ERP Projects - How to do it all wrong

Subtitle - spend too much, deliver late, annoy everybody

I have now seen quite a number of implementation projects, some from the point of a customer, some as a consultant for the prime contractor (where you cannot always tell the customer what you think) and some as an independent consultant. Most of them have been JDEdwards OneWorld, but I have also been involved in implementations of other products. There are a number of things that I see go on again and again that cause cost and time overruns on the project, and then at the next upgrade they usually reverse the crap decisions and have much more sucess from this point on.

Make decisions quickly

It is always nice to have all the information, but in reality waiting for all the information can cost you a lot of time, and you may still make the wrong decision. You will still get it right 95% of the time if you listen to the advise you get.

Be prepared to change a decision

The quicker you admit the wrong direction and change it the better off you are.

No witch hunts

If something goes wrong, fix it, analyse why it occurred, prevent it occurring again. Get on with it. If you work in an environment where the custom is to find a scapegoat and blame them you will suffer from a total inability to make decisions as noone will take any responsibility anymore.

Customise with care

If you are customising the application for your organisation, you need to consider the implications of this for taking upgrades, getting support. You need to ask yourself, do we need to do it this way? What other ways can we do it?
With trival things like descriptions of a field on the screen not being what it was on the old system, train your users for the new one. The long term cost will be much greater than the couple of weeks the users will need to forget the old system.

Manage system changes

All systems are more stable once the changes to setup stop. It is very important that changes are documented and managed so that when you have problems it can be identified what changed.

Spend $$$ on training

Spend money on training your staff. The cost of having people struggle with the new system will be on-going. You will not get the required buy-in from your users and the system will be bad-mouthed. This becomes a very big problem with a staged implementation as you will get more resistance from the other areas that have not moved to the new system yet.

Spend $$$ on consultants

Sound great coming from a consultant, but the reality is that the fastest and most effective way to get your own people up to speed is to use experienced consultants. The key is the knowledge transfer that needs to occur. In some cases you may have the most knowledgeable people, but if they cannot share that knowledge find a consultant that can.
Once you are getting close to taking over the system yourself, send all the consultants on holiday for a week or two. This will help to identify any missing knowledge/skills.

Question anything/everything

Often it is a good idea to source the project manager and the primary technical resource from a company other than your implemenation partner. This means that when the trouble starts you will not run the risk of not knowing, as you sometimes do if it is not in the interest of the implementation partner.