PODCAST | Bayer Consumer Health, Chief Analytics and Insights Officer, North America: Change Management in Data Governance is the Hardest Problem

Baz Khuti: Hello. Good morning. My name is Baz Khuti. I'm president of Modak, and we're partnering with CDO magazine, MIT CDO IQ, and the International Society of Data Officers in a series of interviews. And today I welcome Manik Gupta, Chief Analytics and Insights Officer at Bayer Consumer Health. Good morning.

Manik Gupta: Baz, good morning. How are you? Hope all is well with you. Good to see you again. 

Khuti: Thanks for volunteering to do this. COVID-19 has had a massive impact on all of our lives. What kind of changes have you seen in types of solutions and how data is being used within Bayer to help manage the pandemic, and what have you learned from this?

Gupta: As dramatic as COVID-19 has been, what it's done for data and analytics, in essence, is that it has accelerated data and analytics about five years. What we thought would take five years to do —  that has happened in under a year. That's because of two things — the relentless focus on accelerating the digitization of data and the digital transformation of functions within a business. And then, the virtualization of people and processes and, frankly, the transformation of food businesses. More importantly, what has happened is this whole idea of virtualization has taken itself to the very next level. 

We have obviously virtualized our data. So, our data now sits in modern stacks in the cloud; our algorithms are now working in the cloud. We’re pushing all of this content to stick with consumers in ways that they want to receive content.

We're meeting our consumers where they want us to meet them in the physical world and the online world, and at the intersection of both worlds, which is the omni world. So, in essence, this is virtualization. This is about leveraging the cloud and then pushing content that makes the most impact. 

Khuti: That's good! Great to hear! And you know, COVID really has become an opportunity from the challenges we all face — you now use data to accelerate some of the solutions back into the market, and personalization to your consumers and customers, as you go forward. 

But this technology is continuously changing, it’s evolving all the time. There's new and many fronts around smart sensors, predictive health, precision health, and so on. How does Bayer keep up with these technology trends and their impact on business?

Gupta: There's a tremendous push to study these trends across every facet of the Bayer corporation. I’d love to talk about what we are doing here. But before I talk about that, I want to revisit this idea of bionic capabilities because, at the end of the day, that is about combining the bible of supercomputing with the ingenuity of the human mind.

Just sit back for a second and think about computing and Moore's Law, and then take one step forward. Think about quantum computing. What's really happening is that one technology year is now equal to 200 chronological years. That is just simply profound in terms of the pace of change that we all have to keep up with.

So, we take a very value-chain view of analytics and insights, and at the end of the day, these technology trends are lending themselves to new ways to produce value. We're looking at data signals from consumer-led culture trends. We're looking at digital, including social, mining omnichannel transactions from smart devices that are in the cloud and powered by 5G networks — again, all in the service of providing the right product to the right audience at the right time, which is the very purpose of our business, which is to transform everyday health.

Khut: When you say “value chain,” what are the core pillars of that? 

Gupta: Upstream in the value chain for analytics and insights is the creation, the collection, and the storage of data. So, think of everything from big data lakes to big query platforms that allow you to actually store and organize your data. Somewhere in the middle of the chain are all these goofy workflows that one needs to create and then let algorithms work at scale on the data that you've been collecting, storing, and organizing. Next, downstream, it's really all about producing products, services, solutions, and the answers to business problems with the singular intent of monetizing the data, creating value for the company and its stakeholders, and obviously making sure that we have a very important role to play in taking care of the health of our consumers.

Khuti: That's a good way of putting it — it comes down to the upstream, midstream, downstream approach of looking at the value chain and then, as you said earlier, the feedback loop between these. So, there's continuous learning as we go forward. That’s a great way of putting it. But in all of it, you need good data. How do you get good data? I talk to so many of my peers about the importance of good data governance. What do you think are some of the key traits, habits, behaviors, and processes that should be considered in good governance?

Gupta: Hopefully, this is not up for much debate, but data is an asset in today's environment. Data is monetizable. Data can drive sustainable competitive advantage. It's never been more critical to build a comprehensive data governance program, and that has four facets to it.

First, at a very fundamental level, it is about data quality. It's not the number of data sets you have, it's the largest data sets you have. It's the 1% of data that is actually monetizable. And within that, it's how clean, accurate, and complete that data is because garbage in, garbage out. Number two is about data access and security. There has to be an intense focus on protecting this asset. That's a very critical, non-negotiable equation within a good data governance program. Then comes this idea of data stewardship, which I think is relatively misunderstood. Folks think that data stewardship is all about IT managing your assets. That's simply not true. Every function that owns the data must steward that data, and that's all about maintaining it, refreshing it, and running restatements on it.

So, stewardship is very important. It's this recognition that you have a role to play as you manage a data life cycle. And finally, it is not data oversight. We partner with the North America IT Advisory Board, and in essence, the responsibility of the board is to ensure we have a great understanding of our use cases and the value that these use cases should produce.

We then tailor IT investments and data analytics investments in line with those use cases. When you do that, you obviously pay a lot of attention to the underlying data assets you've got in relation to those use cases and the monetization of those use cases.

That's really how I think about highly cohesive, well-structured data governance.

Khuti: That's a great way of putting the full data structure down for good data governance and housekeeping. What kind of behaviors or cultural values do you think are important to ensuring those four pillars are institutionalized in the company?

Gupta: When it does come to stewardship, whether you go to product supply or begin planning, or in operations, manufacturing, marketing, or sales, you do have ownership of data. You're bringing data into the company either because you're generating it yourself through your operation or you're buying. And that, in essence, means that you have stewardship, responsibility, and accountability for those assets. The realization that you do have a responsibility like that — and it doesn't necessarily always sit with IT or the chief data officer, or the chief analytics officer — is a very critical realization.

And I see a lot of progress in that space. A lot of people now understand why managing data, even within your function or within your space within a business, is so critical to the overall functioning of that business. I think that realization is very critical to the success of a good governance program.

Our data management routines, protocols, and systems technology are eminently solvable problems. I do tend to over-rotate towards the change management aspect of data governance because that is the hardest problem to solve and takes the most time and effort.