💡 Download the complete guide to AI-generated synthetic data!
Go to the ebook
Episode 47

47. The secrets of unlocking business value with your data strategy

Hosted by
Alexandra Ebert
In the 47th episode of the Data Democratization Podcast, Alexandra Ebert talks to Maritza Curry, renowned Head of Data at RCS South Africa, on the essentials of crafting an effective data strategy. They deep dive into pivotal elements of data management and governance, shedding light on why a traditional AI strategy might not be imperative. Maritza, with her vast expertise, underscores the significance of incorporating business acumen into AI and data strategy endeavors.
Key Moments:
  • [00:00:04] - Introduction
  • [00:03:25] - AI and data support for business strategies.
  • [00:05:21] - Discussing the misconception of AI as a standalone strategy.
  • [00:08:21] - The role of data strategy in organizations.
  • [00:11:16] - The importance of people and change management in data strategy.
  • [00:16:34] - Lessons learned in formulating data strategies.
  • [00:22:15] - Challenges of demonstrating value in AI and data strategy.
  • [00:25:43] - Explaining the value of metadata management to stakeholders.
  • [00:32:43] - The pivotal role of people in data governance.
  • [00:37:49] - Impact of AI on data governance.
  • [00:41:06] - Exploring future focuses in data and AI.
  • [00:47:56] - Advice for young professionals in AI and data fields.

Transcript

[00:00:04] Alexandra Ebert: Welcome to the Data Democratization Podcast. This is episode number 47, and I'm Alexandra Ebert, your host, and MOSTLY AI's Chief Trust Officer. With me on the show today is a fellow member from the Women in Data Executive Forum, Maritza Curry, who is the head of data at RCS in South Africa. Maritza has shared all of her tips and secrets, how it's getting real business value out of your data strategy. She covered why it is so important to get the basics right, namely data management and data governance, and how data governance needs to look differently to fit your organizational structure depending on which type of business you run.

She also shared why it's so important to communicate not only with executives, but also with the business users, and develop a data strategy that comes "from the people to the people" to unlock business value. Plus, she also shared why not only AI and data literacy initiatives are important, but business literacy is critical and many, many more shares throughout the episode. I would suggest let us dive right in.

Welcome to the Data Democratization Podcast, Maritza. I'm very happy that we have taken the time to record this episode, and before we dive into all the topics that we want to discuss, could you briefly introduce yourself to our listeners and maybe also share what makes you passionate about the work that you do?

[00:01:27] Maritza Curry: Good morning and thank you for inviting me. I'm really excited about this conversation. I've been in data analytics probably around 23 years, so quite a long time. It feels like I've seen it all, done it all, experienced it all. If you look at how dynamic our industry is, that's not quite true. I feel like it's changing constantly. There's constantly something new to learn. That's part of why I love data and analytics, because there's always something new that you can get into, and you can always learn from others in the industry who are doing amazing, interesting things.

If you have as much experience as I have, then there's always opportunities to share your experiences and your learnings. That's why I love this industry. It certainly keeps us on our toes. It certainly makes us think and challenge what we are doing on a daily basis. That's great. It's never static.

[00:02:46] Alexandra: I can only agree. That's definitely one of the benefits of working in the ever-changing industry of data and now also artificial intelligence. Today I want to talk with you mainly about how executives can get their AI and data strategy right. We have this thought-provoking discussion as part of the Women in Data Executive Forum we were both part of. I think you shared some really interesting things that I believe the audience is interested in hearing. Before we dive into the what and how of an AI and data strategy, maybe let's start out with should there even be an AI and data strategy or is it more that AI and data should just support the overall business objectives and strategy of an organization?

[00:03:25] Maritza: Oh, you're starting off with a really interesting question. If you look at AI as a trend in the industry, in 2017, AI was the number one trend in the industry. If you look at where we are now, AI is not even featuring under the top 10, but AI as a strategy is not featuring. What is featuring is AI-driven data management, augmented analytics, augmented workforce.

