Data is central to how companies compete, nurture customer relationships, and develop brand loyalty through end-to-end customer experience.

In this environment, data strategy is crucial to business success. But, who should be responsible for the data strategy? Who owns the customer and operational data? What are the appropriate metrics and KPIs for a customer-centric data strategy? And most importantly, how does the data strategy support the underlying business goals?

To address these questions and more, we speak with Danielle Crop, Chief Data Officer of Albertsons, and Bruno Aziza, Head of Data and Analytics at Google Cloud. This episode explores how Albertsons, with over $62 billion in revenue and 325,000 employees, uses data across the company to improve operations and deliver better and more personalized products and services to customers.

The conversation includes these topics:

Danielle Crop is the Senior Vice President and Chief Data Officer at Albertsons and is responsible for building and executing a world-class central data strategy that delivers benefits for the customer regardless of whether they shop in store or on the company’s digital platforms. Her work uses machine learning and advanced data science capabilities to enhance performance across Albertson’s businesses and markets.

Bruno Aziza is Head of Data and Analytics at Google Cloud. He specializes in scaling businesses & turning them into global leaders. He has helped companies of all sizes: startups, mid-size, and large public companies. He helped launch Alpine Data Labs (bought by Tibco), AppStream (bought by Symantec), SiSense (bought Periscope Data) and  AtScale. He was at Business Objects when they went IPO (after acquiring Acta and Crystal Reports, and before SAP bought them for $7B). He was at Microsoft when they turned the Data & Analytics business into a $1B giant.

Transcript

Michael Krigsman: The combination of data and customer experience has become essential. And so, today on CXOTalk, we're speaking with two profound experts: Danielle Crop, Chief Data Officer of Albertsons, together with Bruno Aziza, Head of Data and Analytics for Google Cloud.

Danielle Crop: My role and responsibility as chief data officer at Albertsons is data science, data platforms (so data management and data governance), as well as data products (so all of those wonderful things that enable the enterprise to deliver on customer experience with data).

Michael Krigsman: Bruno, tell us about Google Cloud and tell us about your role.

Bruno Aziza: I run advanced product management for Google Cloud data analytics portfolio. That's products you might have heard like BigQuery and Dataproc and Dataflow. We have lots of news, lots of new customers we can talk about today as it relates to the customer experience challenges and opportunities.

Michael Krigsman: Danielle, maybe you can start us off by giving us some insight into this relationship between data and customer experience.

Danielle Crop: Today, it's essential. Customers expect that you're going to know things about them that (30, 40 years ago) they would have never expected that you would know about them.

They expect you to have an understanding of who they are, what relationship they have with you as a company, and what purchases they've made in the past. This is an expectation that they have and that you're going to customize your experience with them based on the information that you know about them. That's a fundamental difference.

Data and CX are inextricably entwined at this moment in time. I consider those probably going to get even more entwined in the future.

On data collection for customer experience

Michael Krigsman: When we talk about customer experience, what kinds of data do you think about gathering, Danielle at Albertsons? Then, Google, you're talking across a range of different organizations. What are the kind of patterns of data collection that you see?

Bruno Aziza: To add to what Danielle said is this notion of real-time, which is becoming really important for, I think, organizations but also customers who are going through their shopping experience. Just like Danielle said, they expect that not only are you going to understand them, but you're also going to understand cohort analysis, and there are some people just like them, so you can create really compelling experiences, help them find products that they need faster, and maybe consider products they didn't think about before.

I think, if you break down the types of data, there are at least three or four data types you have to be able to collect, transform, and augment around the experience.

The first one is, of course, the data about your particular customers and visitors, understanding who they are as individuals, but who they are as maybe groups of individuals, and understanding what's going to augment (if you will) their experience.

The second type of data is everything around engagement. When you have a multichannel experience online, a retail store, and maybe through partner stores, you need to understand what are the click-through rates, what are some of the conversions that you're experiencing, because that's a sign (if you will) that you're presenting your customers with the right information.

The other aspect is everything around behavior, so implicit behavior around purchase history, just like Danielle was saying, or abandonment of shopping carts. Why are people abandoning their cart? What does that tell us about the experience? Is it about the products now being available? Is it about something wrong about the site or maybe they're dropping subscriptions that you were hoping they would keep?

Then finally, is everything that's explicit – if you run surveys or if you ask them, "Hey, how was the experience in the store?" or "How was the experience here?" – you can get a sense of how well did you deliver on this promise of having a customer experience.

Maybe one last comment on real-time is if you think about non-retailers like Uber or any of these organizations that are in the business of informing people of where are my goods, and they are real-time platforms, I think retailers also need to be able to provide this type of capability. We certainly see companies like Albertsons leading the way here because they're able to not only optimize their inventory; they're able to optimize the experience while shopping but also, optimize the experience while delivering the goods (if they are being shipped to the customer).

Michael Krigsman: Danielle, I see you nodding furiously as Bruno was talking.

Danielle Crop: The real-time aspect in the way that the cloud and more modern infrastructure allow us to gather signals from our customers in real-time and then make decisions in real-time about what their needs are is really unprecedented in human history. It's an exciting place for somebody in data, like myself, to be.

How do we take all of this infrastructure and this amazing amount of data that we have and enable incredible experiences for our customers? One of the ones that appeal to me the most, which we're still working on, is if you're shopping in the store. You're in the store, holding onto your phone.

If you're like me, you do that. I do that every time I'm in the store. I've got my list in front of me. I've got my app in front of me.

What if we could push to people what they bought the last time they were in the produce aisle while they're standing in the produce aisle? Wouldn't that be super helpful to our customers? While that takes a lot of infrastructure and capability that we don't necessarily have at this moment in time, it will be here soon, and I'm excited about those types of use cases.

Bruno Aziza: We used to think there is an online shopping experience and there is an in-person shopping experience. But, increasingly, it's a hybrid shopping experience. You had this multichannel experience before, but what if you could bring them together even beyond the experience inside the store?

We have retailers that we work with who are able to provide real-time inventory information with a specific aisle where you would find what you're looking for while the person is traveling to the store to really make it efficient for them to not have to waste a bunch of time to find what they're looking for. What Danielle is describing here – I really do think is the future of retail – is understanding this multichannel, multiformat relationship and bringing it together to the service of excellent customer experience.

In the end, they're going to buy from you because you have created an environment that is really focused on getting them to what they're looking for, and sometimes even recommending things that they might not have thought about, in very effective ways. There's a huge opportunity here beyond just the experience itself. The challenge is, as there is tremendous growth in data, there's tremendous diversity of the nature of data you'd have to bring in order to accomplish that.

We tend to think about searches, clicking on boxes, and so forth. But images, audio files, what's consumed that really has been part of the universe of these consumers that if you truly have a modern data platform, it allows you to bring it all together to really create something that's different that maybe, frankly, we never thought would be possible – just like what Danielle is describing.

On data sources that drive customer insights

Michael Krigsman: What I find to be fascinating about what you're both describing is, very often, we think of customer experience as the screen looks nice, attractive buttons, but you're really going to customer experience at the most foundational level where you're deep into the operations, deep into the processes, and really reflecting (at an intuitive basis) how the customer is behaving and what they actually want from you.

