. I’ve been reading the usual sources for these things and find the logic in favor of APR compelling. On the other hand(s) I find the logic in favor of another half-dozen or so management initiatives equally compelling. And I’ve been reading KJR long enough to remember your thoughts on the importance of maintaining focus (“Nailing governance,” 8/7/2006), which I find even more compelling.

I know we can’t do everything we really ought to do, which brings me to my question: What’s your take on APR? Should this make my top-three list?

You’re a consultant, so I know your answer will be “It depends.” What I want to know is what it should depend on.

– Seeker of Free Wisdom

Seeker …

Oy.

Before I can answer we’ll have to wade through something that’s frequently misrepresented or misunderstood, starting with the phrase itself.

APR is built around a metaphor – that IT can manage its application portfolio the way investors analyze their financial portfolios.

If you’re an investor, you assess and constantly reassess the funds and individual equities that constitute your portfolio. You determine their health and anticipated future value, use that insight to assign a disposition – buy, hold, partially divest, sell – to each of them, and act according to those dispositions.

APR theory says you should do think about your applications portfolio the same way – assess each application’s health, assign dispositions accordingly, and act on those dispositions.

This theory isn’t so much wrong as it is, as someone once put it, insufficiently right.

The applications your enterprise relies on to get its work done do constitute a portfolio. But (I am, after all, Dr. Yeahbut) that isn’t all they constitute.

That’s because of a fundamental and retrospectively obvious difference between a financial portfolio and an applications portfolio: With financial investments, buying or selling shares in a particular fund or equity has little impact on your investments in any of the other funds and equities you own. But in a portfolio of applications, you can’t decommission an application or replace it with another with impunity.

Beyond the portfolio view of its applications, IT also needs to consider them from the perspective of holistic design.

Holistic design is the difference between a pile of lumber, Romex, screws, nails, and so on and a structure you can inhabit. (Okay, this isn’t the best metaphor I’ve ever come up with, but let it go – it isn’t a rabbit hole worth falling into.)

Holistic design is how you figure out which collections of applications the enterprise needs to support each of its major areas of functional responsibility, how the applications within each collection interconnect to provide complete solutions, and how the collections interconnect with all of the other collections to support the enterprise-level need for information integration and process standardization.

In theory, if you’re handling holistic design well you wouldn’t need APR, because your holistic designs would determine exactly which applications you need and where you need them.

But as Benjamin Brewster once said, “In theory there is no difference between theory and practice, while in practice there is.”

Bob’s last word: When it comes to an organization’s applications environment, it turns out the holistic and portfolio views aren’t just equally important.

They’re complementary.

More than that: APR is usually chartered as a once-and-done project. It takes the application inventory and, if you’re smart, your holistic application design as inputs, and produces a remediation roadmap as its primary output.

But (there’s that word again) … there’s a lot more to the story than that. But we’re out of our self-allotted space this week, so the rest will have to wait.

Bob’s sales pitch: This is a big, complicated subject, and there’s only so much I can handle in these weekly bite-size chunks. So more next week. And if you need a more in-depth conversation than what fits into this format, you know who to call.

Comments

|In: Industry Commentary|Last Updated:

What’s Microsoft waiting for?

Somehow or other, as, over the past four decades, its Office suite of applications has evolved into a powerhouse force for individual productivity and group collaboration …

Somehow or other Microsoft never bothered to fix an annoying and easily fixed Excel “feature” (as in, it isn’t a bug; it’s a feature). Namely, just how many versions of blankness do users need (maybe 2 but I doubt it) and why are they so hard to find?

There are, so far as I can tell:

  • String nulls: Cells that have never been touched but are formatted as character strings.
  • Numeric nulls: Likewise, but as numbers.
  • The other numeric null: What you get after unchecking File/Options/Advanced Show a zero in cells that have zero value.
  • “”: Empty character strings.
  • “ “: The blank character.
  • Zeroes formatted to not display anything in the cell.

By themselves a few of these might actually be useful. What makes the bad outweigh the good is that there’s no one test that reveals them all.

Surely Microsoft could provide (for example) a modified ISBLANK() function that returns a TRUE value for any sort of blank cell; likewise it could add a FILE/OPTIONS/ADVANCED parameter that unifies the behavior of all blanks; or add a parameter to any function that handles blanks telling it to either ignores blanks of all types, or treats them all as zeroes. That would save legions of Excel jockeys from having to rely on =OR(ISBLANK(A1), A1=””,A1=” “,A1=0,LEN(A1)<1) or some such formulaic nonsense.

Why do we have this mess? I can only speculate. I imagine it’s rooted in one of the laws of organizational dynamics: When the problem is diffuse and merely annoying there’s always something more important to work on instead.

While I’m beating Microsoft up over this admittedly less-than-world-shaking grievance I have to admit: I’m guilty too. In my case the problem I’ve let fester is (was) failing to update WordPress to a supported version of PHP. I’ve known for a couple of years that the platform needed updating. But as I’m no longer a good enough programmer to troubleshoot any problems that might emerge from the process I decided to leave well enough alone.

Until this week (and let me know if you spot any glitches that need fixing – thanks!)

So I’m as guilty in my own way as Microsoft is in its way, but on a small enough scale and with such a small number of victims if it goes wrong (one, namely, me) that I can handle the remorse. But at the other end of the scale I know of Fortune 500 companies running applications on out-of-support server hardware and operating systems … thousands of them … with no path to safety because procrastination has saved them money every year for the past decades or more.

Not to mention all the other large but not enormous enterprises for which infrastructure maintenance is never quite important enough, right up until an attack takes their core applications down for a week or more.

Manufacturing mavens figured out a long time ago that preventive maintenance is less expensive than break/fix maintenance. That the IT infrastructure is harder to understand than the motors, gears, pulleys, and belts that constitute a modern factory doesn’t change the principle.

Bob’s last word: In IT, maintaining healthy platforms and infrastructure isn’t best practice.

It’s the minimum standard of basic professionalism.

Bob’s sales pitch: There’s a reason I named this blog “Keep the Joint Running.” It’s because of Principle #7 of the KJR Manifesto, which says: Before you can be strategic you have to be competent.

It isn’t the only principle worth adopting, either. Be a person. Buy yourself a copy. If you already have a copy, write an Amazon review. KJR needs your support. This is one way of providing it.

Comments