AI as a strategy has evolved into AI as an enabler of strategy. For me, that's a sign that the industry is maturing around the topic of AI. I think that's how you need to think about it, not AI as your strategy, because I think you very quickly run into challenges with that because it's not really use-case driven. It's not really defined around what value can we get from AI.

A lot of organizations have run into this challenge, and I think this is one of the reasons why only about 12% of organizations can say, okay, hand on heart, we are getting real measurable value, ROI, from our AI strategies. Think of AI as an enabler of your data strategy or your organizational strategy, not a strategy in itself. I think that's when you're really going to get to the value because you're thinking around use case and what AI can do for your organization.

[00:05:21] Alexandra: It makes sense.

[00:05:21] Maritza: Not AI as a strategy, AI is an enabler.

[00:05:26] Alexandra: Understood. Basically, if you would focus too much on AI as a strategy, then you would land in the realm of having AI as a toy exercising the innovation department but potentially failing to connect it to the business units and then not producing real business outcomes. Is there anything else that-

[00:05:41] Maritza: 100%.

[00:05:42] Alexandra: -you see organizations-- Understood. If you now have to think of AI as an enabler of our business strategies, executives just open newspapers or scroll through the social media feeds, the takeaway is AI can do everything everywhere all at once. What are the areas that you would say that are most prone or that you should actually start thinking of AI using to enable within your strategy?

[00:06:11] Maritza: I think it's going to differ by industry and it's definitely going to differ by organization. For me, the starting point with data strategies always going to be what are the problems we're trying to solve and what are the opportunities that we are trying to unlock? If AI is an enabler for some of those problems that you want to solve or the opportunities that you want to unlock, amazing.

For example, if you are quite a mature organization, you've got a very mature data platform, but your next step now is to start thinking about automation, to get greater efficiencies to deliver faster, to shorten that time to insight, then maybe what you need to think about is how can AI help us to do data management better? How can AI help us to do proactive data quality management? How can AI help us to automate some of the very manual processes in our end-to-end data value chain, including analytics, including data science?

If you are looking at opportunities around customer experience and how to make customer experience or improve customer experience, then AI can certainly help you there. For example, in banking, facial recognition is one way that you can get people to get through your branch faster and more efficiently so they don't have to go to a counter in your branch. They can just-- almost like self-service, self-help in a branch. It really depends on your organizational strategy, your data strategy, where you can use AI as an enabler.

[00:08:21] Alexandra: Understood. You mentioned now data strategy twice. Is it your opinion that you shouldn't have an AI strategy, but that a data strategy is important? If so, why?

[00:08:33] Maritza: I think that data strategy is important and I've heard people say, "Well, you shouldn't have a data strategy. Data should just be part of your organizational strategy." I think that's quite a dangerous approach. The reason is that if you're not focusing on data as a separate strategy to your organizational strategy, and yes, I know everyone's going to jump up and down now and say, "No, but Maritza, your data strategy should support and be aligned to your organizational strategy." Of course, that's not what I mean.

I mean that you still need a data strategy that from an analytics point of view, maybe, that definitely you want to support your organizational strategy. There are components in a data strategy that you can't ignore and that you can't just make hygiene factors. That is the data management and the data governance components of your data strategy. Those are the basics that you absolutely always, always have to have in your data strategy. If you don't, your analytic strategy, and if you have an AI strategy, will simply not be successful. It's always going to come down to the data and the quality of your data and how you master your data, how you manage it, how you govern it.

Then, of course, there's a people component to your data strategy, which I think a lot of organizations forget about, because data strategies, inherently, they drive change and they drive transformation. If they don't, it's probably not a very good strategy. If your strategy is not driving change in your organization, you really should think about it and rethink it. It should drive transformation, and that means that you have to manage the change.

