On Death and Dying – The Legacy System Way

Remember Swiss-born psychiatrist Elizabeth Kubler-Ross and her book on death and dying? She outlined the five stages of grief a person goes through when dying. I read something recently about the fight to save a legacy system and it reminded me of this very book I read well over a decade ago.

In case you forgot, let me refresh your memory.

The Five Stage of Dying
- Denial and Isolation
- Anger
- Bargaining
- Depression
- Acceptance

The list was praised by some and criticized by others. It reminds me of how so many companies struggle when they prepare to let go of their beloved legacy system.

Let’s revisit those stages and add in the legacy system concept.

Denial and Isolation – My legacy system is great: functionally rich, reliable, and cost effective. I will ignore the isolated complaints and this will go away.

Anger – How could someone not love this system? Obviously they are technically inept.

Bargaining – I will be able to keep my legacy system, if I just add in the few extra lines of code needed to keep them happy.

Depression – How could this be happening? How could they not love my system? How am I ever going to find something suitable for our company, something affordable, something supportable?

Acceptance – Where is the consultant, project tool kit, something? If I’m losing my legacy system, let’s just get it over with now so I can retire.

We frequently run into companies who we consider “tire kickers”. These are companies that seem to resurface every few years to “see what is new” in the world of ERP solutions. They have old legacy systems that are fifteen to twenty years old and still believe they have benchmark technology. Yet for some reason, know enough to review the latest ERP software to review what they “might” be missing.

I just don’t understand. If your business management system is older than your oldest home appliance, how can it possibly help your company stay competitive? How can it provide your company with any edge against your competition? How can it be considered anything but a hindrance?

I am confused, I am upset. Maybe I’m the one in denial...

[scru]

at least it's something....

[jwenting]

Just because something is old doesn't mean it's no good...
Your attitude is quite typical of the young dogs in the industry (and probably anywhere) who think that high tech gadgetry is the holy grail towards solving any problem and that anything that's not bleeding edge is by definition fundamentally flawed.

Let me tell you a lesson: those systems aren't flawed. They're functional, they do what they're supposed to do, and they're chugging away generating income at almost no cost to their owners.
Sure you can spend $10 million+ to create something new that does the same but has a shiny new AJAX user interface instead of a textbased mainframe terminal and runs on a cluster of Linux PCs instead of an RS6000.
But you're sacrificing stability for looks, performance for buzzwords.

While you young dogs assume those systems are all but dead because they've been around a few years and are implemented using technology that's not at the moment state of the art, those systems are actually alive and kicking the butt off systems that were supposed to replace them but fail consistently to deliver.

All those companies by now have had projects to replace those "legacy" applications, started by managers who think just like you do.
All those projects failed to deliver on time, on budget, and within specs.
Most of them failed to deliver at all.

I've been there, done that, on both sides of the fence.

[Flannel Avenger]

Being able to get parts and maintain a system is a big part of the equation.

My answer is almost invariable, "it works fine, it meets your needs, but if it gets hit by a surge, it's all over, I can't fix it, the parts just can't be had anymore"

[azimuth0]

Perhaps all the 20-year old legacy systems are so good because 99% of them have already been thrown in a dumpster? At this point, you're really looking at the best of the best. It's easy to be nostalgic about the past and think that the programmers of yore were better, but in reality, there were just as many missed schedules, budgets, and specifications then as there are today. The real tragedy is that it hasn't gotten much better in 20 years.

[RebeccaGill]

I'm flattered by the "young dog" comment. I rarely get called young anymore and it is refreshing.

I'm not saying ALL legacy systems are bad. I'm saying companies who know they have issues, but can't let go of they system driving those issues, have a fundemental problem. This is what I take issue with from a legacy standpoint more than anything else. This is what I do not understand.

If your company and your industry can allow for an older system, then "have at it". Most small to midmarket companies do not have this option.