Bruno Aziza: In addition to this is data that they might not have had access to before. If you think about the context of the data that influences your decision as a buyer, of course, you've got information about your preferences, and you've got information about the inventory (and if it's available), but what about weather conditions? What about things that are outside the store and outside of your own data repository (as an individual, if you will) that are going to make your experiences better?

What if we knew that, two days from now, it's really going to rain and you should really consider that, or the conditions are going to change? There's also a lot of opportunity in bringing external data to really create this experience that, up until now, if you think about the world of retail 30 years ago, it simply just was not possible.

Michael Krigsman: Danielle, as you're thinking through data and customer experience at Albertsons, how is this all resonating with you, what Bruno was just describing – again, this strong operational and infrastructure piece?

Danielle Crop: The amount of data and the amounts of types of data that are able to be brought together now, almost in the real world you can replicate the same kind of data streams that you have online. Think about it in terms of if someone wants, and they opt-in, you can know where they were shopping before they came to Albertsons.

Online, you can do that. You've got the signals of, we know what other sites you were at, and we can use that information. It's almost like that's getting replicated in real life, and that can be a really powerful driver of customer experience (as long as it's used for good).

Michael Krigsman: With this kind of broad spectrum of what's possible, how do you isolate, prioritize where to actually focus? You can do anything, so what do you choose to do and how do you decide what to do?

Danielle Crop: We decide what to do a lot based on size of opportunity. The first thing that we try to ask ourselves is, "Is this something that is essential to the business at this point in time, and how valuable is it?" because we always have limited resources in every business to decide what we should do.

I really encourage my data science team to focus on, okay, how much business value is there in this idea? Let's get an idea and size that before we start going down into a possible rabbit hole that might be very interesting but not very valuable. That is really how we do it at Albertsons.

Bruno Aziza: What we've observed with the organizations we work with is it's highly correlated with the level of pain that people are experiencing in their experience. I'll just give you a few examples.

Searching a large catalog, for instance, is really hard for customers (even if they know what they're looking for). Working with Cartier, for instance. It's a company that was started a long time ago; 174-year history of Cartier watches.

What they did is they now enable people to take a photo of the watch and find that, across their catalog, (within three seconds) with a high level of accuracy. That's a huge pain point because the customer already knows what they want, and you want to make it as fast as possible for them to find that product. It's a great example of just augmenting the customer experience in ways that really removes a lot of friction.

Another example is an organization like Loreal, for instance. It's a French organization. Michael, you know I'm French, so here I'm giving you two French examples. I apologize for that.

L’Oreal has created this application. I think it's called ModiFace where you can experience makeup virtually, so augmented reality.

I think, if I look at what retailers and organizations that are providing great service and products like that to their consumers, how they're enhancing the experience is very much related to what Danielle is talking about. They're coming to your online properties, or they're coming to your physical locations to accomplish a job. That job might be finding something they already know or experiencing something that they're very curious but would be really difficult to do.

Changing makeup is hard. It's a very physical experience. If you can use digital experiences to suggest here – based on what you're wearing right now, what you should be putting on as makeup – is really quite amazing and something that we just couldn't do before.

On how to use data science for customer experience and personalization

Michael Krigsman: Would it be fair to say that this focus on data and customer experience ultimately comes down to using data to help model digitally the things that customers care about that they want to do, and that they want to engage in? Is that a fair way to put it, or not really?

Danielle Crop: I would say it's pretty fair. I would add in that the way I look at it is, "Is this helpful? Is it useful to the customer?" because those are the features that they're going to use.

A lot of times, we'll focus on, "Oh, this is really cool." Right? But is it actually going to be useful to the person and their life?

You look at Uber, right? Why was Uber super successful? It was incredibly useful and filled a need that people didn't even know they had.

Those are the types of things I think that are most powerful when you're saying CX meets data.

Bruno Aziza: There's this theory called "jobs to be done," Michael. I'm a big Chris Christensen fan. He's a professor at MIT and has written this book on jobs to be done.

I think people coign to your store or visiting your online storefronts (if you will), they're not coming to just do that. They're coming to accomplish a job that sometimes is about finding a specific item or sometimes even creating an experience for people they're hosting.

To the extent to which you can make that process a lot easier is when you're really going to provide the best service. That's what's going to make them come back because they're going to get a sense that you get them.

You understand who they are. You understand the job they're trying to accomplish, and the way they're hiring you (the retailer) to accomplish that job and enable them to make this progress toward the end goal.

The end goal is not coming to visit your store. The end goal is coming to create an experience for people they love or buying something for themselves so they can feel good, things that are really around the human experience.

Now, it's a beautiful time to be in 2022 because we have tremendous technology that allows you to get there a lot faster. Really, we couldn't do that before.

The cloud has really helped us accomplish a lot of that. Data sharing platforms have allowed us to do this at a very large scale for lots of different data types while the data is governed in a saleable manner.

We're really in a good time now to create amazing experiences for customers shopping, wanting to create experiences for themselves and others.

On ethical considerations of data in customer experience

Michael Krigsman: We have a question that just came up from Twitter from Arsalan Khan. Arsalan is a regular listener who asks great questions. Thank you, Arsalan.

He's asking about the ethical considerations. What about that? How do you think about the ethical lines, as Arsalan mentioned in his tweet?

Danielle Crop: The key thing for me is that it has to be in service of the customer. If the usage is in service of the customer (and you have to request consent), then I think you can say this is for good, this is data for good. If it is not in service of the customer and it is in service of just your shareholders, then you really need to think about whether or not you want to do it.

Fortunately, our use case is not one of those dopamine use cases at Albertsons, so we don't need to worry about that so much. But I think that is a huge consideration for the industry at large, which is, are we deliberately making people addicted to their devices? That is something that, as a culture, I think we have to tackle.

Fortunately, at Albertsons, that's not our use case. We want people to be in and out of their app, and we want it to be easy and useful and sticky, but not addictive.

Michael Krigsman: Bruno, thoughts on this issue of the ethics and where you draw the line?

Bruno Aziza: The security of the data, this is not a compromise. This is something you have to build in, design upfront from that.

An example is opt-in applications. Customers need to opt into that experience, so you can be sure that the data is secure and it's only accessed by them.

It's their data. Ultimately, they own that data. The retailer really doesn't.

You first have to have that contract (that ethical contract, if you will) where it's very clear that you're never going to compromise on that.

Secondly, the way we see retailers work with data, in general, is the aggregation of the data, so you never really look at individual information. You really look at trends, and that's how you maximize the relationship between the data you have and the experiences with your customers.

What Danielle was saying here is that you have to align to what the job is and how you're creating the experience to the service of the people who are buying that from you. You have to focus on where do we align here.

That's why I was connecting earlier in the buyer's journey, if you will. That's where you're going to be able to provide the best value for your organization and the customers is when they're aligning. Any time they're not aligning, I would say, is probably a red flag.

On data science talent and the data team at Albertsons

Michael Krigsman: Let's shift gears a moment and talk about the organizational aspects. Building a data machine, so to speak, is complex. Danielle, can you tell us about the composition of your team?