Don't just expect that the capabilities that you enable through your data strategy, that people will rejoice and just use your new data products or use the insights that they get from analytics and data science or embed those in analytics in your business processes. It doesn't just happen on its own. You've got to manage it and you have to make it happen. There's a huge change management component to your data strategy as well. That is why it's so important that you have a data strategy that you've thought through and have all of those different components in it. Otherwise, you're definitely going to run into some issues.

[00:11:16] Alexandra: That makes sense and this is also something that I've heard from a few other guests that you definitely need to take people and change management into account. Since you've been in this industry for a while, have you any lessons learned in what order organizations should tackle these three important topics or what should you have in place before you go in the data management, data governance part on the people side to make sure that your change process is a smoother one?

[00:11:44] Maritza: These two are very painful lessons that I've learned in my career when it comes to data strategy. The first one is, it's not something that you can formulate on your own. When I started off my data strategist career, my very first data management role that I landed in, I googled everything that I could about data strategy. I read all the books, there was actually just one, [laughs] two, there were two books on the market at that time, and-

[00:12:16] Alexandra: Oh, nice. [laughs]

[00:12:16] Maritza: -there wasn't a lot in the knowledge base. [laughs] There wasn't much in the knowledge base around data strategy because it was a fairly new thing. I found a lot of white papers, and that's what I based my data strategy on. I didn't talk to a single person in the business, not a single person. The outcome of their data strategy is that it wasn't successful. I mean, I'm being very generous if I said wasn't successful. I don't even think we got to the first year of rollout.

What I've learned is that data strategies are unique to organizations, and data strategy has to come from the organization, it has to come from the people. I had this colleague at one of the companies I work for, and he said it so beautifully, he said to me, "Maritza, data strategy must come from the people or be from the people for the people." I think that is just perfect because every organization's problems that they want to solve are unique.

The opportunities that they want to unlock are unique. You have to talk to people, that is your first step, talk to everyone in the business. Don't just talk to the executives or the operating board, talk to people who actually have to use data on a daily basis. Talk to your data scientists. Talk to the data team who supports the data scientist and the analyst. Talk to just normal people in the business who have to make decisions with data. Get everyone's opinion. Get everyone's pain points, everyone's ideas about what better looks like. Then you formulate your data strategy based on that.

As a data strategist, I see my job as a facilitator. Yes, I am the one who has the knowledge and the experience to bring all of those ideas together and formulate the strategy. At the end of the day, I'm the facilitator of all of those conversations. That also helps with the change management because people can see their ideas, their aspirations in that data strategy, they recognize it and they can connect with it. It mustn't be this cold, static 100-page Word document that you throw out people and say, "Hey, here's your data strategy, go forth and conquer." You just cannot do that. That was the first lesson.

The second lesson was that data strategies are 1% formulation and 99% execution. You cannot have a successful data strategy if you don't have a well-thought-through roadmap on how you're actually going to execute it. Then you have to manage that execution. You have to make sure you've got dedicated people who can execute it, whether that's a ring fence team or an outsource team, but you have to focus on the execution because it will not happen on its own. I think, again, as a data strategist, this was a hard lesson because the formulation part is the sexy part. Let's be honest, the execution can be not the most exciting thing in the world if you are a data strategist.

If you don't focus on it, then your data strategy will not happen. You have to think of all the things that you normally would when you executed a strategy. You have to think about how are we going to measure the success? How are we going to present that to the stakeholders who would be interested in are we actually getting ROI from this investment? Let's be honest, data strategies are massive investments for organizations. You can't just be blase about the execution part. Those were two of the most painful lessons I've had to learn as a data strategist. There are many, but I'm going to say those are the two most important ones.

[00:16:34] Alexandra: Thanks a lot for sharing. I think it's already very important lessons to hear. Many follow-up questions, but maybe to the first lesson that you shared. I was curious, you mentioned this, what sounds like paradise land, just two books about data management versus today, we have to conquer information overflow in any fields of data and AI.

