You need to plan the roll out of the NPD process just like a project, because that is what it is. It will need a plan, resources, a budget and a project manager. You also need to decide what strategy you will adopt, e.g. ‘big bang’, i.e. rolled out to every project in one go, or ‘little by little’.
Big Bang Implementation
This is for the brave. In this version everything builds up to ‘D-Day’ when the whole company switches over to the new process. For it to go well, the preparation has to be extremely thorough, everyone has to know what to do, and it needs to have undergone thorough testing prior to the big day. Even so there will probably still be some issues, so a big enough support team needs to be on hand to resolve any issues quickly.
Phased Implementation (Little by Little)
In this type of implementation more gradual approach is taken in order to reduce the risk of a big bang descending into chaos. This can take many forms and there is no ‘right’ way to do it.
However one fairly sensible plan is to select a number of projects to be the guinea pigs. The selection criteria for these might be:-
- all new projects
- projects of a particular type
- projects run by a particular department or site.
- one or 2 particular projects
- Introduce different sections of the process at different times, e.g. the front end, the middle and the launch phases (this happens naturally if you start it with new projects, but you may have to wait a long time until you get to the launch stages)
The projects that are not chosen to follow the new system continue with the old one, or perhaps a hybrid version, taking the ‘easy’ and/or the important bits.
It is always a good idea to run some pilot projects. These can trial the whole process, or separate parts of it.
My personal suggestion is to do the following:-
- Run some pilots on chosen projects. I favour having 3 pilots running at more or less the same time
- One that is just starting, to test the front end.
- One that is somewhere in the middle to test the middle stages
- One that is about to be launched to test the launch stages.
- Make the necessary modifications to the process and then release it
- Do all the necessary training and then do a phased implementation, e.g.
- All new projects
- Chosen existing projects that are at various stages. In doing this there might be a temptation to choose the ‘good’ projects. However it may be more beneficial to the company to choose the projects that are in trouble.
- Chose a mixture of stages or concentrate on one stage at a time
- All remaining projects
You will have to figure out how to deal with projects that started on one process and are having to change to the new one mid-stream. Some leeway should be allowed and plenty of discussion with the project manager and the team. The key here is to be sensible and pragmatic, not dogmatic. You don’t want to cause unnecessary delays just because of changing processes. On the other hand you do want want to find and correct any major project issues that perhaps would not have been picked up in the old process.
Gateways are the key to a successful process and I find that the best way to implement a change from the old process is to assess roughly where each project is in the new process, i.e. which stage it is at and run the appropriate Gate Review. Inevitably the project team will not have done everything that is required at that stage in the new process, and may have done some other things that are not required until later, so some judgement will have to be used.
Also you will need to decide whether to run the Gate Review that would have happened at the beginning of the stage or the one that is scheduled for the end of the stage. Personally I would say unless it is very close to the end of the stage do the one before, as it may pick up some missing activities, which will need a decision on whether to force the team to do them or not . Then do the Gate at the end of the stage when appropriate, and follow on from there. However be sensible about the judgements that you make. You don’t want to demoralise the team unfairly or kill a project unnecessarily.
By the way I am not talking about installing new NPD software as part of the implementation. Often this is not necessary and only complicates the implementation, as it becomes more about getting the software working rather than about getting the process working. Therefore if at all possible make the process mods without installing new software. If you do install new software then make sure it is flexible and can cope with a bastardised interim process for projects that have already started. The last thing you need after you have told a Project Manager, “Don’t worry about that aspect. We know that it wasn’t in the old process and we won’t force you to do it on this project”, is for him to come back with the message “Computer says ‘No’.”.
Here is an extract of the song ‘Little by Little’ by Beatrice Verdi & Buddy Kaye, as sung by Dusty Springfield. She must have been a team member during a big bang NPD process implementation, just yearning for a Little by Little approach.
I don’t know where to turn
Don’t know what to do
I’m walking on thin ice
And it seems I’m falling through
Little by little by little by little
I’m afraid Dusty is no longer with us to offer advice, but if you want to discuss your NPD process implementation, then contact us and we will try to help.