Danielle Crop: A team of product managers and data scientists that own the data lakes (so the data platforms, data management, data governance teams), as well as data product leaders, and then the data scientists that drive all of those algorithms. That is the composition of the data office at Albertsons. They're responsible for just really driving this change.

We're 11 months in, at this point, of the data office at Albertsons, and so we're still building, growing, and changing. But really exciting opportunities to build some platforms at scale for Albertsons to drive really omnichannel.

I would say that omnichannel has been a buzzword for over a decade, and everybody has been like, "Oh, this is..." I think we're actually at the point now, between data and infrastructure, that omnichannel is going to be a reality. That's really exciting for us to be at the center of building those products and capabilities and the data science that underlies it to drive true omnichannel experiences.

On building a data culture

Michael Krigsman: Bruno, the notion of a data culture, where does that fit in?

Bruno Aziza: First of all, I'd say that Danielle is an exceptional chief data officer. What's amazing in our industry is that we have now graduated and matured to having a lot of folks like Danielle driving data strategies, and that's where data culture starts. It starts at the top.

I think, 10 years ago, if we look at the latest data, only 12% of companies had hired a chief data officer. Now, in 2022, according to the latest data, I think it's 74% of organizations.

We're making a lot of progress. Culture, as vague as it might sound, also needs to be supported by an organization that has a team, that has a charter, that is recognized and brought to the executive table to drive the data strategy.

It has now been recognized just like your CFO. You have a finance organization. Well, now you have a data organization. I think that's where it's starting.

We have a long way to go because, if you look at surveys in the market and we ask organizations, "What is the number one problem to being successful with data?" 91%, almost 92% of people that have been surveyed will tell you that culture is their greatest challenge.

I think Danielle has a perspective on why that is, so I don't want to steal the thunder from her on that. But I think it starts at the top, and then there are, of course, some principles that you've got to go and apply.

Michael Krigsman: Danielle, tell us about data culture and your thoughts on this?

Danielle Crop: It's really interesting to have been in the world of data since 2001, in the corporate environment, and see the change that has occurred. I think, when I first started, the data culture challenge was more of, how do we get people to use data to make decisions? Now, the problem is more, what data do you use to make decisions, because there's so much of it?

It's almost like an analysis paralysis. You can get into so many different metrics and looking at metrics. I think this is where big data and data science is going to have to come in, in this "Fourth Industrial Revolution," is in making sense of all of this data and making it actionable, which is why I think data science is at the center of all of this.

You could look at data all day long, metrics all day long. Are you really making the right decisions? This is where models become very important.

Then going back to our data for good conversation, they have to be ethical models. So, it's an interesting moment in time for data, but I think that that's where the culture needs to shift.

I think we have a lot of very traditional leadership at this moment in time, and they have a tendency to want to look at their reports. But I think, ten years down the road, what we're going to be doing is they won't have to look at those reports necessarily, except for at the highest P&L level. It will be automated.

The decisions in the business will be far more automated through data science and algorithms. I think that that will free up so much capacity within our organization to focus on higher-order strategy. I think that will be fantastic, but I think we're at that moment of, like, a lot of leaders don't really know what this new Fourth Industrial Revolution means. As data folks, we have to move them forward in that direction.

Bruno Aziza: Or even trusting that we'll get there. There's still a lot of education to occur.

The encouraging piece is the rise of the chief data officer and the rise of really good best practices around: How many data people should you hire? What is the makeup of the data team? What is the role of the engineering and data science?

There are a lot of best practices there that just go way beyond what we used to talk about ten years ago on, "Hey, let's have your data culture principles, print them on the wall, and then hope people will respect them." Of course, you need to have that, but you also need to have an organization, and you need to have practices that allow you to remind people of that.

What we see organizations do is data literacy programs. They do office hours, and they train people on what does it look like to understand and work with this data, this dashboard, this insight.

You'll be surprised that people are really interested in learning that. Not the tooling, not the dashboard in themselves, but they want and they recognize that they're sitting on a gold mine of information. Up until now, until you really bring the chief data officer in the organization, they really have not been able to use them beyond just the standard reporting tactics that Danielle is talking about.

I think all of us are consumers and we're experiencing the tremendous benefits that occur when you have intelligence that surrounds you and helps you throughout your day. But I think if we take an honest look at most organizations, they're not there.

In fact, there's research that shows that only a third of organizations are able to get value out of the data that they have. I'm encouraged by the progress we're making, but there's still a lot of work to do. The number one barrier is the culture and the application, the deployment, if you will, the execution on that culture strategy, which is still in progress right now.

On using data science to deliver business value

Michael Krigsman: We have a really interesting point from Robin on Twitter who says, "It's analysis paralysis, but it's also the data story and value that should be connecting with the data products," so company, visions, and value for the customer. Any thoughts on this? I think it's a really interesting point.

Danielle Crop: Your data program and data strategy has to connect to value, and it has to be quantifiable. That's part of the role that I have, which is to develop this organization, strategy, and then make sure that it's very clear that this is tied to business value. Increase in sales, whatever the core metric might be, and that the program is driven towards that.

I am curious, as to Bruno's comment earlier about people, like how few organizations are getting value out of their data. That doesn't surprise me, and yet it concerns me because, to me, data is so valuable that it's hard for me to even imagine a scenario in which you don't get value out of your data.

Bruno Aziza: Everyone should learn from the work that you're doing. It's just still really complex, I think, for most organizations to get full visibility on their data because the data that they work with is often highly distributed, like we said. It's in all types of shapes and formats, and it's really challenging if you are evolving on a platform that might not be modern.

It's really challenging, first of all, to even just get basic visibility. Then building data products on top of that is certainly really complex for organizations to get there.

Beyond the technical aspects, what we are seeing is the organizations that succeed really connect to this point that the listener is making on the value of the storytelling. The business of making decisions is a very emotional one. We're not logical machines that also have emotions. We're emotional machines that also use logic (sometimes) to make decisions.

What we see is people using internal marketing vehicles to make sure that people are reminded of the culture. Do you have a brand around your initiative? Is your brand memorable? Are people recognizing this brand as they're going through their reports?

A lot of organizations I work with will brand their reports or their dashboard saying, "This is certified data based on the brand that we're using." And so, there's definitely a logical component to it, but a big component is also how you're connecting at the human level, not just with your customers but also with your employees who are trying to make the right decisions with their data and sometimes they just can't connect with that.

Michael Krigsman: I just want to invite everybody to subscribe to our newsletter, so we can keep you up to date on our shows. Just hit the subscribe button at the top of our website. If you're watching on CXOTalk.com, subscribe to our YouTube channel too, and tell a friend.

On aligning data strategy and business strategy

Okay. We have a great question. This is from Suman Kumar Chandra on LinkedIn. He says, "How do you align the data strategy of your organization with the business strategy when the business is changing very rapidly?" It's a really good question.

Danielle Crop: I like to simplify this. How you do it is you say, "Okay, your business strategy is always tied to your P&L (in one way or another)." And so, if you keep your objectives very clear (and three-year in nature), then you can tie your data strategy to your objectives quite clearly.