I was curious if you experienced change when you said, okay, you were talking with the different business units, the different individuals within the organization back then when the focus of the conversation was more around data and the problems that they experienced there, or what they would like to see differently versus today where there's much more coverage even on topics like AI, which also leads to sometimes overinflated expectations. Has this conversation become different, more challenging since AI entered the game? What's your experience on that?

[00:17:22] Maritza: I would definitely agree with that, but these conversations have always been challenging. For example, when I had to put my very first data strategy together, and this was probably a good 18 years ago, the absolute rage at the time was BI 2.0. Basically, what that was, was real-time analytics, but, of course, at that time, we didn't have the technology that we have today that enables real-time analytics. This is what the execs asked for.

They wanted real-time analytics, not understanding how it will solve problems in my organization. Just knowing this is a trend and this is something that they felt we needed to do. Anyway, we went ahead, we implemented it painfully, because, again, like I say, there wasn't actual technology that enabled it. You had to code this real-time data ingestion, data provisioning capability, and almost immediately after implementation, we realized no one's able to take a real-time decision based on the information. It didn't fit into-

[00:18:48] Alexandra: Why was that?

[00:18:48] Maritza: -the business process. That problem has repeated itself throughout my career, but with different topics, including AI. There is an understanding of the value that organizations can get from new concepts. I'm going to throw data mesh in here as well. People hear the concept, and they say, oh, that sounds good, and there's one or two companies who have had success with it or have shown some ROI, but at the point where I have these conversations with an exec, the value that we see other organizations get from these capabilities, it hasn't been around long enough to see if there is longevity in that value.

Often the value is very point in time, and then five years later, you look at it and the value isn't there anymore. These are always really, really hard conversations, but how you get around that is, I think your data strategy always has to be bimodal. You have to focus on the basics, the basic problems that you want to solve, and sometimes those problems are really basics. It could be as basic as making sure that data is available for analytics at a specific time in the morning.

Maybe you have a data ingestion problem. That sounds like an operational problem. I don't think that's an operational problem that affects the entire organization that should be part of your data strategy to solve that problem, but at the same time, you also have to look to the future, and if you are very clear and if you can guide your stakeholders around the use cases that they define where AI or any new capability will enable, then absolutely include it, but have your head in the clouds, but your feet anchored in the basics and the foundational stuff that you have to do. Make sure your data strategy has more components.

[00:21:14] Alexandra: Another good point. I definitely want to follow up on how to guide the business units in terms of choice of the use case. I want to come back to something that you said earlier, which was about demonstrating the value and showing the value.

I was just curious, because one thing that I oftentimes hear, specifically when I have conversations focused on AI, innovation, how to bring AI systems into organizations, that here are the challenges that with AI, obviously, you have a different timeline than from traditional software development, and some of those can't predict, "Okay, is it going to take X weeks or X months until you have XYZ result?"

Do you also experience similar challenges when it's more about the foundational parts of the data strategy where you have to invest a lot upfront but not necessarily show or can demonstrate value immediately? Are there any tips you can share with our listeners how to, if there's endurance and trust and believing to executives that it's still worthwhile pursuing a data strategy and subsequently also AI enabled approaches?

[00:22:15] Maritza: I don't think we have the luxury of not showing value very quickly in our data strategy roadmap. I don't think that our stakeholders will tolerate investing in a data strategy and not see the value very quickly upfront. The best approach that I've experienced myself is to be very clear on what your business use cases are, but also what your data management, data governance use cases are.

There has to be some kind of decisioning body that makes decisions around the data strategy. As a data leader, you should not make the decisions. That should sit with your stakeholders. If you have a data board, or a data steering committee, that is where you want those decisions to be made. Your job as a data strategist, or a data leader is to articulate those use cases, the possible value to your data board or data steer co. so that they are enabled, and they have all the information that they can make the decisions.

