Constant new business requirements driven by the innovative demands of many business leaders may cause dilemmas for the CIOs, materializing in conflicting targets. On the one side, IT leaders must realize the digitization of business ideas as efficiently as possible. Thriving economic success is, for sure, the dominant goal for any company. However, on the other side, the ‘show must go on.’ Any addition, expansion, or change of the running digital backbone must not disrupt day-to-day business processes. In fact, the IT leadership must ensure to keep the lights on at the actual IT system landscape and guarantee running business operation at all times. Needless to say: business applications and data networks must function at the best performance.

Changing parts of technology components, adding new capabilities, or enhancing the landscape with new software applications can be risky.

Reflecting on some of the software implementation projects I was called in to rescue, not all missions were finished with brilliant shining lights. Despite all the preparation efforts and testing regimen of the implementation teams, the final switch from the old to the new system went dark. And during its aftermath, the business was impacted over many months. I documented some of these stories about project recoveries in my book Run IT – Dominating Information Technology.

It’s not a coincidence that the following idiom offers insight into the IT leaders’ minds: ‘Never change a running system.’ The older the computer systems get, the riskier the change situation becomes. In my understanding, the system includes all computing layers comprising the entire technology stack, from the ‘metal’ (servers and physical components) up to a magnitude of UI devices. While the IT guys have to do regular maintenance activities – including software updates on OS and applications – the need for an architectural refreshment of the entire digital landscape might be needed, driven by business and data demands. Think about this: today, large corporations still operate R/2 mainframe systems that were introduced in the early 1980s. The hardware and the software have been out of maintenance for more than 20 years. Spare parts for the hardware has to be found on eBay or other third-market sources. Such a situation describes a real keep-the-light-on in our modern digitized world. But where does such a situation lead to?

I think any company running on outdated hardware or software stack corners itself in a situation that gets more difficult to manage the longer the time drags on. As such, a typical ‘solution’ for the problem would be to build and establish a new digital platform, side by side to the existing one. However, this side-by-side implementation imposes some questions.

  • Duplication of data: “Which data reside where?”
  • Additional IT efforts: “How to minimize any maintenance efforts for the old system?”
  • Confused business users: “Which system, old or new, to use when?”
  • Heavily distributed data: “How to create and run business analytics reliably?”

Any side-by-side deployment of new systems can easily cause nightmares for the company’s leadership.

This situational storyline demands an escape route from the keep-the-lights-on dilemma, resulting in a complete system replacement strategy instead of a side-by-side implementation. I admit that the total replacement of old IT might require more time to analyze and implement all business processes. But typically, business processes must be migrated to the new system anyways. This situation then leaves the question about the data, particularly old data residing within the old system. Calling it Historical Data (see Fig.1), it might be needed for regulatory or legal purposes. Therefore, one crucial step is migrating the data and its business context to a data management platform. That data management platform must fulfill the requirements of the auditors and regulators. Securing the Historical Data into a secure database, accessible to all business users and external authorities, is the escape route from the keep-the-lights-on dilemma.

Keep-the-lights-on itself is not an approach the deserves the word strategy. It is an approach to avoid decision-making, particularly on the data layer. A new digital backbone empowers innovation for all lines of business. The businesspeople can try out much-needed new omnichannel tactics. The new system allows state-of-the-art defense mechanisms against security threats and attacks on privacy. And eventually, it offers modern real-time databases supporting business models that were unthinkable a decade ago.

Therefore, my advice to all business and IT leaders: “Switch off the Old and turn on the New – exactly that will keep the lights on for years to come.”




andreas graesser









Author: Andreas Graesser, Chief Innovation Officer at innovad LLC. ( )