It's a phrase almost guaranteed to send chills down the spine of an IT leader: "Rip and replace."
Just like it sounds, this do-or-die strategy is the ultimate admission that an organization's computer systems have become too restrictive and inflexible -- no longer meeting business needs, satisfying employees and customers, or enabling innovation.
It isn't a strategy to be undertaken lightly, and the task involves painful soul searching about what to do with complex legacy systems and clunky enterprise applications. And it isn't an exercise for the faint of heart. These efforts are always extremely disruptive to the business and can be awfully expensive, and many efforts fail despite all the best intentions and investments made.
So why then would any organization want to throw out the baby with the bathwater and start its core technology infrastructure all over again? Simply put: Necessity.
"There are good reasons and bad reasons to do a rip-and-replace project," explains Michael Krigsman, an industry analyst and founder of CxOTalk.com. "A good reason is if there is no direct path from your current technology state to where you need to be, and that happens."
Krigsman offers the example of a manufacturing company that has grown in size and complexity over the years and frequently changes its operational processes.
"They have custom software that they built 15 years ago, and every time they want to make a change they're getting into compliance issues," Krigsman says. "Or, they buy a new company, or they add new products, or whatever it may be. It requires a very significant programming effort to adapt the system to the business change."
"That is a good reason to rip and replace, because it means you need greater agility from your systems in terms of responsiveness to the business change, and there is really no alternative," Krigsman says. "In a case like that, you may not have a choice but to rip and replace."
Out with the old and in with the new
That was the case for Union Mutual of Vermont Cos., which took on the task of replacing a deficient insurance claims processing systems.
Gary Ouellette, senior vice president of operations and enterprise risk management, says the 2012 project took 24 painful months to complete and the price tag hit "seven figures." Despite the difficulties, he says the strategy worked. "I believe the approach we took was the only way to go," Ouellette says.
A century-old insurance company located in Montpelier, Vt., Union Mutual had been running on a Cobol-based policy and claims system that was first installed in 1987. It was an old dog that couldn't be taught new tricks. The company needed to streamline processes, find new efficiencies and undertake new strategic initiatives -- none of which were possible with the existing system.
Perhaps most importantly, agents found the system difficult to use.
"Our agents had very little -- and I'll put quotes around the word 'real' -- access to our system and data," Ouellette says. "What they had was basically a Web screen that we had built that may have looked and felt like a system, but it really wasn't. It had no workflow; no interaction; no ability to go on and look at policy documents or do endorsements or anything like that. For us, that was probably one of the biggest functionality points that we were really seeking to correct."
The first step in the long road to reinvention
Union Mutual executives, including the chief financial officer (who has since become the company's CEO), first examined the various options available and the pros and cons of each. "We had looked at not only different technologies, and different vendors, but also different approaches," Ouellette recalls.
"We looked really hard at the best-of-breed and multivendor-type of approaches, and that would definitely not have been rip and replace," he explains. "That would have been integrating a piece at a time from a different vendor -- a claims piece, a billing piece, eventually a rating piece, and an image piece and a workflow piece. We had taken a long hard look at that because that was sort of the world we lived in prior to where we are today."
But company executives quickly decided that approach would only prolong the pain caused by business disruption and increase the risk of failure, Ouellette says. They decided it was ultimately less painful to rip the bandage off quickly and have a single system ready to replace it.
Getting by with a little help
While the decision to go the rip-and-replace route was made internally, Union Mutual then brought in a consultant to help with the vendor selection process. The need was for a new system that would provide full functionality now, while also scaling for future needs.
The company also wanted to be able to move data from the legacy system to the new system as quickly and seamlessly as possible. Ouellette was well aware of horror stories about other organizations unable to retire their legacy systems even after a rip-and-replace project. Many organizations are unable to move all the old data onto their new system and are forced to keep the legacy system on life support for years after a supposed completion date to the effort.
Union Mutual had neither the budget nor staff to maintain two systems.
"For us the timeline was critically important. The quicker we were able to get on the new system, the fewer disruptions we would have in the marketplace. And the quicker we could sunset the other system, the easier it is for our team and our agents to only work on one system rather than multiple systems. For us, that was absolutely key," Ouellette says.
After years of living with complexity, Ouellette's team also longed for the simplicity of a single system, along with full functionality and flexibility. After all, to go through all the intense pain and expense of a rip-and-replace effort, the end results had better dazzle everyone concerned.
"We chose a suite [the SurePower Innovation platform from ISCS] that had basically all functionally built into one [product], under one roof, with one partner. Our new system does everything -- ratings, issue, agent portal, insured portal, reinsurance, finance, interfaces, printing, image, workflow; and it's all one continuous unit, which is nice," Ouellette says.
Driving business and revenue growth
Niceties aside, what the new ISCS system really does for Union Mutual is allow for business process change, provide for greater efficiencies, reduce operating costs and enable new revenue opportunities.
"One of the big things that we've seen since the day that we've gone live with the system is double-digit growth," Ouellette says.
That -- in a nutshell -- is what a rip-and-replace effort is all about. It is not a technology project. It is a business project enabled by technology.
"When I listen to rip-and-replace discussions, I think there's a danger that it is a technology initiative, and that runs a big risk of running afoul of what the business thinks ought to happen," says Jim Duggan, an analyst at Gartner.
"The issue tends to be that 90 percent of what the business is running they're perfectly happy with. It doesn't matter if it's high risk or on bad platforms or on an unsupported version of Oracle. As long as it's running, they don't care. So if you can't relate it to a business risk, it's a tough sell," he says.
Getting sponsorship and support from the business
Union Mutual was fortunate in that the idea of a rip-and-replace effort had early support at the top.
"It's critical to make sure that the people who are in the business know and understand what the new system capabilities are, and that from the CEO of the company on down, the message is very clear on why we are buying this new technology, and how we are going to utilize this new technology," Ouellette says.
If the CIO fails at that, employees will end up using the new system in the same way they used the old one, with little real gain.
"That is one of my biggest lessons learned. Although I know that intrinsically, it's worth mentioning again and again, because when you spend millions of dollars on a new system it is so easy for key people in the business to try and make it what they already know -- when what they need is for it to be modernized and allow for new workflow and processes," says Ouellette. "That is really where the payoff is for them."
Finally, he says there were two significant advantages to adopting this strategy. "One, it got us the new technology that we need to meet our strategic goals. Again the strategic goal being the ease of doing business externally, being more efficient and sustainable. So that was really key," Ouellette says.
"The second piece -- which I believe was just as important -- is that rip and replace allowed us to do this in a much more compact timeframe, so it had minimal impact on our markets, on our external clients, on our insured, and on our agents. Those to me are the two most important points where the rubber meets the road," Ouellette says.
This story, "Rip and replace: When it pays to make a total systems change " was originally published by Computerworld.