I think you have to be completely transparent, because it's easy to over-promise when it comes to data strategies. I've done it so many times in my career. I sometimes still do it in spite of the very painful lessons I've learned where we over-promise and we under-deliver. It's just so easy when it comes to technology and people. People and technology, it's just not predictable. I think as a data, you have to be very transparent with your data board, data steer co., and explain why certain use cases you can deliver quickly and what the trade-offs would be if you deliver quicker than what you believe or your team, the data strategy team believes you can deliver it in.

You've got to be transparent about-- there are certain things, especially the foundational things that might take the full three years of your roadmap. There has to be an understanding of why that is so. It's your job to articulate that, to simplify those concepts, because some of those concepts are complex, but it's also not in the frame of reference of the stakeholders that you are engaging with. It's not your job as a data strategist to make those decisions, it is the job of your data board, data steer co, but it is your job to articulate it and to make sure people have the right information to make the right decisions.

[00:25:26] Alexandra: Makes sense. Maybe could you also give us a tangible scenario of how to articulate a specific data management or data governance use case that's a little bit smaller chopped than the three-year fully-fledged plan of, okay, this is where we want to be at after rollout events?

[00:25:43] Maritza: I'm going to choose quite a sticky one. [chuckles] Metadata management. Metadata management is really difficult to explain to a data board because the concept is just so why the heck do we need this? It sounds like we have to invest a lot of money in it. It's just what is the value that we are going to get from it? This is a real experience.

How I explained it was that one of the problems that emerged through my interviews, my pre-data strategy formulation interviews, was there were two problems. The one was that there isn't any ownership in the organization. People don't take ownership of the data that is produced through their business processes. That ended up in a really serious data quality issue. The other problem was that this company had a really big data science team, probably around 30, 40 people at that time. It grew really quickly.

The data scientists and the analysts, there was another team of analysts. All in all, almost a hundred people. They said one of the biggest problems that they had was they had to answer business questions really quickly because it was a very dynamic business, very fast paced. They really struggled to understand where to find data to answer those questions because of the complexity of the technical landscape. Those were two problems we had to solve.

Metadata management and the metadata management tool could solve those problems. Obviously when I had to explain this to the business stakeholders, to the let's call it investors, in the data strategy, I couldn't even utter the word metadata management. I would've lost everyone. The way that I articulated it was around those two problems and the cost because both of those problems we could actually attach a actual cost to in terms of productivity hours lost.

The cost of those productivity hours lost and also the opportunity cost of not being able to answer business questions quickly, not understanding who do I go to, who owns this data, who must I go ask about this data, what's the quality level, et cetera. Just the cost of the productivity hours and the opportunity cost was far higher than the cost of the technology we had to implement to solve the problem. That is how we presented it to the data board.

Actually, at that time, it was an executive, it was the operating board, and that business case was approved. That business case did go through. It's years later now, and I know that that company, I was a consultant, so I did leave the company, but that company did implement metadata management very successfully with and definitely did get the value from it.

[00:29:03] Alexandra: Awesome. Thanks a lot for walking us through this example. Maybe, since you mentioned importance of communicating and not using complex work where you lose your executive group or data committee, do you have any other generalizable tips on how to communicate with executives about data and AI initiatives?

[00:29:22] Maritza: I think it's always going to be the starting point is the use case. Because, if I put myself in the shoes of an executive who has to buy into the data strategy, I want to know what's the value going to be for my division, what's the value going to be for the organization. If I'm going to invest all of this funding that you're asking for, what is the value? It's always going to come down to value.

Like I said earlier, if you want people to connect to your data strategy, they have to see how this will change their lives, how it's going to change their divisions, how it's going to enable their people, how it's going to enable the organization. The conversation's always going to start with value, but the value is specific to that organization, not the value that we read about in the knowledge base, the value that other organizations can get from it. Sometimes that conversation could be, "Right now, this organization is in terms of their maturity. Maybe AI isn't the answer. Maybe AI is not going to solve your problems because you've got other problems to solve."

