Don't eat it all at once: Snack size core banking transformation

Dr Leda Glyptis 11:FS Foundry CEO
5min read

You know they say you shouldn’t start a diet and give up smoking at the same time? Or get married and start a new job? Or whatever?

It’s established wisdom that if you are about to do a thing that will stress and strain you, don’t do two of those things at the same time. We know this reduces chances of success for either of the things you are doing.

So when it comes to core banking system transformation, why don’t we apply this basic principle?

We resist doing it at all; core banking is big, expensive and complicated. So we let it run in the guts of our organisations while we do business on top. And the business keeps evolving and the system starts to strain. Both because of age but also because of fitness for purpose. It was built in a time and a place which no longer make sense for modern banking. But it’s big, expensive and complicated, so we make do. And we try to transform our business within ‘legacy constraints’ performing herculean feats of creativity that still get us raised eyebrows from pundits and challengers for being inadequate.

And then when the system is at breaking point, because it reaches end-of-life because it still calculates all pre-Euro FX rates daily, phantom currencies in neat rows, because, because, because, we finally have to swap it out. All at once and with all the risk that entails. While the precariously perched new business initiatives wobble and the world marches on.

The big bang core banking transformations are the stuff of legend and terror in banks. They are notorious as potentially career-ending projects that start off budgeting an eye-watering five years and half a billion dollars plus the absolute conviction that both those numbers will slip.

But what choice do we have?

As an industry, we keep doing it with the hope this time it will be different while knowing it won’t be.

So we avoid it for as long as possible and then go for the painful process, where we know that something will go wrong, cost more and take longer. Not to mention that by the time we are done, the world will have moved so far, so fast that we will not have solved any problems other than replacing the-end-of-life platform leaving ourselves gasping for breath and on the back foot. A place as familiar as it is unhelpful.

Surely there is another way.

If we take a moment to reflect what core banking capabilities are for and before they stop working, avoid replacing them like for like for a Next-Gen but otherwise identical solution. A little faster. A little more real-time. A little more flexible. But still a monolithic, Big Boy System. Because if it wasn’t that, how would it justify the price tag?

And we don’t stop to ask ourselves why on earth we would pay this price tag at all. Whether we want more of the same. Whether we used all of what we had and was it what we actually needed.

So we carry on buying into the shape we had when we didn’t know what we know now.

Even though we know.

Breaking the cycle of madness

What are we solving for? There are two reasons to replace your core banking infrastructure.

It doesn’t do what you need it to do to deliver on your digital vision

It works okay for most of your traditional stuff but your digital propositions are balanced precariously on top of it. That means they can only run their happy path. The linear task-by-task fulfilment of things that don’t need extensive processing (read: more than one system), historic data queries or extensive system interaction. It feels like seamless real-time connectivity to the client even though things get batched and processed in the back through creaky, leaky ledgers.

Their unhappy path - something needing to be validated through an in-flight query or calculation, a request needing real-time processing, approval or reconciliation - almost always requires a manual workaround. You need to call a human and have the human override the system because the system is not designed for the real-time seamless connectivity that the digital interface requires and the only way to get the synapses to connect is to have a human manually input data in fields. That is what is happening while people put you on hold to investigate an issue. And that’s not good enough.

How do people solve for this? Mostly they don’t. They just suffer on through

Some do try to solve this in a more lasting manner. They find a challenger who provides similar functionality to their core banking provider, but is better at this real-time malarkey, they go with that for digital work and continue to use their existing infrastructure for the other stuff. They create an architectural problem (this is the legacy burden of the next guy), they now have two parallel infrastructures to maintain and the risks that go with each, plus the risks that go with the fact that they are not talking to each other.

In case it’s not obvious so far, that’s not the way to do it.

Don’t replace old problems with new ones. You are building a new system because your old ones are not talking to each other. Make sure the new ones do. Not just talk to each other. But also talk to the old ones.

It’s reached end of life

All good things must come to an end at some point. Okay, so your core banking system has had a good run. It was monolithic and tuned to a different cycle of banking but it was good at what it did for when it was meant to do it. And it was expensive but there was no other option back then. And now it’s dying. And you resent the effort replacing it will entail and the cost and the bother. But you will go ahead and do it anyway.

Now is the time to ask why. Why default to thinking you need a new one at all? Why sigh and go buy something as expensive, as monolithic albeit less batchy? Why replace your problem with a slightly newer version of itself?

Why reduce the pain by 10% when you can remove the pain?

Technology has changed since you last invested in a core banking system. It doesn’t have to be monolithic any more and it shouldn’t be. What we now know about security and stability points against that kind of design.

More significantly, what we now need our systems to do in terms of processing, data access, dissemination and performance means that ‘monolithic’ doesn’t work.

Just because people are still trying to sell it doesn’t mean you have to buy it…

Smaller portions, more often

Here’s the thing. You don’t have to replace it all at once. And you don’t have to replace it all. Partly because you weren't using it all in the first place so you don’t actually need all the bits and partly because some of it still works fine for what you need it to do. You see, it does a lot. Most of it essential. Not all equally complicated.

Seriously. Focus on the things you need, the things the technology currently available allows you to have and the things you want to achieve your aspirations and deliver against your vision.

Keep the focus on your business and don’t buy the narrative that you have to purchase something similar to what you had, especially when you are trying to move away from existing problems.

You are running a profitable, highly complex business that needs to stand up digitally native propositions. You know you can’t do that on your existing core. But you don’t have to do that on a separate set of systems altogether.

If you take your small bites as and when you need them, by the time your old core banking system reaches end of life you will have moved on without any of the pain and a fraction of the cost or risk of any alternative.

Modular is the name of the game. Do what you need. In the order you need it. In the size you need it. That means that core banking is a business tool. And changes to it should be driven by business decisions. Not the specific decision to do a specific thing but by the decision to stop asking ‘where do I have the greatest need and fewest legacy constraints’ but actually start asking ‘where do I have the greatest opportunity and what tools do I need to go after it’.

Put your business first. Build to serve it.

Allow your business to feel the hunger again, unconstrained by what you have. Build capabilities to feed that hunger, not ‘to the extent permitted by your legacy’ but to the horizon required by your appetite.

Take one bite at a time, before things break, in line with what you need and without replacing what you had with an identical looking thing. Don’t accept your legacy as your starting point because the pain of changing it is just not worth contemplating.

Contrary to what the naysayers believe, you really don’t have to eat it all at once.

The binge approach to transformation didn’t work before and definitely isn’t fit for purpose in the digitally native age. Which is just as well. Because it was expensive, painful and didn’t steal a march on anybody.