I have had a number of discussions lately with companies implementing a new project. These projects run the gamete from full SAP implementation, moving to a new distribution partner, to simple online work order entry. However, the major headaches and issues they have had with those implementations are all the same. When digging deeper into the processes they used for roll-out they all have one underlying similarity. They used the Big Bang approach to implementation. For those unfamiliar, the Big Bang approach is a term used in Enterprise Resource Planning (ERP) implementation. The hallmark of the strategy is to move to the new system all at once on a single date. I hear you groaning, Big Bang enthusiasts, and I understand that if you plan and test and plan and re-test some more, this “should” be a smooth process. In most instances though, it is not.

The major issues with the Big Bang approach is that it is high risk. Small problems become more pronounced, testing can be tough until the system is in a live environment, and there is no fall back if a major issue comes up when the system is live.

Let’s look at an example.

A multi-national company had decided to outsource its distribution to streamline the process and cut costs of distribution in-house. The due diligence was completed, and a distribution company was selected. Due to the experience of the distribution company, the multi-national decided to take the Big Bang approach to shutting down their current distribution network and shifting to the new one. They closed their active facility on a Friday and shipped all the product to the new facility to begin shipping in a week (the distribution company said they needed a week to get it all set up and out the door). You can probably guess what happened next. Things did not go as planned and long story/ comedy of errors later, it took 3 months to get a shipment right. You can imagine the customer rage and the ripple effect that had on their businesses.

The one major question to ask yourself when thinking of what strategy to use is “will a small glitch in this implementation greatly affect my customer and my revenue?” If the answer is remotely close to ‘yes’ or even bordering on ‘maybe’, reject the Big Bang and use a Phased or Parallel strategy for your roll-out.

Phasing your implementation (again we are speaking in broader terms than ERP implementations here) involves making small changes over a period of time. The implementation is phased, or small chunks moved to the new way while phasing out the old way. The old process is not completely gone all at once. This drags out the process but allows all players in the implementation to learn at a manageable pace and de-bug any issues before the next phase goes live. Minimizing impact on your customers and day-to-day business (I understand I am using a lot of software implementation terms in this post but really, these core strategies and issues can apply to so many project launches).

The Parallel approach is more labour intensive and results in twice as much work in the short term, but takes less time than the Phased approach. The benefit to this implementation strategy is reduction in risk to the client, and more options for problem solving while the new system kinks are worked out. The gist of it is that you are running both old and new systems at the same time. There is double entry, however this allows testing of the new system in a live environment and until the new system AND old are producing the same outcome to the end user/customer, you don’t turn off the old system.

Let’s look back to our distribution outsourcing example. If they had run a parallel implementation strategy, the order could have been sent to both old and new distribution centers. Both can be packed and ready to ship. If the new system produces the product shipment in the agreed upon service level window, then the new distributor ships the product to the customer and the old in-house distribution center puts the shipment aside to go to the new facility. If the new distributor is not able to meet the service levels or a bug is determined, the bug can be worked out while the old distributor sends the product to the customer. The key to this is a seamless experience for your customer. At the end of the day, your customer should not feel any pain from the changes you are making internally.

Yes, parallel systems are expensive, and the Big Bang approach is faster, but the cost of lost customers and lost revenue can be so much worse.