Saturday 15 January 2011

Foreign parts

From the very beginning of our ERP selection and implementation project, it was thought that we should choose a product that could be used by all of our sites, including those overseas. The reasons for this are that with one system, we should be able to improve processes, make costs more transparent and improve the communication, analysis and decison making for the whole group. This simply makes so much sense, that it would be difficult to justify any other course of action.

At the time, every single site was using a different system, with different processes, and it was not easy to tie all the data together, as much of it was in different formats. Once the decision had been made to go with SAP, it was made very clear that the sites in the home country would be done first, then the product would be rolled out across the rest of the sites in other countries.

The project was kicked off back in 2007 and the go live was planned for the second half of 2008. It was suggested by the System Integrator that we would then be able to get the first site overseas done in the first quarter of 2009, then another about 6-8 months later, and the sites in other countries at approx the same interval. However, due to a number of other outside concerns, a decision was made that the third site was put back to just over a year after the second site, and then the others were to follow on again at 6 month intervals.

As you'll be aware if you've followed my previous posts, the initial go-live was put back, first by a month, then again and again until a date 14 months after the original planned date. Of course, as we hadn't gone live, it was impossible for the second site to do so. Their date was also delayed and a tentative date of second quarter 2010 was agreed. But as so much work was being done on the main project, none was being done for them. It later turned out that of the little work that had been done, none was really of any use, so in the end, they had to start from scratch.

There was a meeting just under a year ago between the director of the SI firm, our CEO, our project lead and the VP from the site for that country. I wasn't involved and didn't get very much information about what was agreed - I do know that the director from the consultancy sent an email with his take on what was agreed, and it was just 4 bullet points! I was told that he would supply a project plan for the implementation at that site, and that they would be going live on a date just after the middle of the year.

I made myself a little unpopular at that point - I pointed out that the date that he had indicated was actually during a period when the majority of the staff at that site would be away. They have a number of national holidays and various events, and to set a date for a go-live then would have been pretty darn silly. As it happens, once the VP concerned got his copy of the email, he made exactly the same comments. It appeared he had made that very point during the meeting, but it seems that this had been ignored. It took a while but a new project plan was eventually drawn up, this time with a more sensible date, and this was distributed.

So the new plan was that this second phase would see the site go-live at the end of October 2010. I had a few reservations about that, as I knew that there was a lot of work to be done, and altho the site is smaller, and there would have been slightly less work, as they have fewer staff to spare to do the tasks, it seemed that the plans were optimistic at best. This was not made any easier by our internal project team. Most of them were not that willing to spend as much time working on the project for another site as they were for their own site.

I spent a few days at different stages over there, and I tried to keep in contact with their people. It became obvious that they were falling behind very badly almost from the get go. I tried to see what I could do to help, but it just seemed that as we had experienced with the project for the main site, nothing seemed to stay on schedule. There was little communication, and we had to try to get information for ourselves on the progress.

There was a meeting organised last September - the director for the SI apparently made a statement that we were on schedule and the site would be ready to go live on the date agreed in October. When I heard this, I couldn't believe it - I went into the CEO and told him quite bluntly that this was garbage and showed just how little work had been done. There were some very tense audio conference meetings at which I was accused of being an "Organizational Terrorist" by the consultants. (I'm not quite sure what they meant by it, but it was obviously not intended to be complimentary.)

As it happens, it became abundantly clear that the second site is not ready (and they still aren't). The consultants working on that site were still making config changes right thru to Christmas, and none of these have yet been transported. No data has been extracted from their old systems, so of course it hasn't been loaded. None of their staff have yet had a single training sesson - and as a result not one piece of testing has been carried out. The consultants were complaining that the language packs had not been applied as there were several errors. We have since identified that in fact all of these are in place and working - the missing text is all stuff that they would have to add manually.

At this stage, I don't think that we have a date for the go-live at the second site. Someone did suggest May this year, but I can't see that happening due to the amount still to be done. We are then almost into their main summer holiday period, plus the national holidays again, so I have a feeling that we could in fact be looking at September or October of this year before they will be ready. In other words, they will be going live about a year after the date set for them as part of the revised project.

Now I have been trying to understand why these dates are so clearly out of sync with the project plan, and I feel that I have something to add. I'm sure that what is happening is that the consultancy are using a generic plan that says, X many days for step 1, Y many days for step 2 and so on. This is fine if the allocated amount makes sense and also if they then actually communicate that information so everyone knows what is required.

I am going to say that I don't believe that any meaningful communication is taking place. The reason that I say this is the consultants doing the work didn't seem to know about the stages and had no idea of when anything was supposed to be completed by. But for me the biggest issue is that it appears the director of the consultants is not even checking to see if the work has been done. His plan says that stage 4 should be complete on day 71, this is day 72 so stage 4 is now complete. If it isn't, he just ignores it and carries on to stage 5.

I'm sure that for some people, this makes sense - but I really cannot see it. If the work is not complete, it is not complete. I accept that if you have a fixed date on a project, sometimes you have to make a decision to trim the workload to make sure that you achieve that fixed date. But in this case, trimming the workload is not an option as that means the product won't work - and it then makes sense to make sure the work is complete even if the date has to slip.

So I suspect that I'm going to be doing some travelling later this year - going to have to dig out my old language tapes to brush up!