Legacy Systems: To Replace or to Not That is the Question
Time is the ultimate luxury in business. Yet, several businesses, including market leaders, are taking time mulling over when to replace their IT applications.
There are obvious and not-so-obvious reasons for switching over to cutting edge IT solutions. But with all the good things of new technology, the scale of change and the temporary disruption that it might bring during the transition, along with the cost and the toll it takes on the organization and its resources can be equally big. This article examines reasons why organizations withhold plans for replacing their IT solutions, why they must replace, and when it is the right time to change.
Why the dilemma?
“What is the definition of a legacy system? One that works!” said somebody in satire. That might be the reason why managers face a dilemma about replacing them: it still works. But there’s more to it. Legacy systems have come a long way to entrench themselves almost inextricably in organizations, making managers and investors equally perplexed as to whether and when to replace them. Here’s how:
IT systems have become the backbone of businesses. They need to be up and running, constantly. Even a glitch for a day could cost a fortune. If that sounds exaggeration, you might want to read Comair’s story where the failure of their legacy system for one day resulted in a loss of $ 20 million. No wonder, even the idea of replacing the systems might send shivers.
It’s not uncommon to have black-box type systems: they work wonders, but few understand how, if not nobody. Those few geeks who coded the system or had clues about it might have left long ago, and the documentation might have vanished, too. Things become worse when these black-boxes are mission-critical applications.
It’s tempting to patch up and keep going just a little longer: web services, middleware, mid-tier platforms, virtualization can alleviate several problems and put a modern face on the ancient systems to show off to the customers; all at an attractive cost and much lower a risk – something that would entice any manager.
A lot is on the line: more often than not replacing mission-critical applications is like an organ transplant surgery. Business continuity, reputation, profitability and a lot more is at risk during the transition along with other unexpected project delays, glitches and gray areas that compound the risk and uncertainties.
It’s not just about the system: as they say it, troubles never come alone. Replacing the system is often synonymous with process re-engineering, restructuring, massive training programs, and new skill are all on top of the responsibility of keeping the show going on as usual.
To sum up, it’s the comfort of maintaining the status quo, the stakes involved in the change, the scale of change, and the cost that make the leaders want to live yet another day with the legacy systems; after all, it still works.
Here are some reasons why we don’t replace legacy applications:
Business paradigms have transformed and are shifting ceaselessly but adapting legacy systems to new era can be no less than a nightmare. The lack of skills in those dying technologies, the costs involved, and the limitations on the capability of bygone-era-systems prove to shackle managers every time they think of new products or processes. Even the giant players in the market can be seen trapped in the systems that once made them leaders; shackled; they watch agile and innovative startups nibbling away their market share.
Systems Get Outgrown
Just like office buildings, structures, policies. The problem is that most people can’t see that coming. It is easy to notice that the manpower will outgrow office building when the company is expanding, but who pays attention to the systems? Often there are restrictions on the volume of data and processing capacity of old systems. Unknown to nearly everyone, these limitations can be time-bombs waiting to explode during the period of extensive growth.
The Elephant in the Room
In every sense of the phrase. Over the years, the systems evolve to the extent that they lose their agility, require tremendous resources for upkeep, and become more of an obstacle; just like an elephant. And what about the security vulnerabilities? We better recap the meaning of the phrase ‘the elephant in the room’: an obvious problem or risk no one wants to discuss…
They Don’t Like to Communicate
In a bid to stay ahead and stay alive, organizations have to go with newer systems. But when it comes to integration, the old and the new don’t communicate so easily. At times, legacy enthusiast vendors might devote time and provide ‘glue’ code to make them talk easily with newer systems. But you might not be that lucky always.
Nothing Lasts Forever
Those who know the legacy technology might well be grandfathers already. The legacy service providers are already struggling to justify whether they should continue supporting and if yes, how long. On the other hand, young entrepreneurs are disrupting the marketplace with novel concepts. The day when your system might not be supported, or when market factors necessitate to retire them, might well be tomorrow.
As Bill Gates once said, “Business will change in the next ten years more than it has changed in the last fifty years.” Change isn’t just fast, and it will be even faster tomorrow. Replace we must, but the trick is to know when!
When to replace?
Business is a game of value: delivering value to the customers and deriving a value for the investors. Above the value comes the strategy for sustained long-term value. Whether to replace the legacy system or to carry on depends on two factors: its impact on value and its compatibility with strategy. By now you might have realized that there’s no rule of thumb for replacement, but you ought to consider it seriously in some situations.
When It Stops Evolving
When Lehman gave laws of software evolution, he foresaw it back in 1974 that software – just like managers – must continually evolve and develop new capabilities or else face obsoletion. Lehman said that it becomes lesser and lesser satisfactory, but you know that when the software stops evolving, so does the underlying business to a great extent. Now, considering the tech and market revolution of our times, lack of evolution is an assured way to extinction. Look for the signs of a slowdown in the ecosystem that supports your software. The moment you see the loss of expertise, reluctance and delays in support, maintenance and change requests for your system. Better start planning for the replacement.
When It Starts Decaying
Yes, they do. It’s a result of aging and evolution. Over the years a legacy system will gain a lot of orphaned, duplicate, redundant, and less than optimal source-code. You may wonder ‘why?’ Better ask HR department how many developers, analysts, and architects came and went. Better yet ask IT department what were standards of documentation back then. The legacy systems invariably grow fat with broken architecture in it, popularly known as ‘software rot.’ This is the point of no return. Unfortunately, from here any action (or inaction) to improve the system will lead to disasters. Get more inspiration from Royal Bank of Scotland’s story, where they faced multi-million dollar public fiascos due to their rotting system’s meltdowns.
When It Has a Huge Technical Debt
If you haven’t upgraded to version X yet, the costs and risk for upgrading to x+1 would be even higher. Successive upgrades not implemented are a kind of technical debt. And, the bigger this debt, the larger the stakes. At a certain point, it makes more sense to replace the system than to clear this technical debt.
There are many more signs, and they will pour in from everywhere: the application, its ecosystem, the employees, the environment. They all will start sending the signals. It’s not difficult to spot these signs; they are obvious and abundant! When you see them, the time has come! Call it expenditure or call it an investment, there’s no escape from replacement. It’s then your call to decide whether to wake up now or hit the snooze button.