I don't see a difference between the business strategy and the data strategy. They are one strategy. And so, to me, it's about my role and my team's role is to serve the organization.

What is the business strategy? What do we need to build in order to be able to support that business strategy?

It's about making sure you're connected in and not in a silo (as a data team). That you're connected into what the business is doing and how they're doing it.

Michael Krigsman: How do you connect data and business strategy when things are changing really fast?

Bruno Aziza: The data strategy and business strategy are the same. What we see organizations do to make sure they are connected is shared OKR, so a shared goal, a shared goal between someone in the data team and someone in the business team, and that they can't achieve unless they partner. You just have to be able to advance together.

I will say one more thing is that data strategy has the potential of actually advancing your business strategy. There is another French retailer for you. Carrefour just launched an initiative called Carrefour Links.

It's the result of the maturity of their data abilities that now enable them to create a data platform where they can share their best practices on customer behaviors with the ecosystem, a business that they were not in before. But it's the result of connecting data strategy and business strategy.

Now, because you have become such a mature data organization, you can advance the business itself. Again, not many organizations are there, but that's where the opportunity is.

On the Chief Data Officer role

Michael Krigsman: Arsalan Khan comes back, and he says, "If culture is the biggest challenge for data, then where should the CDO report: CFO, CIO, CEO, board? To whom should the chief data officer report?"

Danielle Crop: Depending on the business strategy at the time and what agenda you're trying to move forward, the CDO should sit probably closer to whatever you're trying to move forward at that point in time. Lots of times, the CDO will report to a CTO or a CIO. I think that if you're at the stage in which you need to move your platforms from very old infrastructure to modern infrastructure, that's probably a good place for the CDO to sit for that period of time.

At Albertsons, I report to the chief customer and digital officer because that's the agenda we're moving forward at this moment in time. But I do think that where the CDO sits is different in every organization (in a lot of ways), and so that's not inappropriate, honestly. It is very appropriate, and I think that where the CDO should sit is dependent upon the business objections of the company at that time.

Bruno Aziza: We're seeing the same thing. Certainly, we asked (through surveys) where do CDOs report to. People had the choice: the chief product officer, the chief technology officer, the CIO, the CFO. It was interesting because we got about the same percentage across these, and I think that's probably because of what Danielle is saying is it really depends on who is leading the charge on the most important business imperative inside your organization.

At Albertsons, your goal is to provide the most compelling digital experiences or hybrid experiences for customers, and it makes a lot of sense that data is going to be the fuel that's going to create those experiences. You find organizations where they run under the CFO because they have the CFO that is very mindful.

We focus, just like what Danielle was saying earlier, on the initiatives that drive the best value for customers. Therefore, the way to assess that value is through the conversion into revenue and the ability to sell more and then sell different things that customers come back to us with and so forth, and really optimize our processes.

Then you have organizations where, interestingly enough, they work under the chief product officer because their business is to build data products. And so, they have data product managers who will take the assets they have and then create experiences (like recommendations and others) that are creating those products that really are creating value for your organization and your customers.

There is not one answer. It really depends on the culture and the makeup of your organization and its business goals.

Danielle Crop: The role of the CDO kind of matures over time because I think the first CDO was actually appointed in, I want to say, 2006. It's still a relatively new area and discipline.

I think that if it's driving really core business objectives, there are places in which the CDO reports to the CEO, and maybe that will become more common over time.

On customer experience metrics and measuring data performance

Michael Krigsman: How do you measure? How does one measure data initiatives, and especially as it relates to customer experience? Robin, on Twitter, comes back. He just wants to be very clear that data should not just sit in a dashboard. It needs to be informative, insightful, personalized, actionable, and ultimately lead to accountability.

Metrics, how do we measure these types of initiatives, and especially with an emphasis on customer experience?

Danielle Crop: Data, if it's in a dashboard, isn't data. That's a metric that you track. Metrics are informed by data but they are not data. Data is what's underneath and behind, and so that's another kind of tweak to what he said – I would provide.

Michael Krigsman: I can sense Robin, in the ether on Twitter, is smiling now, with you saying that.

Danielle Crop: Again, that's part of the data culture concept as well, going back to that. A lot of people think metrics or reporting are data. No.

Michael Krigsman: Bruno, how do we measure? How do we decide? What are we measuring here?

Bruno Aziza: There are many ways to decide it, and I'll probably just take one and the opposite. The first one is, how do you measure value inside your organization?

I think one of the issues in the success of the chief data officer today is the inability to connect the data with its value. I think your listener here is hitting on a very specific one.

There are many ways to look at the ROI, if you will, of data analytics. The first one is, of course, just the simple level of adoption.

Are your employees engaged with the data? Are they making decisions based on data? Are they ignoring your initiatives? That's clearly the first thing you want to look at it because if nobody is looking at what you've built, well, if you don't have data-driven changes inside your organization, that's probably a red flag.

The other piece is customer satisfaction because, ultimately, if you think about what you're trying to do here, you're trying to use data so you're more informed about what your customers want, so you can provide the best experiences for them, so survey your customers.

Just like you see on the back of these trucks that say, "How is my driving?" you probably should have the same level of interest on how are we delivering on the promise. You came to our store. You came to our site. Do we really get you? Do we know you?

Then probably the other area that I would look at is how are your products evolving as a result of the information that you have. Is the inventory you have today the same inventory that you had last year? If the answer is it's a 99% overlap, then you have to ask yourself, did you really not learn anything or you were just right on when you started?

I think there are so many factors that are going to affect your inventory. Hopefully, with the knowledge you have from your customers, your industry has to evolve in places maybe you didn't expect.

Look at these three things. It's probably a good way to get started: adoption, customer satisfaction, and the nature of the inventory you propose, the type of company that you've become because of your knowledge of data.

On using data to deepen customer relationships and customer loyalty

Michael Krigsman: What advice do you have for folks who want to use data to deepen their customer relationships? I'll ask you to answer pretty quickly.

Bruno Aziza: Connect with the community. There are many ways to learn from leaders. I'm excited that they're watching you today but start a conversation with people you admire in the industry.

It really starts with the dos and don'ts. You want to learn from best practices. You want to learn from the worst practices. People are really good at sharing that.

We have our own video program we call Data Journeys (every Tuesday) where I interview customers. It's really designed to do that.

I would say the best practice is connect to the humans, the human beings that are behind these best practices, and reach out to them. Use LinkedIn. Use platforms like that to ask your questions. It's going to be the best way that you learn.

Michael Krigsman: Danielle, you started us off and you're going to get the final word here. What advice do you have for folks who want to use data to deepen the customer relationships?

Danielle Crop: Know thy data and know where it is. You can get creative and inspired by data, but you have to know what it is and where it is.

Start there. Start with the simple stuff of, like, "Okay, where does this data reside? How do we pull it? How do we know it? How do we understand it?" Then you can come up with some really great ideas about how you can use it.

You can always come up with data, the things of how you're going to use it, but maybe that falls apart when you actually go get it because it's not there. So, I always say, "Start with what's there, and then grow from there," because if you try to abstract back from an idea, you may be disappointed. If you start with the data, you can come up with some really interesting and creative things.

