Data isn’t king if the ‘royal’ C-suite doesn’t buy in

The role of chief data officer is relatively novel on Wall Street. While bank CDOs first started sprouting up before 2010, there’s been limited uniformity around what the title actually means. Do they have sway at the CIO/CTO/COO level? Will the CEO/CFO/board give them the time of day? Will the bank’s various business heads ask: “Why the hell are you talking to me?”

As the CTO of a global tier-one bank tells it, the CDO role only became one of prestige when the bank “got very serious about data a few years back”. Data has always been important, but who owns it? The chief information officer? The various business siloes? Or is it every man, woman and child for themselves?

The reason for the need to build out the CDO function was that the bank had just “ring-fenced a sizable amount of money to create a golden record of our securities, products and customers”.

On top of that golden source of data sat a newly built system of data governance controls: one version of taxonomies, one version of data quality controls and one structure for the business to access the data they need. 

“We’ve put a lot of focus on getting the foundation right,” says the CTO. “ If you can imagine how many petabytes of data we have across the sea of business and products we have – that work was foundational.”

But the undertaking was not exactly straightforward. 

Voice of the cee-DEE-ohh

The first three parts of this ‘Voice of the CTO’ series have largely focused on the balance between managing legacy systems while investing in new, innovative technologies. But at the heart of these projects is data … and data is never an easy fix.

The CDO of the bank mentioned above says the firm spent “a ridiculous amount of money” on an enterprise-wide data lineage project. The goal was establishing that aforementioned “foundation”. The problem they encountered, though, was one of sustainability. The CDO likens it to needing to fix the interior lights of a car, but instead you’re given the wiring diagram of the house that the car is parked in – there’s overlap, but a lot of interpretation is needed. 

If a regulator comes in, we need to ‘commercialize’ that lineage so it tells a story. Here’s what the data quality roles are, here’s where the manual processes are, here are our clearly defined data lineage guidelines. That’s the story

CDO of a tier-one bank

“Nobody knew the right questions to ask – the light in your car has nothing to do with your house,” they say. “So we did all of that work, and we were not rewarded.” 

After a rough start, the data team went back and asked itself, what does sustainable data lineage look like?

“If a regulator comes in, we need to ‘commercialize’ that lineage so it tells a story,” says the CDO. “Here’s what the data quality roles are, here’s where the manual processes are, here are our clearly defined data lineage guidelines. That’s the story.”

Data professionals are often overlooked when it comes to planning a firm’s IT roadmap. The problem that is too often overlooked is the concept of “bad data in, bad data out”, says the CDO, although they use a different word for “bad”. 

“We had conviction, but we still didn’t land it,” they say. “Now, everybody is saying the same things. If this doesn’t become religion at the top of the house – like cyber became religion – then it becomes really dangerous. We almost have to make everybody at the bank a data practitioner.”

The key, say sources, is making data management a priority with the C-suite and even the board. Data needs persistent budgets and workforces that don’t get cut due to changing environments. “We need to change the way we live – not just, ‘do the next project,’” says the CDO.

The CDO at a large asset manager also talks about “data as religion”. As a CDO at multiple companies over the past decade, they’ve seen firsthand where the role falters. They believe that so-called line-of-business CDOs – who are responsible for a particular silo’s data management but don’t answer to the firm-wide strategy – are toothless because they create fiefdoms. 

“This has been my theory: they are all good at reporting to different executive committee members. The problem is you end up with one COO of fixed income – or whatever – who is well ensconced in the details, but no one else is,” says the asset manager CDO.

Essentially, it’s tough to find religion if everyone has a different god.

The 9 to 5

As noted previously, parts two and three of this project look at how bank technologists are constantly under pressure to reduce technical debt related to legacy systems while at the same time make smart decisions to increase innovation at the bank. These projects, though, are often at odds with the boring fundamental work of improving data quality, lineage, timeliness, accuracy and completeness.

Those unsung data professionals do care about innovations in the fields of cloud, AI, open-source and APIs. But their day-to-day lives revolve around satisfying regulatory obligations, talent acquisition and retention, getting budget to solve problems holistically rather than with popsicle sticks and Band-Aids, and proving ROI for establishing a culture where ‘data is king’, if not religion. If a bank doesn’t get that data foundation in order, it can spend millions on cutting-edge tools only to reap minimal results. 

There was a time when the phrase ‘big data’ ruled data conversations: “We need more data! MORE! MOOOOORREEE!!!” In recent years, though, the conversation has shifted. There’s almost too much data, and looking at it on a spreadsheet can be overwhelming. So now the key to winning is around providing context around that sea of information.

The CTO at a large regional bank in the US says data management needs to fit hand in glove with the company’s overall enterprise architecture strategy. It can’t be two separate functions – otherwise, it’s impossible to get the most out of cutting-edge tools. This is what the CTO experienced upon joining the bank a few years ago.

“When it came to data, it was always too difficult to get folks what they wanted. And when they did get the data, the quality was sometimes dodgy,” they say. 

That led the bank to create a brand-new data management function about five years ago. The fifth and final piece of this ‘Voice of the CTO’ series will look at challenges related to systems’ interoperability. But the CTO says that if you don’t have the data management piece in place – and the C-suite has bought into the plan – technical debt increases and datasets spread like an expensive weed throughout the organization, failing to provide impactful insights while adding to costs. 

To better manage the bank’s data, it decided to enlist Snowflake to create “a modern data management warehouse”, they say. As the bank has built new datafeeds within Snowflake, it also created a new template for building feeds into the warehouse that adhere to the bank’s data governance guidelines. 

“The hard part is getting the data there,” the CTO says. “Once the data is there, we use Qlik Sense, which is very quick to use and [an end user] can stand up a dashboard in a week or two, rather than a couple of months.”

But more importantly, it actually prevents the spread of shadow IT, a pernicious practice through which business siloes find workarounds on their own rather than through the IT department, which adds to technical debt and cost. 

“It felt like we were building a foundation for a long time,” says the CTO. “And then – all of a sudden – you start to see the business value popping up all over the place.” 

But the foundation would never have been laid if not for the C-suite taking an active interest in the bank’s data management overhaul. Also, though, whether the title is chief data officer or some other fancy concoction of words that means ‘head of data governance’, what is being learned by heads of technology is that their lives become easier if there’s someone by their side telling the story of data. And, ultimately, the hope is that the story becomes more palatable to the bank’s top executives.