Pushing an AI agenda because, as a data leader, that's going to be great for my experience and my career growth to implement AI use cases. Sometimes you have to put your ego a little bit almost to the back of what is valuable to the organization and be just really honest with the organization, "But you don't have the maturity. You cannot implement AI use cases without this, this, this, and this being in place. You've got to do that first." That's an incredibly difficult conversation to have. Again, it's going to come down to the value and making sure that you articulate very clearly value trade-offs cost, et cetera.

[00:31:50] Alexandra: That's a good point and also is in line with one of my recent guests here, who is a data and analytics leader in an insurance company. He also shared that oftentimes the data scientists are inclined to work on the newest technologies, the shiny new difficult modeling approaches and so on, but what the business units need is sometimes something that just takes them half an afternoon to code out that it's a real-life changer for the business unit.

I think it's a good point that you reiterate to putting your-

[00:32:18] Maritza: 100%

[00:32:18] Alexandra: -egos behind and having the business needs and what the business units need upfront. Maybe also one other thing that I'm super curious about to discuss with you because you mentioned the important of data governance. First, maybe if you could lay out a little bit for our listeners what the most important elements of data governance are in your opinion, and why organizations should pursue data governance as an important piece within the data strategy.

[00:32:43] Maritza: Well, the most important component of data governance for me is its people. Data governance, I think people see it as a very annoying compliance-driven initiative that's driven by IT and IT security in the data office, and why should we do it? It's just standing in the way of innovation. I hear that a lot and it just makes everything else take longer. Data governance for me is, it really is about people, and it has to start with people, because data governance can absolutely be an enabler in your organization. As a data leader, what you need to do is you need to shift the perceptions about data governance in your organization.

One of the really important lessons I learned in my career was that your data governance strategy or the data governance component of your data strategy has to align to the culture of the organization. For example, if it is a very traditional bank that you are working with, then there is quite a high tolerance for governance, because banks are operating an environment-

[00:34:05] Alexandra: In their D&H.

[00:34:05] Maritza: -that's very compliance driven. Retailers are far less compliance driven than a bank, and then the tolerance for governance is far lower, rigid governance is far lower. Again, you absolutely have to think about data governance and the implementation of data governance as bimodal. Of course, there are going to be components of it that's more rigid, but you have to almost ring fence the rigidity component to very specific areas. Maybe compliance reporting or sharing data with third parties. You have to be very rigid around data quality.

You have to be rigid around the safe transfer of that data. You have to be rigid about does that data contain personal data? Then rigidity is justified, but at the same time, you have to realize that there are parts of your business like maybe your data science team where governance can stand in the way of progress if it's too rigid. You have to adjust your data governance approach when it comes to your data scientists and your analysts. You almost have to have this continuum when it comes to data governance. On the one side there's very rigid data governance, on the other side there's flexibility, and you have to be clear about how do we implement that in your organization?

[00:35:48] Alexandra: That definitely makes sense. One thing I'm also curious about in this context, how does AI change the picture of governance for organizations? Is it just the same thing, a little bit different, or are there any fundamentally new aspects that need to be taken into considerations when organizations want to use AI?

[00:36:07] Maritza: I have to think about this one. I don't think it really changes the fundamentals of data governance because if your data governance strategy or approach includes data ethics, that's going to include AI ethics. If your data strategy includes, for example, a model governance component, that's going to cover advanced analytics in all of its guises whether that includes AI in your organization or just analytics, advanced analytics.

I don't think it changes your data governance approach very much other than what I've just talked about that maybe you have to be a little bit more flexible in terms of how you implement data governance and data management standards. It doesn't change the fundamentals of data governance. I don't think so. Data governance has to be transversal in your organization. Can't just apply to AI, can't just apply to analytics. It's got to apply to everything that you do with data and all of your different types of users or data consumers. It's got to cover all of that. It maybe expands it to the topic of AI, but it doesn't change fundamentally how you do data governance.

[00:37:49] Alexandra: Interesting point. I'm wondering, does it change on the spectrum of scale that when we assume that organizations have become more mature with AI and data, also have these AI enabled tools in the hands of significantly more employees, is there an importance to teach the fundamentals of data and AI ethics to a broader part of the organization's employees?

