AMROBANK PROJECT 1982/3

After the Ericsson project I did a major, year-long task force for Philips Banking Division that involved moving a new banking mini-computer system from Philips Research Laboratories in Paris who didn’t want to lose it, to Philips Banking headquarters in Stockholm who didn’t want to receive it because it wasn’t invented by them. At the same time we had to deliver a large quantity of these same systems to three customers in Switzerland, Holland and France.

Just by chance the British company, Halifax Building Society, asked for bids for a first-in-Europe system of through-the-wall cash dispensers. And Philips Data Headquarters decided not to pass the responsibility for bidding to Stockholm but to handle it themselves snd asked me to assist them.

All of this came to pass. I set up a test centre in Holland and checked each configuration before it was shipped to the three banks, Philips got the order for the Halifax business that Stockholm was only too happy to handle, even though it was for the French systems, because it was the biggest order in Philips banking history and they were glad to get the kudos and the commissions.

My first assignment for Philips was considered a success, and it became doublly important because it led to my being offered a contract by the Philips Dutch Sales office to implement Amrobank’s first network. When I signed the deal, I demanded to clarify the strength of the position. I said I wanted a guarantee that I would be be supported if things got hot which they usually do in projects, and especially on banking projects that have failed once already. I got the guarantee, they said they knew exactly what I meant.

The situation was that Amrobank had earlier signed a contract with Philips for the development and delivery of a network based on X25 software. A joint bank and Philips team had just completed the development part of the project when Philips withdrew their X25 support. It was a tragedy for Amro. They had asked advice from a hot shot consultant who advised transferring to IBM’s new system network architecture SNA. It just so happened that Philips had just completed SNAPPI its SNA product, so Amro decided to give Philips another chance but insisted on a contract with very tight deadlines and huge penalty clauses. For example, the schedule was 3 months implementation, 3 months testing and 3 months installing branch systems at a rate of two per day. And the penalty was that Philips would have to pay all the costs for the new development which is something they didnt usually do. Philips signed me up to make it happen.

At my first visit I was introduced to the man selected by Amro to be their project manager. He admitted he had never been a project manager nor ever been on a computer project, but he was confident of his ability to perform the job, he said, because he was from that part of Holland where large scale farming was concentrated and that he felt that experience there would stand him in good stead. I wondered what planet he had been living on, but said, kindly I think you’d agree, that I was sure that between us we would make a success of it. We decided to alternate as Chairmen of the status reviews, and he asked if I could agree to him taking the chair at the opening meeting, and I agreed.

The bank computer centre was a new modern building without windows that could open to let in fresh air. The conference room was an internal circular shaped place that, like a night club, was heavy with the smell of cigar smoke. That’s because in Holland in those days it was a macho thing to smoke, in particular cheroots or small cigars, and support Ajax or other football team. So I was wary of the first meeting and how I would perform in that hateful atmosphere with a virgin customer project manager.

I was right to be wary. The room was horrendous for people with allergies and the Amrobank project manager opened the meeting by making two announcements that he hadn’t the decency or sense to have told me. The first was that the bank’s SNA expert had resigned and that an advertisement for a replacement would be placed in the newspapers soon. The second was that Amro had not yet decided which network control system they wanted. Then he proceeded as if to continue the meeting according to our agenda.

I stopped him, by saying “Hold it, you do understand that you have just announced two project showstoppers, don’t you”. He said “That is Amrobank responsibility and it will be addressed shortly”. I said “No, its not Amrobank’s responsibility. Its our responsibility, yours and mine, to get the system designed in three months so we had better replace the SNA specialist today or tomorrow from a consultancy and decide what network control to use right now;  there’s only one that can be implemented in the project timescale so we’d better chose that one.”

He got hot under the collar and pompously stated headmaster-like “No it’s an Amrobank decision”. I spoke again, in a fatherly style, and said “What part of this don’t you understand? In farming terms, you are taking a lorry load of sheep to market when a lion jumps in and starts eating the sheep. You say you are going to phone the farmer to ask him what to do, wait for his answer and then do what he advises. I say to hell with that, we grab the rifle and shoot the lion, as fast as possible.” He closed the meeting immediately and I guessed went off quickly to cry on the shoulder of his manager. I was right. Within the hour I had a phone call from the Philips manager telling me that Amrobank had complained about my what they called outburst. And that Philips had replied by asking “Are the dates in the contract deadlines or not? They were told “Yes they are, and our jobs depend on them being met”. Then Philips had said “In that case listen to him, we got you the best man in Europe for crash projects so listen to him.” The Amro man had said ” OK, but ask him to be more polite in future’ please.”

After that first meeting and the outcome, I realised that from now on I would have to run the project with the Amro guy following behind me, trying to look important. But in fact he wasn’t very visible after the start.

There were three outstanding features of the project that followed.
One, the Philips developers of SNAPPI, two very bright guys, had arranged a SNA environment including a dual purpose computer-based machine that could act like a mainframe to test terminals and act as a terminal controller to test the mainframe.
Two, the Amro man responsible for the installations was ex Philips and knew their weakness when it came to testing, so had inserted in the contract that Philips would set up a pre- delivery test area in which up to ten systems could be installed and put into test and kept under test for the recommended time in which components on the edge of breaking down will break down;  in other words a two week burn-in in test.
Three, we decided to set up and run an exceptionally effective help desk during installation period by having developers man the desk to handle problems as they arose.

With the correct level of expertise to advise and guide them the joint Amro and Philips software team produced the design on schedule, then implemented on schedule and then tested it thoroughly on schedule. And the installation process starting from the burn-in centre delivered all systems on schedule at two per day without fail.

In short it was a very successful project and some months later Amro invited us all to a castle for a very high-class congratulations dinner, at which they made fancy speeches praising us all on a great project wonderfully managed.

But my opposite number didn’t attend.

I was offered the job of consultant on projects for Philips Data.