Michael Krigsman: With that, we are out of time. I want to say a huge thank you to Danielle Crop. She is the chief data officer of Albertsons. And to Bruno Aziza, who is the head of data and analytics for Google Cloud. Thank you both so much for sharing your valuable time with us today. I really, really appreciate it.

Danielle Crop: Thanks.

Bruno Aziza: Thank you so much for having us.

Michael Krigsman: Everybody, thank you for watching, especially the folks who asked such awesome questions today. Before you go, please subscribe to our YouTube channel, hit the subscribe button at the top of our website so we can keep you up to date on our shows. We have just amazing shows coming up. Everybody, thank you so much. I hope you have a great day, and we will see you next time. Bye-bye.

Michael Krigsman: The combination of data and customer experience has become essential. And so, today on CXOTalk, we're speaking with two profound experts: Danielle Crop, Chief Data Officer of Albertsons, together with Bruno Aziza, Head of Data and Analytics for Google Cloud.

Danielle Crop: My role and responsibility as chief data officer at Albertsons is data science, data platforms (so data management and data governance), as well as data products (so all of those wonderful things that enable the enterprise to deliver on customer experience with data).

Michael Krigsman: Bruno, tell us about Google Cloud and tell us about your role.

Bruno Aziza: I run advanced product management for Google Cloud data analytics portfolio. That's products you might have heard like BigQuery and Dataproc and Dataflow. We have lots of news, lots of new customers we can talk about today as it relates to the customer experience challenges and opportunities.

Michael Krigsman: Danielle, maybe you can start us off by giving us some insight into this relationship between data and customer experience.

Danielle Crop: Today, it's essential. Customers expect that you're going to know things about them that (30, 40 years ago) they would have never expected that you would know about them.

They expect you to have an understanding of who they are, what relationship they have with you as a company, and what purchases they've made in the past. This is an expectation that they have and that you're going to customize your experience with them based on the information that you know about them. That's a fundamental difference.

Data and CX are inextricably entwined at this moment in time. I consider those probably going to get even more entwined in the future.

On data collection for customer experience

Michael Krigsman: When we talk about customer experience, what kinds of data do you think about gathering, Danielle at Albertsons? Then, Google, you're talking across a range of different organizations. What are the kind of patterns of data collection that you see?

Bruno Aziza: To add to what Danielle said is this notion of real-time, which is becoming really important for, I think, organizations but also customers who are going through their shopping experience. Just like Danielle said, they expect that not only are you going to understand them, but you're also going to understand cohort analysis, and there are some people just like them, so you can create really compelling experiences, help them find products that they need faster, and maybe consider products they didn't think about before.

I think, if you break down the types of data, there are at least three or four data types you have to be able to collect, transform, and augment around the experience.

The first one is, of course, the data about your particular customers and visitors, understanding who they are as individuals, but who they are as maybe groups of individuals, and understanding what's going to augment (if you will) their experience.

The second type of data is everything around engagement. When you have a multichannel experience online, a retail store, and maybe through partner stores, you need to understand what are the click-through rates, what are some of the conversions that you're experiencing, because that's a sign (if you will) that you're presenting your customers with the right information.

The other aspect is everything around behavior, so implicit behavior around purchase history, just like Danielle was saying, or abandonment of shopping carts. Why are people abandoning their cart? What does that tell us about the experience? Is it about the products now being available? Is it about something wrong about the site or maybe they're dropping subscriptions that you were hoping they would keep?

Then finally, is everything that's explicit – if you run surveys or if you ask them, "Hey, how was the experience in the store?" or "How was the experience here?" – you can get a sense of how well did you deliver on this promise of having a customer experience.

Maybe one last comment on real-time is if you think about non-retailers like Uber or any of these organizations that are in the business of informing people of where are my goods, and they are real-time platforms, I think retailers also need to be able to provide this type of capability. We certainly see companies like Albertsons leading the way here because they're able to not only optimize their inventory; they're able to optimize the experience while shopping but also, optimize the experience while delivering the goods (if they are being shipped to the customer).

Michael Krigsman: Danielle, I see you nodding furiously as Bruno was talking.

Danielle Crop: The real-time aspect in the way that the cloud and more modern infrastructure allow us to gather signals from our customers in real-time and then make decisions in real-time about what their needs are is really unprecedented in human history. It's an exciting place for somebody in data, like myself, to be.

How do we take all of this infrastructure and this amazing amount of data that we have and enable incredible experiences for our customers? One of the ones that appeal to me the most, which we're still working on, is if you're shopping in the store. You're in the store, holding onto your phone.

If you're like me, you do that. I do that every time I'm in the store. I've got my list in front of me. I've got my app in front of me.

What if we could push to people what they bought the last time they were in the produce aisle while they're standing in the produce aisle? Wouldn't that be super helpful to our customers? While that takes a lot of infrastructure and capability that we don't necessarily have at this moment in time, it will be here soon, and I'm excited about those types of use cases.

Bruno Aziza: We used to think there is an online shopping experience and there is an in-person shopping experience. But, increasingly, it's a hybrid shopping experience. You had this multichannel experience before, but what if you could bring them together even beyond the experience inside the store?

We have retailers that we work with who are able to provide real-time inventory information with a specific aisle where you would find what you're looking for while the person is traveling to the store to really make it efficient for them to not have to waste a bunch of time to find what they're looking for. What Danielle is describing here – I really do think is the future of retail – is understanding this multichannel, multiformat relationship and bringing it together to the service of excellent customer experience.

In the end, they're going to buy from you because you have created an environment that is really focused on getting them to what they're looking for, and sometimes even recommending things that they might not have thought about, in very effective ways. There's a huge opportunity here beyond just the experience itself. The challenge is, as there is tremendous growth in data, there's tremendous diversity of the nature of data you'd have to bring in order to accomplish that.

We tend to think about searches, clicking on boxes, and so forth. But images, audio files, what's consumed that really has been part of the universe of these consumers that if you truly have a modern data platform, it allows you to bring it all together to really create something that's different that maybe, frankly, we never thought would be possible – just like what Danielle is describing.

On data sources that drive customer insights

Michael Krigsman: What I find to be fascinating about what you're both describing is, very often, we think of customer experience as the screen looks nice, attractive buttons, but you're really going to customer experience at the most foundational level where you're deep into the operations, deep into the processes, and really reflecting (at an intuitive basis) how the customer is behaving and what they actually want from you.

Bruno Aziza: In addition to this is data that they might not have had access to before. If you think about the context of the data that influences your decision as a buyer, of course, you've got information about your preferences, and you've got information about the inventory (and if it's available), but what about weather conditions? What about things that are outside the store and outside of your own data repository (as an individual, if you will) that are going to make your experiences better?

What if we knew that, two days from now, it's really going to rain and you should really consider that, or the conditions are going to change? There's also a lot of opportunity in bringing external data to really create this experience that, up until now, if you think about the world of retail 30 years ago, it simply just was not possible.

Michael Krigsman: Danielle, as you're thinking through data and customer experience at Albertsons, how is this all resonating with you, what Bruno was just describing – again, this strong operational and infrastructure piece?