[00:38:12] Maritza: Yes, that's actually a good point. If you think about just very normal, traditional, self-service BI, you can't implement self-service BI successfully without a self-service governance framework because then it just becomes a free for all and you cannot manage how people use and share data. It has to happen within a self-service governance framework.

Now, if you look at where we are now and what the trends are in the industry, we have self-service 2.0, and that is augmented analytics. We give normal people in our organizations access to advanced analytics tools, and that they don't have to have a background in statistics or mathematics to use these tools. It's basically a black box. They just need to use it and make decisions on the outcomes. That does not negate the fact that you need self-service governance. Even more so, I think, do you need self-service governance in that scenario.

Does it mean that you have to rethink your data literacy program? Probably. You probably want to think about-- Because these are two very different types of users doing very different types of things, using data in very different ways. Yes, you probably want to think about your data literacy program and how it applies to these new types of users. Your citizen analyst, citizen data scientist.

Again, the fundamental stays the same. Maybe you just expand your data literacy and your self-service governance program. You have to rethink it a little bit. You can't just stand back and say, "It will be fine." Nothing's going to happen. You have to keep up with what's happening in your business and how that affects your data governance approach.

[00:40:24] Alexandra: That maybe wouldn't be the wisest choice to just let it happen since we spent a lot of time on the basics of a data strategy, which I think is important, particularly since you emphasized that these are the deciding factors in the successes of so many things that build on top of it.

For organizations that already have reached a maturity level where they say, "We are quite happy with the state of our data governance, our data management," what would you say are other important areas that executives should look into to set them up for success, particularly also when we look a little bit ahead, maybe generative AI technologies and many organizations being curious how they could incorporate that in their strategy? What else is there that executives should focus on?

[00:41:06] Maritza: I think if you are lucky enough to be very mature in all the components of data and analytics, then you can definitely start looking at what are the trends in the industry and how can I expand the capabilities I already have? We talked about self-service BI and how you can evolve that into augmented analytics or analytic self-service data science.

You can look at your data platform and how you can use AI to get greater efficiencies out of your data platform and, as I said earlier, to shorten that time to value, or time to insight, you can expand your data literacy program to business literacy. Not only making sure that-- Data literacy, that's the most basic thing you can do with your business users, is to teach them how to interpret data, how to use data. You want to take it beyond that. You want to make sure that when they look at a dashboard or they use a self-service application, that they can actually make a business decision based on that data that's going to lead to an actual business outcome. I think that's often forgotten.

Everyone's jumping on the data literacy bandwagon and designing and rolling out these programs. I think absolutely that's very important, you have to do that, but you have to think beyond that. Beyond that is business literacy.

[00:42:57] Alexandra: Absolutely.

[00:42:57] Maritza: Just assume that someone in the business has business literacy. Don't assume that. You've got to bring those two things together. You can expand your self-service as well with generative AI. Think about this. If you can incorporate generative AI into your data and analytics platform, your business user can use a very sophisticated bot on your platform to ask quite sophisticated questions.

You could ask a question, maybe you want to say, I want to know what was the turnover for this region this year versus last year. Generative AI can interpret that question and bring back an answer. Generative AI can be used to help our business users interpret dashboards. It can actually write the story around what's in that. You can expand what you've already done. I think this is the key. If you don't have those basic things in place, how can you go to that next level? It's just not possible.

[00:44:18] Alexandra: Absolutely.

[00:44:18] Maritza: Have the basics in place. Do it well because, by the way, for me, innovation is also doing what everyone else is doing, but doing it a heck of a lot better than anyone else. That's also innovation, not just the new sexy stuff.

[00:44:35] Alexandra: Makes sense. Yes, absolutely. I like what you're sharing because oftentimes it's just that organizations that simply don't yet have the maturity level to jump on the AI wagon or generative AI wagon and say, okay, how can we bring this-

[00:44:49] Maritza: Correct.

