Modernizing talent one of the keys to analytics success

The necessity of the CDO

It's been 20 years since Capital One hired the first known CDO. Widespread adoption of the role, however, took time.

As recently as 10 years ago, only 12% of blue-chip organizations -- large enterprises such as American Express, McDonald's and Walmart -- had CDOs. By 2021, however, that number grew to 76%, according to a 2021 survey by NewVantage Partners.

And that's a strong first step toward analytics success.

But it takes more than just hiring a CDO to achieve analytics success. That CDO has to be empowered, according to Halper. They have to be part of their organization's leadership team rather than just the leader of a team -- a move that shows the CDO is a true leader within the company rather than just a figurehead.

In that leadership role, CDOs should be responsible for developing and implementing a strategy around data that is designed to lead to value. That strategy includes establishing policies and standards around data -- which is a data governance framework -- and processes for collecting, preparing, organizing and using data.

"It's not enough just to have a CDO," Halper said. "That's very important. But we saw the companies that had a CDO as part of the C-suite where the CDO could actually get visibility. [They] had more top-line success."

Similarly, DosSantos said the role of CDO must be defined -- and valued -- and that its definition has evolved.

When organizations first started hiring CDOs, the data executives' role was essentially to protect their organization. They were there to make sure their organization's data was in order to stay within a given industry's regulatory guidelines. In a sense, they were like accountants, according to DosSantos.

But as the use of data to drive business outcomes has advanced, so too has the role of the CDO. If they're to be entrusted with helping their organization increase revenue, they need to be empowered to do so.

"Analytics isn't new," DosSantos said. "What's new is the idea of trying to utilize the analytics as a centerpiece to make money. So now there's a transition from a defensive-minded CDO to an offensive-minded CDAO. Now, we expect CDOs to help make money, to be close to the business outcomes. We're transitioning from protecting the data to harnessing the data."

And in harnessing their organization's data, CDOs should be overseeing and operating the data supply chain, determining what the data is going to be used for, and determining how it will be used to try to make more money, he continued.

More talent modernization

Talent modernization goes beyond hiring and empowering a CDO. The people working with the CDO and carrying out the CDO's plan also need analytics knowledge. That means hiring data engineers.

Data engineers are the ones who work with their organization's data. They prepare it for use by the data scientists and analysts who build the reports and models that result in the data-informed decisions that demonstrate analytics success.

They build data pipelines -- which are now ideally automated to save time and effort -- so their organization's data doesn't get isolated. They integrate their organization's data, cleanse it and structure it. They make it accessible, not hiding it in the depths of a data repository but housing it in a data catalog where an analyst can easily search for specific data and find what they need.

And they implement the data governance framework laid out by their CDO. They ensure that their organization's data is secure and its use doesn't run afoul of their industry's regulations. They also aid end users who know they're using data in a safe manner that won't cause harm.

"When you think about data as a product and the manufacturing process needed to [make data trustworthy], data engineers are critical," DosSantos said. "They're the ones that make sure that data you need is standardized, trustworthy and available on demand. What they do is say, 'Here's all the data you can trust,' and [they] make it available with the touch of a button."

In addition to adding data engineers, upskilling existing employees in how to use data is also part of the talent modernization needed for analytics success.

That means creating a data literacy program and appointing data literacy enablement teams to lead that program.

Learning data is like learning a new language, according to Halper. And data literacy is the ability to use that language to understand data, derive insights from it and communicate the results.

But many organizations aren't proficient in the language of data, Halper noted.

While full fluency across the organization isn't needed to give self-service end users the authority to make decisions using data, those end users do need to reach at least a conversational level of proficiency with data to achieve organizational success with analytics.

We're dealing with supply chain issues, a pandemic, [and] unrest in the world. And companies are realizing -- now more than ever -- that they need data and analytics to compete.

But many organizations are struggling to reach a conversational level with data, Halper said.

According to a data literacy assessment done by TDWI, organizations' average score was just a 53 out of 100, demonstrating that the vast majority are far from an ideal level of data literacy.

"There were lots of good things happening around data literacy, but they still had a ways to go -- they were in the middle of the road," Halper said. "More organizations are supporting the notion of data literacy. But there's a lot of work to be done in terms of putting programs together and the training."

Data literacy programs need leaders, and they need buy-in from the executive level, she added.

DosSantos noted that data literacy training does not mean forcing non-data people to learn coding and turning them into data scientists and data analysts. Instead, it means making sure employees understand the decisions that need to be made, the definitions of the words used to speak the language of their business, and the data that supports the decisions.

"The goal is to help someone, in the context of a business decision, have information and take action," DosSantos said. "They have to understand the context of their problem, the context of their organization, and then they can understand the data and analytics that are presenting them with meaningful help in terms of making their decision."

Beyond talent

In addition to talent modernization, Halper identified four other factors common to organizations that claim success with their analytics programs.

A modern data foundation is key, including deployment in the cloud to enable levels of speed and agility that can't be achieved on premises.

They need a data catalog -- an organized, governed repository where employees can find the data they need to inform their decisions.

Automation is another crucial part of analytics success. When data workers are forced to manually repeat the same tasks, it significantly slows down the time it takes to reach insight.

In addition, without automated tools that can augment humans and search through millions of data points to surface insights, users fail to notice many potentially important data points, leading to missed opportunities.

The importance of communication cannot be understated, according to Halper. Organizations need to clearly communicate what they want out of their analytics to create the buy-in that leads to success. They need to communicate what they want out of their employees to get them to train and be part of their organization's transformation.

"If you're going to all of a sudden implement a new architecture, people are concerned about their jobs," Halper said. "They get scared. If you're going to modernize anything, you need to communicate what you're doing so people buy into it."