Danielle Crop: The amount of data and the amounts of types of data that are able to be brought together now, almost in the real world you can replicate the same kind of data streams that you have online. Think about it in terms of if someone wants, and they opt-in, you can know where they were shopping before they came to Albertsons.

Online, you can do that. You've got the signals of, we know what other sites you were at, and we can use that information. It's almost like that's getting replicated in real life, and that can be a really powerful driver of customer experience (as long as it's used for good).

Michael Krigsman: With this kind of broad spectrum of what's possible, how do you isolate, prioritize where to actually focus? You can do anything, so what do you choose to do and how do you decide what to do?

Danielle Crop: We decide what to do a lot based on size of opportunity. The first thing that we try to ask ourselves is, "Is this something that is essential to the business at this point in time, and how valuable is it?" because we always have limited resources in every business to decide what we should do.

I really encourage my data science team to focus on, okay, how much business value is there in this idea? Let's get an idea and size that before we start going down into a possible rabbit hole that might be very interesting but not very valuable. That is really how we do it at Albertsons.

Bruno Aziza: What we've observed with the organizations we work with is it's highly correlated with the level of pain that people are experiencing in their experience. I'll just give you a few examples.

Searching a large catalog, for instance, is really hard for customers (even if they know what they're looking for). Working with Cartier, for instance. It's a company that was started a long time ago; 174-year history of Cartier watches.

What they did is they now enable people to take a photo of the watch and find that, across their catalog, (within three seconds) with a high level of accuracy. That's a huge pain point because the customer already knows what they want, and you want to make it as fast as possible for them to find that product. It's a great example of just augmenting the customer experience in ways that really removes a lot of friction.

Another example is an organization like Loreal, for instance. It's a French organization. Michael, you know I'm French, so here I'm giving you two French examples. I apologize for that.

L’Oreal has created this application. I think it's called ModiFace where you can experience makeup virtually, so augmented reality.

I think, if I look at what retailers and organizations that are providing great service and products like that to their consumers, how they're enhancing the experience is very much related to what Danielle is talking about. They're coming to your online properties, or they're coming to your physical locations to accomplish a job. That job might be finding something they already know or experiencing something that they're very curious but would be really difficult to do.

Changing makeup is hard. It's a very physical experience. If you can use digital experiences to suggest here – based on what you're wearing right now, what you should be putting on as makeup – is really quite amazing and something that we just couldn't do before.

On how to use data science for customer experience and personalization

Michael Krigsman: Would it be fair to say that this focus on data and customer experience ultimately comes down to using data to help model digitally the things that customers care about that they want to do, and that they want to engage in? Is that a fair way to put it, or not really?

Danielle Crop: I would say it's pretty fair. I would add in that the way I look at it is, "Is this helpful? Is it useful to the customer?" because those are the features that they're going to use.

A lot of times, we'll focus on, "Oh, this is really cool." Right? But is it actually going to be useful to the person and their life?

You look at Uber, right? Why was Uber super successful? It was incredibly useful and filled a need that people didn't even know they had.

Those are the types of things I think that are most powerful when you're saying CX meets data.

Bruno Aziza: There's this theory called "jobs to be done," Michael. I'm a big Chris Christensen fan. He's a professor at MIT and has written this book on jobs to be done.

I think people coign to your store or visiting your online storefronts (if you will), they're not coming to just do that. They're coming to accomplish a job that sometimes is about finding a specific item or sometimes even creating an experience for people they're hosting.

To the extent to which you can make that process a lot easier is when you're really going to provide the best service. That's what's going to make them come back because they're going to get a sense that you get them.

You understand who they are. You understand the job they're trying to accomplish, and the way they're hiring you (the retailer) to accomplish that job and enable them to make this progress toward the end goal.

The end goal is not coming to visit your store. The end goal is coming to create an experience for people they love or buying something for themselves so they can feel good, things that are really around the human experience.

Now, it's a beautiful time to be in 2022 because we have tremendous technology that allows you to get there a lot faster. Really, we couldn't do that before.

The cloud has really helped us accomplish a lot of that. Data sharing platforms have allowed us to do this at a very large scale for lots of different data types while the data is governed in a saleable manner.

We're really in a good time now to create amazing experiences for customers shopping, wanting to create experiences for themselves and others.

On ethical considerations of data in customer experience

Michael Krigsman: We have a question that just came up from Twitter from Arsalan Khan. Arsalan is a regular listener who asks great questions. Thank you, Arsalan.

He's asking about the ethical considerations. What about that? How do you think about the ethical lines, as Arsalan mentioned in his tweet?

Danielle Crop: The key thing for me is that it has to be in service of the customer. If the usage is in service of the customer (and you have to request consent), then I think you can say this is for good, this is data for good. If it is not in service of the customer and it is in service of just your shareholders, then you really need to think about whether or not you want to do it.

Fortunately, our use case is not one of those dopamine use cases at Albertsons, so we don't need to worry about that so much. But I think that is a huge consideration for the industry at large, which is, are we deliberately making people addicted to their devices? That is something that, as a culture, I think we have to tackle.

Fortunately, at Albertsons, that's not our use case. We want people to be in and out of their app, and we want it to be easy and useful and sticky, but not addictive.

Michael Krigsman: Bruno, thoughts on this issue of the ethics and where you draw the line?

Bruno Aziza: The security of the data, this is not a compromise. This is something you have to build in, design upfront from that.

An example is opt-in applications. Customers need to opt into that experience, so you can be sure that the data is secure and it's only accessed by them.

It's their data. Ultimately, they own that data. The retailer really doesn't.

You first have to have that contract (that ethical contract, if you will) where it's very clear that you're never going to compromise on that.

Secondly, the way we see retailers work with data, in general, is the aggregation of the data, so you never really look at individual information. You really look at trends, and that's how you maximize the relationship between the data you have and the experiences with your customers.

What Danielle was saying here is that you have to align to what the job is and how you're creating the experience to the service of the people who are buying that from you. You have to focus on where do we align here.

That's why I was connecting earlier in the buyer's journey, if you will. That's where you're going to be able to provide the best value for your organization and the customers is when they're aligning. Any time they're not aligning, I would say, is probably a red flag.

On data science talent and the data team at Albertsons

Michael Krigsman: Let's shift gears a moment and talk about the organizational aspects. Building a data machine, so to speak, is complex. Danielle, can you tell us about the composition of your team?

Danielle Crop: A team of product managers and data scientists that own the data lakes (so the data platforms, data management, data governance teams), as well as data product leaders, and then the data scientists that drive all of those algorithms. That is the composition of the data office at Albertsons. They're responsible for just really driving this change.

We're 11 months in, at this point, of the data office at Albertsons, and so we're still building, growing, and changing. But really exciting opportunities to build some platforms at scale for Albertsons to drive really omnichannel.

I would say that omnichannel has been a buzzword for over a decade, and everybody has been like, "Oh, this is..." I think we're actually at the point now, between data and infrastructure, that omnichannel is going to be a reality. That's really exciting for us to be at the center of building those products and capabilities and the data science that underlies it to drive true omnichannel experiences.

