Culture Eats Strategy for Breakfast—and in Data & AI, Poor Culture Eats Careers Too

UWEBC Podcast Introduction

Customer experience, technology, people, and supply chains are shaping how organizations operate today, but keeping up isn’t easy. At the UWEBC, we connect leaders navigating transformational change every day. Each week, we bring you the journeys and perspectives of Wisconsin leaders shaping the future of business.

From the University of Wisconsin–Madison, this is the UWEBC podcast.

Doug Barton:
Hey, good afternoon, folks. It’s Doug Barton with the UWEBC, and this is another episode of Catalysts in Cars Getting Coffee.

We’ve gotten coffee. Hey, look!

I’m here with Karan Dhawal, and we’re going to have a rousing conversation about culture, artificial intelligence, data, and analytics.

I’ll allow Karan to introduce himself in a moment, but I want to say that he has been a tremendous resource to our community of data leaders here in Wisconsin and beyond. It’s exceptionally fun to have you with us, Karan.

Karan Dhawal:
Thank you, Doug. First of all, thank you for the friendship and mentorship and for having us here.

UWEBC is a great community builder for Wisconsin companies and helps keep talent here, so it’s a pleasure to do this.

I’m Karan Dhawal. I work for a strategy consulting company. I also teach chief data and AI courses at a couple of universities.

It’s a pleasure to be here. I’ve been in the data space for a long time, including enterprise data management, AI data management, and governance principles.

We’ll talk a lot about this in this short five-minute video.

Doug Barton:
Thanks.

I noticed that, unless I’m wrong, you didn’t mention culture there. Let’s talk about that.

Why is culture the frontier now? Why now?

Karan Dhawal:
Everyone has been spending a lot of money on new titles like chief data and AI officer, and there’s a lot of AI growth happening.

Everyone is talking about AI and its multiple versions, but no one has talked about culture, why AI fails, or why chief data and AI officers only stay in their jobs for less than two years.

It’s all attributed to culture.

In fact, I’m working with a couple of colleagues to research this topic further.

Culture is important. It’s almost like company culture, but now think about data and AI culture. It involves knowing what decisions you’re making, what data is being used in the AI, and talking about it.

It’s becoming paramount because it is needed for the successful implementation of business policies and for creating business impact.

Doug Barton:
I’m curious, then. As you look at a particular organization, what might constitute a maturity assessment? How do you know whether you’re making progress?

Karan Dhawal:
That’s a good question about maturity.

It’s hard to tell because there is no formal framework for data culture and AI maturity.

I’ve been working on a culture framework and presented it last year to help determine whether a company’s culture is mature.

I once worked with someone who was a leader at a company that sold software. They had an e-commerce website selling software to companies.

I talked to one of the engineers and asked, “What is the top-selling software?”

This data engineer was working on the e-commerce application, loading the software and prices, but was not able to answer the question, “What is the best-selling software?”

That tells you something about the data culture.

The person was just doing their job. They weren’t linking their job to business impact or business goals.

That shows a culture that isn’t mature.

On the other hand, think of a company where everyone, from the top down, knows what their job is, what data they’re using, and why they’re using it.

They don’t say, “This is Karan’s Power BI report,” or “This is Karan’s report.”

They say, “This is our impact report. This is our KPI report.”

A lot of people can hold onto those reports and data and not provide access.

You can easily judge a culture by asking two or three key business questions and seeing what is happening in the company.

Doug Barton:
This raises a question.

It’s common to hear that we’re trying to build data fluency, data literacy, or analytical literacy. But to some degree, it’s all business literacy and fluency.

Maybe relate some of the concepts people talk about to this idea of culture and maturity.

Karan Dhawal:
A lot of people will say, “My data isn’t good,” or, “The business doesn’t trust my data.”

You’re always going to hear about quality and things like that.

Culture comes from knowing that we are doing the right things.

I always use the term “key business questions.”

For your business strategy, what are the key business questions you want to ask other team members?

As part of those key business questions, you identify the key business data elements or data domains you need to do your job.

Then talk about them. Who has the data? Can they share it? Can you create that data and put it in an internal company marketplace where other people can use it?

That helps define how you build the culture and measure it going forward.

Here are a couple of key examples.

I’ve had clients who always talk about wanting a certain technology for data. It could be a cloud database, a new security database, or a process.

But they never talk about the key impact they’re trying to make in the business.

Why do they need the software?

If a senior leader made the decision that this is the software they want, they should define and document that decision.

Five years later, if the software doesn’t have enough feature upgrades, the market has changed, and they need something different, they may be spending millions of dollars to replace that software.

At least it won’t become a blame game where people say the prior management did it.

It will be documented to show why that decision was made at the time.

Having the key business questions answered, along with the decision metrics, helps create a data culture.

Then it isn’t about the prior management, the prior data scientist, or the software being at fault. It was a decision that was made.

Doug Barton:
That’s a great example.

One of the questions I have related to culture, and I think this is true of organizational culture, is who owns it?

Is there such a thing as an owner?

Karan Dhawal:
There is.

It’s almost like asking who owns the company culture.

Data and AI culture is a subset of company culture.

It’s also like asking who owns security in a company. How can you keep your company secure?

It’s everyone’s job not to click on a phishing email.

Data and AI culture is very similar. I think everyone owns it.

What you can do to demonstrate that you own it is know what data you’re using, confirm that it is governed data and good-quality data, and talk about the data you have used with others.

Be a champion for the data you are using.

You may or may not be the owner of it, but champion it and say, “This is good data.”

Also, link the data to the key decisions you’re making from it.

At the leadership level, someone might say, “Create this new AI algorithm.”

But if the data being used is poor quality or has bias, and the leader is enforcing a culture of using the data anyway, that reflects an immature culture.

From an ownership perspective, if a data scientist uses something that has bias in it, they aren’t doing the right thing for the culture.

I think the culture is owned by everyone who touches the data, uses it, or is empowered as the owner of a dataset.

They all own the culture.

Doug Barton:
That’s an interesting observation.

In the ultimate case, we all become owners of the data culture.

There is a sense in which it is diffused but felt. It has a felt presence in governing our actions and ways of working.

Let’s wrap this up with your most urgent takeaway.

If someone is listening today, what might be their first step toward improving the data culture?

Karan Dhawal:
The key takeaway is that it starts and stops with you.

When you’re making a key decision and using data, talk about it and document it.

If you’ve made a decision that affects the business, link the key business questions to the key business decisions.

Also understand how your job and the KPIs you use are affecting the overall business strategy.

Having that understanding is the key takeaway.

As I said, it starts with you, and it ends with you.

You are the one who will help improve the overall data and AI culture maturity in your organization.

Doug Barton:
That’s great.

We all have a role to play, and it can start close to home.

It isn’t a big, abstract idea that requires a committee. It can begin right at home.

Karan Dhawal:
It begins right at home.

Doug Barton:
Okay, folks. We’re going to leave it there.

Karan, thanks for being with us, and On, Wisconsin.

Karan Dhawal:
On, Wisconsin. Thank you so much, Doug.

UWEBC Podcast Closing

If you want to keep the conversation going, check out the UWEBC website for more events, resources, and details on what’s coming next. Thanks for listening, and from the UWEBC, we’re glad you joined us.