[00:44:49] Alexandra: -into our organization? These initiatives usually don't lead somewhere meaningful. I also love the point that you shared on ensuring business literacy. I recently had the CEO and the VP of curriculum from Data Camp, one organization that's really focusing on democratizing AI and data skills on the podcast. We were also talking about this challenge that when organizations invest in their AI and data literacy programs, that there oftentimes is disconnect to the business problems oftentimes fueled by privacy because they can't work on the real organization's data but have some toy data sets that don't necessarily relate to the day-to-day context of the individuals.

This is also where we, with synthetic data, build a bridge in there so that organizations can democratize access to meaningful data. I think this is also one important element to make sure that data literacy also happens on the organization's data, to have a better chance that this actually then also results in using the data and AI skills to drive business decisions to do things differently. Really enjoyed that you brought up this point.

[00:45:56] Maritza: 100%.

[00:45:58] Alexandra: Since we're approaching the end of our episode, maybe two last questions. One thing I would be super interested to hear from you, which areas are you most excited to focus on in the next, let's say, one to three years ahead?

[00:46:13] Maritza: For me, it's definitely democratization in all of its forms. Still focusing on traditional self-service BI and making that successful because, let's be very honest, there's a reason why self-service BI has been a trend for 15 years, because we're just not getting it right. I would love to get it right and get real value from it.

The other component of that is augmented analytics, is to really take self-service to the next level. That could be augmented analytics, maybe, as I mentioned earlier, incorporating a generative AI into some of our data products. I think making sure that you are decentralizing the capabilities that often sits in a central data team to your business users, that's what gets me excited and that is what I would like to focus on.

[00:47:26] Alexandra: Maybe as a last question, since we're both members of the Women in Data Executive Forum and you have this extensive experience in the field, what would be your advice to young professionals in an AI and data field who have either already started their career here or are interested in joining the field? What would be your advice for them to make sure that they're successful in their career and enjoying working in that space as much as you obviously do?

[00:47:56] Maritza: I would say be open to other people's ideas and collaborate. Data analytics is a team sport. When we started off, we talked about how dynamic this industry is and how quickly it changes, but at the same time also remains the same. I would say to any young person entering the field is to collaborate with other people. People with a lot of experience who you can learn from, but also people not just in IT or in technology, but people in the business who can teach you about the business. I think where we're going as data professionals is that you have to have equal knowledge in technology and data analytics capability, data management capability as your business knowledge.

That's where you're really going to provide value to your business stakeholders, is when you can bring those two things together. You're going to understand the opportunity, you're going to understand the problem, but you're also going to understand how can I solve that problem? The reason I'm saying you need to collaborate is, you can't be a one trick pony. There isn't a single solution to every single problem or realizing every single opportunity. If you're a data scientist, and you said it earlier, it doesn't mean that data science solves every problem.

You have to be open to the possibility that there are other solutions, easier solutions, a straight line between two points. That would be my advice. Collaborate with other people, be open to other people's ideas, learn from other people, and put your ego on the back burner, because if you don't, you're not going to learn from other people.

[00:49:59] Alexandra: Thank you so much again. Very, very wise words. I just admire how you talk about all these important concepts in just such a concise and easily to understand manner. Really, really appreciate you taking the time and coming to the podcast today. Thank you so much, Maritza.

[00:50:14] Maritza: Thank you, Alexandra. Thoroughly enjoyed the conversation. Thank you.

[00:50:20] Alexandra: I hope you enjoyed this episode as much as I did. I really appreciate how on point Maritza is with her recommendations and points. If you have any questions, feedback, comments, concerns, as always, you can reach out to us at podcast at mostly.ai or simply commenting below the posts on LinkedIn. Until then, I'm looking forward to have you tune in soon.

Ready to try synthetic data generation?

The best way to learn about synthetic data is to experiment with synthetic data generation. Try it for free or get in touch with our sales team for a demo.
magnifiercross