Banks and other financial institutions

I spent the very early part of my contracting career, which ran from 1991 to 2004, suffering from imposter syndrome. I had largely got over this by the time I went to work at Barclays in 1993, but even so, I had a slight anxiety that could be summed up as "I bet they do things properly here". And I wasn't really 100% sure what 'properly' actually looked like. Well, I needn't have worried.

Alongside the project I was working on at Barclays was one that was concerned with preparing the bank for the 'millennium bug'. Incidentally, amongst the intellectually unwashed ranks of the anti-vaxxers, climate change deniers, Brexiteers, and people won't wear masks, there is a motley crew of people who claim there was no millennium bug. I can quite categorically tell you they are wrong.

Anyway, I got talking to one of the chaps on this other project, and he was telling me about some of the challenges they faced, including the fact that they had lost the source code for some of the core banking systems that were written in the seventies*. The source code is what enables you to read and amend a programme's processes, so that's quite something to lose. The millennium project was having to reverse engineer those systems. It was the only time I found myself even vaguely interested in millennium related work. (I've always quite fancied being a detective.)

A few years later I was working at another financial institution. The project for them was to migrate data from one system that wouldn't survive the millennium to a new one that would. My job was to test that that migration process was bringing everything across successfully. The first step was to ensure that the migration process resulted in as many customers and accounts as there were on the old system appearing on the new one. So far, so good.

But then I had to ensure that we'd successfully and accurately brought across the relationships between those customers and their accounts. And that was where I couldn't get the numbers to work: not every account appeared to have a customer and vice versa. As you can imagine, I flagged this up but no one seemed to want to address the problem.

I've often said that the attributes of a good test manager include common sense, stubbornness, and a lack of concern about being popular: I stuck determinedly to my guns. Eventually, I was taken to one side and told that the relationships I was looking for didn't exist; at some point, they had been lost. Consequently, the company didn't know which of the orphan policies were being paid for and which weren't. Equally, there were some customer payments that could not be assigned to a policy because no one knew which policies a customer owned.

And so it was that there was a certain sweet spot where you could have taken out a policy with this company (no, I won't name them), then never made a payment into it, yet twenty years later take in your paperwork and collect the rewards. Incredible, eh?

These legacy systems were a millstone around the necks of many companies and when I went to the Royal Bank of Scotland to test their new internet banking platform, the biggest problems we had arose from trying to 'talk' to the old systems. (This is why something as apparently straightforward as setting up a direct debit might have taken forty-eight hours to complete.)

I remember at the time, the big retail banks were very envious of the newcomers like Egg - remember them? - that sported wholly new systems with no baggage. It seemed like a change was on the cards in the banking world yet in the subsequent twenty years, none of them seems to have really dented the status quo, at least not until Monzo came along. I am already a big fan.

I'm telling you all this to compensate for the fact that my only photo from today is a screenshot of my Monzo account. Soz.

*Funny to think that seemed like ancient history at the time when it's now nearly thirty years since I worked at Barclays.

****
-12.9 kgs
Reading: 'Troubled Blood' by Robert Galbraith