Guiding Your Team Through a Data-Driven Transformation

The Challenges of a Data Environment

Governing a transactional application—like amazon.com, for instance—is very different from governing a data environment. Most people in the industry mainly have experience with transactional applications.

Let’s take a look at data scientists. They generate insights based on information that is extracted from applications. They need data, but it’s restricted because of the personal information it contains. It needs to be cleaned first; and there are numerous processes involved. So, how do you get people to change their ways and give you different governance models in production?

Getting People to Accommodate Transformation

I specialize in data-driven transformation and have overseen four transformations throughout my career: at a startup, at Capital One, for the US census, and at my current company. Traditionally, you have to get executive support for cloud and data transformations. So in the process of drawing out a roadmap, I recommend communicating with stakeholders. A great deal of managing transformations is listening to people, understanding their concerns, and noting why they’re hesitant to change.

While controlling data is necessary, people also need access to it to get their work done. The biggest challenge is acquiring the data in the first place. Copying it may seem like a solution, but having a person do that can lead to inaccuracies. There needs to be an external process that copies the data, which safeguards its legitimacy; only then can people generate insights from it. At Capital One we set up Databricks, a Big Data SaaS provider, in order to accomplish this goal.

When making substantial changes in the way you govern or in your work processes, you can’t go too fast. You have to meet people where they are. Even if the technology is impeccable, if it’s too big of a change or if you can’t communicate it well, people aren’t going to embrace it.

How Amazon solved this problem: they had a tool that let workers use the programming language they wanted.

How we solved this problem at Capital One: similarly, we let people use the tools they wanted to generate insights. We didn’t mandate that everyone uses whichever tool is cloud native. We got people used to working in the new environment without people having to to learn a whole new skill set to do their job. Then, and only then, did we begin transitioning over to newer toolsets.

Helping Employees Keep Up with Change

• The less change you put people through, the better.
• Make incremental changes towards your ultimate desired position.
• Let people ask whatever questions they have, and let them “grow up with the technology,” in Gene Kranz’ words.