FINANCIAL CRIME DATA

Why can’t we answer a simple question?

In twenty years of working with financial crime teams, there’s one problem I’ve seen more than any other:

Firms struggling to answer seemingly simple questions about their clients.

For example, ‘How many clients do we have?’ sounds like the simplest question, but many banks can’t answer it due to poor data.

 

Why is it so hard?

In commercial banking, complexity is the norm - hierarchies, legal entity structures, international groups and so on. Some connections are obvious from public data; others, like personal associations, aren’t. Cross-border firms face even bigger challenges. The result is inconsistent, unclear, and often inaccessible client data.

New regulations add to the problem. In one case I remember where a new rule meant a critical legal document had to be on file, one firm’s onboarding team had no idea which clients had it or how it was stored. No one had ever needed to standardise its filing or naming convention previously so the related data points had not been captured.

Sanctions are another example. When Russia invaded Ukraine in 2022, firms had to scramble to answer: What’s our exposure? Very few had the data at their fingertips.

And these problems aren’t limited to risk and compliance.

Function heads signing off budgets need to know: How efficient are we? How many cases can we handle with fewer people? What happens if we add 10,000 new clients?

New business opportunities bring still more questions, for example: If we were to add this number of new clients to our book, what impact would that have on operations?

Struggling to find figures
Struggling to find figures
‘We’ll find it when we need it’ won’t cut it any more

Boards and regulators expect clear, immediate answers. One board member I worked with was convinced he knew how many clients his firm had. His figure was widely quoted in meetings and he had no reason not to believe it. Until we checked the data – when it turned out to be wildly off.

After investigation, we found that different teams defined ‘client’ differently, data was duplicated, and records were inconsistent. There was no reliable source of truth.

After some tricky conversations (and a lot of data-cleaning), we fixed it.

At another bank, a remediation backlog of corporate clients was thought to add up to 400 cases. Digging into the data, we found three separate lists of around 400 clients. Some of these overlapped, but many didn’t. In fact the real number was closer to 1,000.

Why does this happen?

In my experience, issues like these boil down to three things:

  1. Lack of transparency – The links between systems and databases are unclear. There is poor accountability, and no single owner of the data.
  2. Lack of clarity – Data fields mean different things to different teams. A ‘renewal date’ might be a deadline for one team, but a 90-day warning to start the renewal process for another. Taken at face value, it can look like everything’s overdue when it’s not – or vice versa!
  3. Lack of control – If something goes wrong, how do you stop it happening again? Without strong data governance, firms just patch things up and move on.

No-one designed these problems into the system. They emerged over time as firms reacted to new regulations, adding new data fields and quick fixes that weren’t built with long-term use in mind.

How bad is your data? A quick test

If you are reading this, thinking, ‘Our data seems reliable – I’m not sure this is an issue for my teams!’, I have a challenge for you.

Pick one or two key metrics. Ask three people in different teams (who should know) for the latest figures. See how long it takes them—and whether their numbers match.

If they don’t, you’ve found a problem – and a good place to start if you want to fix your data.

Answer one question at a time

Trying to fix all your data problems in one go is usually a mistake. Such large-scale projects often overrun, cost more than planned, and in some cases don’t even finish, due to their unexpected complexity.

A better approach is to clean up data by answering one critical question at a time. That way, you see immediate value rather than waiting for a big-bang solution that may never land.

My advice is to get a data analyst working alongside operational people. But don’t just accept their raw numbers. I once worked with an analyst who insisted, The data doesn’t lie. And I’d explain to him, Yes, the maths is right, but your figures don’t mean what you think they do! By combining his technical skills with my business understanding and real-world context, we got to the right answers.

 

A problem you can’t afford to ignore

Recently more firms are waking up to the need for clean, accessible data. It’s not just about compliance. It’s also about competitive advantage.

Say you want to offshore operations to cut costs. In that case you’ll almost certainly need stronger data controls to maintain quality in a new or less experienced team. Or if you’re planning to introduce AI - to automate risk summaries, for example - you’ll need exceptionally reliable data to make it work without the input of experienced people.

If you can’t answer a simple question today, don’t assume you’ll find the data when you need it tomorrow. Fixing it now could save you from a crisis later.

Because once your board (or regulator) loses confidence in your ability to answer a simple question, it’s incredibly hard to win it back.

Let's make change happen.

We help Financial Institutions accelerate digital transformation – delivering improved efficiencies, better risk controls and enhanced customer experiences.