On building a data culture

Michael Krigsman: Bruno, the notion of a data culture, where does that fit in?

Bruno Aziza: First of all, I'd say that Danielle is an exceptional chief data officer. What's amazing in our industry is that we have now graduated and matured to having a lot of folks like Danielle driving data strategies, and that's where data culture starts. It starts at the top.

I think, 10 years ago, if we look at the latest data, only 12% of companies had hired a chief data officer. Now, in 2022, according to the latest data, I think it's 74% of organizations.

We're making a lot of progress. Culture, as vague as it might sound, also needs to be supported by an organization that has a team, that has a charter, that is recognized and brought to the executive table to drive the data strategy.

It has now been recognized just like your CFO. You have a finance organization. Well, now you have a data organization. I think that's where it's starting.

We have a long way to go because, if you look at surveys in the market and we ask organizations, "What is the number one problem to being successful with data?" 91%, almost 92% of people that have been surveyed will tell you that culture is their greatest challenge.

I think Danielle has a perspective on why that is, so I don't want to steal the thunder from her on that. But I think it starts at the top, and then there are, of course, some principles that you've got to go and apply.

Michael Krigsman: Danielle, tell us about data culture and your thoughts on this?

Danielle Crop: It's really interesting to have been in the world of data since 2001, in the corporate environment, and see the change that has occurred. I think, when I first started, the data culture challenge was more of, how do we get people to use data to make decisions? Now, the problem is more, what data do you use to make decisions, because there's so much of it?

It's almost like an analysis paralysis. You can get into so many different metrics and looking at metrics. I think this is where big data and data science is going to have to come in, in this "Fourth Industrial Revolution," is in making sense of all of this data and making it actionable, which is why I think data science is at the center of all of this.

You could look at data all day long, metrics all day long. Are you really making the right decisions? This is where models become very important.

Then going back to our data for good conversation, they have to be ethical models. So, it's an interesting moment in time for data, but I think that that's where the culture needs to shift.

I think we have a lot of very traditional leadership at this moment in time, and they have a tendency to want to look at their reports. But I think, ten years down the road, what we're going to be doing is they won't have to look at those reports necessarily, except for at the highest P&L level. It will be automated.

The decisions in the business will be far more automated through data science and algorithms. I think that that will free up so much capacity within our organization to focus on higher-order strategy. I think that will be fantastic, but I think we're at that moment of, like, a lot of leaders don't really know what this new Fourth Industrial Revolution means. As data folks, we have to move them forward in that direction.

Bruno Aziza: Or even trusting that we'll get there. There's still a lot of education to occur.

The encouraging piece is the rise of the chief data officer and the rise of really good best practices around: How many data people should you hire? What is the makeup of the data team? What is the role of the engineering and data science?

There are a lot of best practices there that just go way beyond what we used to talk about ten years ago on, "Hey, let's have your data culture principles, print them on the wall, and then hope people will respect them." Of course, you need to have that, but you also need to have an organization, and you need to have practices that allow you to remind people of that.

What we see organizations do is data literacy programs. They do office hours, and they train people on what does it look like to understand and work with this data, this dashboard, this insight.

You'll be surprised that people are really interested in learning that. Not the tooling, not the dashboard in themselves, but they want and they recognize that they're sitting on a gold mine of information. Up until now, until you really bring the chief data officer in the organization, they really have not been able to use them beyond just the standard reporting tactics that Danielle is talking about.

I think all of us are consumers and we're experiencing the tremendous benefits that occur when you have intelligence that surrounds you and helps you throughout your day. But I think if we take an honest look at most organizations, they're not there.

In fact, there's research that shows that only a third of organizations are able to get value out of the data that they have. I'm encouraged by the progress we're making, but there's still a lot of work to do. The number one barrier is the culture and the application, the deployment, if you will, the execution on that culture strategy, which is still in progress right now.

On using data science to deliver business value

Michael Krigsman: We have a really interesting point from Robin on Twitter who says, "It's analysis paralysis, but it's also the data story and value that should be connecting with the data products," so company, visions, and value for the customer. Any thoughts on this? I think it's a really interesting point.

Danielle Crop: Your data program and data strategy has to connect to value, and it has to be quantifiable. That's part of the role that I have, which is to develop this organization, strategy, and then make sure that it's very clear that this is tied to business value. Increase in sales, whatever the core metric might be, and that the program is driven towards that.

I am curious, as to Bruno's comment earlier about people, like how few organizations are getting value out of their data. That doesn't surprise me, and yet it concerns me because, to me, data is so valuable that it's hard for me to even imagine a scenario in which you don't get value out of your data.

Bruno Aziza: Everyone should learn from the work that you're doing. It's just still really complex, I think, for most organizations to get full visibility on their data because the data that they work with is often highly distributed, like we said. It's in all types of shapes and formats, and it's really challenging if you are evolving on a platform that might not be modern.

It's really challenging, first of all, to even just get basic visibility. Then building data products on top of that is certainly really complex for organizations to get there.

Beyond the technical aspects, what we are seeing is the organizations that succeed really connect to this point that the listener is making on the value of the storytelling. The business of making decisions is a very emotional one. We're not logical machines that also have emotions. We're emotional machines that also use logic (sometimes) to make decisions.

What we see is people using internal marketing vehicles to make sure that people are reminded of the culture. Do you have a brand around your initiative? Is your brand memorable? Are people recognizing this brand as they're going through their reports?

A lot of organizations I work with will brand their reports or their dashboard saying, "This is certified data based on the brand that we're using." And so, there's definitely a logical component to it, but a big component is also how you're connecting at the human level, not just with your customers but also with your employees who are trying to make the right decisions with their data and sometimes they just can't connect with that.

Michael Krigsman: I just want to invite everybody to subscribe to our newsletter, so we can keep you up to date on our shows. Just hit the subscribe button at the top of our website. If you're watching on CXOTalk.com, subscribe to our YouTube channel too, and tell a friend.

On aligning data strategy and business strategy

Okay. We have a great question. This is from Suman Kumar Chandra on LinkedIn. He says, "How do you align the data strategy of your organization with the business strategy when the business is changing very rapidly?" It's a really good question.

Danielle Crop: I like to simplify this. How you do it is you say, "Okay, your business strategy is always tied to your P&L (in one way or another)." And so, if you keep your objectives very clear (and three-year in nature), then you can tie your data strategy to your objectives quite clearly.

I don't see a difference between the business strategy and the data strategy. They are one strategy. And so, to me, it's about my role and my team's role is to serve the organization.

What is the business strategy? What do we need to build in order to be able to support that business strategy?

It's about making sure you're connected in and not in a silo (as a data team). That you're connected into what the business is doing and how they're doing it.

Michael Krigsman: How do you connect data and business strategy when things are changing really fast?

Bruno Aziza: The data strategy and business strategy are the same. What we see organizations do to make sure they are connected is shared OKR, so a shared goal, a shared goal between someone in the data team and someone in the business team, and that they can't achieve unless they partner. You just have to be able to advance together.

I will say one more thing is that data strategy has the potential of actually advancing your business strategy. There is another French retailer for you. Carrefour just launched an initiative called Carrefour Links.

