The Future of Data – Is Data Management Sexy Now?

I’ve lost count of how many times I’ve heard that a new technology can “solve data quality issues”. But no single technology can address quality issues; only data management solves for it. No matter how much blockchain, cloud, and AI you throw at a data set, it will not solve for data integrity. At the heart lies source management, understanding the use case, tracking data transformations, applying tooling correctly, measurement, and monitoring. Technology is a great enabler, but it’s no magic wand.

Data management is now less point-to-point and more management across the enterprise. This is a good development and has been key to, for example, improved risk management.  However, this also means errors once confined to one process are now enterprise-grade.

AI and ML can also amplify errors and are worryingly effective in reinforcing human bias – not only amplifying data errors, but also building in discrimination. Screening CVs and calculating exam results have exposed this issue.

Alt data is now mainstream. Today, roughly half of investment firms use alt data, according to Bank of America. It’s likely to grow as more firms invest in new technology post-pandemic and as they orient investments increasingly toward sustainability.

Social media feeds, satellite imagery, credit card transactions, geolocation data, and weather forecasts are all being used to derive insights, including identifying responsible investing opportunities.

But data – including alt data – is not Lego. And big data can mean a big mess. While the growth in alt data is enormous, in isolation it’s as useful as buying a trailer when you don’t have a car. It has to connect to something or it means nothing at all. The focus has to be on collection, connection, and relevance.

When it comes to wider standardization, there are plenty of areas where data management is hampered by a lack of global standards. Environmental, Social, and Governance (ESG) data is a prime example. As the focus on ESG investment grows, so too does the need to apply standards for ESG data. We are just at the beginning of ESG data – an inverted pyramid, and we are at the bottom, right at the inverted tip of standardization and application.

When it comes to data quality, issues are often rooted in how a company is organized and operated. In a 2018 report, McKinsey described culture as either a compounding problem or a compounding solution.
It can fast-track application of analytics, amplify its power, and steer companies away from risk, or it can work against the realization of insights and benefits.

Too many companies approach data analysis as a science experiment. To make raw data meaningful, it needs to be sorted, harmonized, and put into context. Hard work still needs to be done to gain sustainable and meaningful results.

Data managers hold information about the characteristics of a data set – what it represents, how it behaves, how it is sourced and processed, and what its behaviors are (timeliness, update frequency, quality parameters). What we casually refer to as “metadata” is a gross under-representation of the skills and knowledge of data management experts.

Understanding data characteristics is key to insights – it’s the gap between business need and data science application, and is what enables effective automation. Unlocking this information is key. There’s gold in those brains.

The most useful thing any organization can do is to improve data literacy (and I mean any organization: The world runs on data now. Whether in retail, insurance, banking, manufacturing, transport, or hospitality, data matters).

Employees need to understand what data their organization has, how to get hold of it, produce and read reporting, and use it to make decisions. And it’s data managers who can help make this a reality.

Data needs to become a second language, commonly understood and spoken in an organization. This isn’t about explaining error rates, primary keys, pivot tables and Six Sigma. It’s about helping people realize their full potential and make smarter decisions. The ability to ‘speak data’ will become an integral aspect of most day-to-day jobs.

COVID-19 was a stark reminder that agility and speed are key to navigating global disruption and market change. Effective data management is core to this: It helps organizations see around corners, adapt, scale up, and seize the moment. Digitization and automation are nothing without data.

While data collection, management, governance, and distribution – the bread and butter of data managers – may not sound exciting, they are building blocks for successful processes, organizations, and decision-making. So, while it may never be sexy, it’s worth investing in your relationship with data, it might just lead to something deeply meaningful.