SAP reality check

I know very little about the States new system except that it doesn’t seem to be working terribly well at the moment.

 
Despite all of the positive statements issued by those who should know better, the cracks can no longer be papered over. SAP just isn’t working as it was supposed to be.

 
Let me now make a number of assumptions.

 
The system itself has been widely and successfully sold so it must have some considerable merit.

 
It is a bespoke system and must therefore be tweakable to fit the circumstances.

 
It must on the whole be suitable for the job in hand otherwise we wouldn’t have bought it.

 
So, if the system is basically sound, why isn’t it working?

 
All change is difficult to implement, especially big change. As a generalisation we do not like change unless it’s clearly hugely beneficial and involves us in the minimum of effort. Unfortunately neither of these apply to this new system.

 
There are long term benefits but mostly they happen somewhere else. For the majority of people, the change is difficult and involves more work so it was bound to be a tricky implementation.

 
It is said that some time ago, Jersey was contemplating a fundamental transport change to come more in line with their continental neighbours. Obviously changing the side of the road people drive on would be too big a move to make in a single step so they decided on a gradual implementation. Month one would see all of the lorries and buses make the change with everyone else following once the big boys had got comfortable with the new system. Fortunately, someone mentioned the idea to a Guernseyman and it never got off of the ground.

Some changes must be all or nothing but is this the case with SAP?

 
The structure of any new system project is broadly similar. Buy/Build, test, tweak, retest, train, pilot, re tweak and retrain, repilot until mostly correct and finally implement.

 
Most of this is relatively easy because you are dealing with trained and specialised professionals. However it’s only when you take the system out into the real world that you really begin to test it properly. You see, the users don’t have the familiarly with the system that you’ve built up and they assume nothing. Some might not have even used a computer before whilst others have cheerfully been using other computer systems for years but which are unlike this new system and in their eyes obviously superior.

 
Training is everything and I suspect this is an area which has turned out to be poorly implemented and rushed. But it wasn’t meant that way. I’m sure the system was piloted in one or two areas and carefully nurtured into being, thus proving it’s success.

 
However, pilot studies are usually carried out in areas where there is a good chance of success as it’s just too embarrassing for a pilot to fall flat on its face. Pilots are also sometimes overstaffed and underworked for the same reason. It’s a fair trial of a new system but it’s not a stress test.

 
A stress test is where you throw the kitchen sink at it and see what happens. Choose your worst scenario then double it and see if the system copes. That way you’ll see where the cracks appear.

 
Fortunately Guernsey has done just that. Unfortunately it’s done it in a real life situation and throughout the entire States system and it hadn’t worked.

 
I know I keep repeating this phrase but it doesn’t seem to getting through to some people; that or they’re in denial.

 
SAP doesn’t work as envisaged and it needs to be stopped.

 
Yes, it’s a huge loss of face but the time has come to face the music.

This system has been poorly implemented.

 
We need to either go back to our previous methodology or, if we’ve burned our bridges, devise a temporary bypass to get rid of the log jam created by this failure.

 
Most importantly, we then need to learn from the mistakes.

 
The system is, I assume, basically sound and will work. It might need a major retweaking for some areas and it may not ever work in others for reasons which are only now apparent. This we have learned from the failure so we have gained a little.
 
Basically you now need at least two teams. A well staffed team of firefighters to clear the backlog and a smaller specialist team to work out everything that went wrong. No sugar coating and no protecting reputations; tell it like it is and learn from the mistakes.

 
Then start again.

 
Start with the department most badly affected by this change and flood them with support staff until they are able to use the new system without training wheels. Leave a small dedicated support team in place to pick up residual problems and move the bulk onto the next department.

 
Keep learning from the mistakes and failures which will keep popping up and keep changing the overall implementation policy to take these into account. Some problems will be department or section specific but others will resonate around other areas as well.

 
All of this will be much harder now, as you’ll be faced with a mindset which knows the new system is rubbish and that it will never work because they’ve seen the proof. That is one of the prices of failure and we now need to prove them wrong. 

 
Show the benefits and show them how using the system will make their lives a little easier.

 
And here is where we hit what I suspect might be the biggest problem because the system doesn’t make everyone’s lives easier.
 
In fact it makes things a lot more difficult for some people so that some small savings might be made elsewhere. In some cases it ends up costing people a lot more time and stress (neither of which appear in a spreadsheet) to show potential bottom line savings at departmental level (where it obviously counts to those who matter).

 
If this is the case, then it needs changing. Moving work loads elsewhere does not save money. Getting all individuals responsible for inputting their own data on a daily/weekly/ monthly basis does not save money. It is spending their time away from their primary duties, which might be totally incompatible to sitting in front of a screen, to save someone else doing the input.

 
But think of all of the time and money saved. No pieces of paper or scribbled time sheets, no input clerks, much more streamlined efficiency. Ok, there may be some teething troubles while people get used to the system but nothing insurmountable.

 
Wrong, all wrong.

 
People will still need to keep their own records of time spent and others will still need to check them. Your data clerks are now your specialist workers who aren’t doing their primary work because they’re spending twice as long either in front of a screen or trying to sort out the problems created. It really isn’t streamlined and it definitely isn’t more efficient.

 
Maybe it’s time for a reappraisal.

 
Maybe it’s time to face up to the reality of all of those unpaid bills, potential court appearances, frustrated suppliers and staff.

 
Maybe it’s time to face up to the fact that SAP just isn’t working.