It's the result of the maturity of their data abilities that now enable them to create a data platform where they can share their best practices on customer behaviors with the ecosystem, a business that they were not in before. But it's the result of connecting data strategy and business strategy.

Now, because you have become such a mature data organization, you can advance the business itself. Again, not many organizations are there, but that's where the opportunity is.

On the Chief Data Officer role

Michael Krigsman: Arsalan Khan comes back, and he says, "If culture is the biggest challenge for data, then where should the CDO report: CFO, CIO, CEO, board? To whom should the chief data officer report?"

Danielle Crop: Depending on the business strategy at the time and what agenda you're trying to move forward, the CDO should sit probably closer to whatever you're trying to move forward at that point in time. Lots of times, the CDO will report to a CTO or a CIO. I think that if you're at the stage in which you need to move your platforms from very old infrastructure to modern infrastructure, that's probably a good place for the CDO to sit for that period of time.

At Albertsons, I report to the chief customer and digital officer because that's the agenda we're moving forward at this moment in time. But I do think that where the CDO sits is different in every organization (in a lot of ways), and so that's not inappropriate, honestly. It is very appropriate, and I think that where the CDO should sit is dependent upon the business objections of the company at that time.

Bruno Aziza: We're seeing the same thing. Certainly, we asked (through surveys) where do CDOs report to. People had the choice: the chief product officer, the chief technology officer, the CIO, the CFO. It was interesting because we got about the same percentage across these, and I think that's probably because of what Danielle is saying is it really depends on who is leading the charge on the most important business imperative inside your organization.

At Albertsons, your goal is to provide the most compelling digital experiences or hybrid experiences for customers, and it makes a lot of sense that data is going to be the fuel that's going to create those experiences. You find organizations where they run under the CFO because they have the CFO that is very mindful.

We focus, just like what Danielle was saying earlier, on the initiatives that drive the best value for customers. Therefore, the way to assess that value is through the conversion into revenue and the ability to sell more and then sell different things that customers come back to us with and so forth, and really optimize our processes.

Then you have organizations where, interestingly enough, they work under the chief product officer because their business is to build data products. And so, they have data product managers who will take the assets they have and then create experiences (like recommendations and others) that are creating those products that really are creating value for your organization and your customers.

There is not one answer. It really depends on the culture and the makeup of your organization and its business goals.

Danielle Crop: The role of the CDO kind of matures over time because I think the first CDO was actually appointed in, I want to say, 2006. It's still a relatively new area and discipline.

I think that if it's driving really core business objectives, there are places in which the CDO reports to the CEO, and maybe that will become more common over time.

On customer experience metrics and measuring data performance

Michael Krigsman: How do you measure? How does one measure data initiatives, and especially as it relates to customer experience? Robin, on Twitter, comes back. He just wants to be very clear that data should not just sit in a dashboard. It needs to be informative, insightful, personalized, actionable, and ultimately lead to accountability.

Metrics, how do we measure these types of initiatives, and especially with an emphasis on customer experience?

Danielle Crop: Data, if it's in a dashboard, isn't data. That's a metric that you track. Metrics are informed by data but they are not data. Data is what's underneath and behind, and so that's another kind of tweak to what he said – I would provide.

Michael Krigsman: I can sense Robin, in the ether on Twitter, is smiling now, with you saying that.

Danielle Crop: Again, that's part of the data culture concept as well, going back to that. A lot of people think metrics or reporting are data. No.

Michael Krigsman: Bruno, how do we measure? How do we decide? What are we measuring here?

Bruno Aziza: There are many ways to decide it, and I'll probably just take one and the opposite. The first one is, how do you measure value inside your organization?

I think one of the issues in the success of the chief data officer today is the inability to connect the data with its value. I think your listener here is hitting on a very specific one.

There are many ways to look at the ROI, if you will, of data analytics. The first one is, of course, just the simple level of adoption.

Are your employees engaged with the data? Are they making decisions based on data? Are they ignoring your initiatives? That's clearly the first thing you want to look at it because if nobody is looking at what you've built, well, if you don't have data-driven changes inside your organization, that's probably a red flag.

The other piece is customer satisfaction because, ultimately, if you think about what you're trying to do here, you're trying to use data so you're more informed about what your customers want, so you can provide the best experiences for them, so survey your customers.

Just like you see on the back of these trucks that say, "How is my driving?" you probably should have the same level of interest on how are we delivering on the promise. You came to our store. You came to our site. Do we really get you? Do we know you?

Then probably the other area that I would look at is how are your products evolving as a result of the information that you have. Is the inventory you have today the same inventory that you had last year? If the answer is it's a 99% overlap, then you have to ask yourself, did you really not learn anything or you were just right on when you started?

I think there are so many factors that are going to affect your inventory. Hopefully, with the knowledge you have from your customers, your industry has to evolve in places maybe you didn't expect.

Look at these three things. It's probably a good way to get started: adoption, customer satisfaction, and the nature of the inventory you propose, the type of company that you've become because of your knowledge of data.

On using data to deepen customer relationships and customer loyalty

Michael Krigsman: What advice do you have for folks who want to use data to deepen their customer relationships? I'll ask you to answer pretty quickly.

Bruno Aziza: Connect with the community. There are many ways to learn from leaders. I'm excited that they're watching you today but start a conversation with people you admire in the industry.

It really starts with the dos and don'ts. You want to learn from best practices. You want to learn from the worst practices. People are really good at sharing that.

We have our own video program we call Data Journeys (every Tuesday) where I interview customers. It's really designed to do that.

I would say the best practice is connect to the humans, the human beings that are behind these best practices, and reach out to them. Use LinkedIn. Use platforms like that to ask your questions. It's going to be the best way that you learn.

Michael Krigsman: Danielle, you started us off and you're going to get the final word here. What advice do you have for folks who want to use data to deepen the customer relationships?

Danielle Crop: Know thy data and know where it is. You can get creative and inspired by data, but you have to know what it is and where it is.

Start there. Start with the simple stuff of, like, "Okay, where does this data reside? How do we pull it? How do we know it? How do we understand it?" Then you can come up with some really great ideas about how you can use it.

You can always come up with data, the things of how you're going to use it, but maybe that falls apart when you actually go get it because it's not there. So, I always say, "Start with what's there, and then grow from there," because if you try to abstract back from an idea, you may be disappointed. If you start with the data, you can come up with some really interesting and creative things.

Michael Krigsman: With that, we are out of time. I want to say a huge thank you to Danielle Crop. She is the chief data officer of Albertsons. And to Bruno Aziza, who is the head of data and analytics for Google Cloud. Thank you both so much for sharing your valuable time with us today. I really, really appreciate it.

Danielle Crop: Thanks.

Bruno Aziza: Thank you so much for having us.

Michael Krigsman: Everybody, thank you for watching, especially the folks who asked such awesome questions today. Before you go, please subscribe to our YouTube channel, hit the subscribe button at the top of our website so we can keep you up to date on our shows. We have just amazing shows coming up. Everybody, thank you so much. I hope you have a great day, and we will see you next time. Bye-bye.