“Data is only as good as its ability to drive value,” is the core belief of today’s guest, Dan Merzlyak, Head of Business Intelligence at BlackRock. BlackRock is the world’s largest asset management firm with over $10 trillion in assets and Dan is focusing on building a new data conversion and business intelligence strategy for the company’s core alternatives platform offering.
The three different types of analytics that can be used to drive a business are descriptive, predictive, and prescriptive. As Dan explains in more detail in today’s episode, his approach to business intelligence is to first identify the business problems that need solving and then work backward towards the data. Artificial intelligence becomes useful at the prescriptive analytics stage and we’re only just scratching the surface of the potential of this tool to drive value.
In big companies, there is more often than not a gap between the people who are driving the analytics and the people who are creating the analytics. However, Dan believes that in the near future it will be essential to have a greater degree of collaboration throughout the company, and for business leaders to adopt business intelligence tools in their daily workflows as opposed to relying on operational teams to present them with data. To enable true transformation in a business setting, people and processes deserve equal attention.
Dan’s wide range of experience working in different companies across many industries has allowed him to witness the trends that are taking place in the business world. Based on these trends, Dan explains the importance of focusing on seamlessness workflow to attract customers.
The world we live in is constantly changing, and changing fast. Data analytics has the potential to drive enormous value for an organization and keep it relevant in an ever-evolving environment. If you’re interested in hearing about the transformational power of data, you’ve come to the right place!
Listen to this episode of More Intelligent Tomorrow to learn:
“Data is only as good as its ability to drive value,” is the core belief of today’s guest, Dan Merzlyak, Head of Business Intelligence at BlackRock. BlackRock is the world’s largest asset management firm with over $10 trillion in assets and Dan is focusing on building a new data conversion and business intelligence strategy for the company’s core alternatives platform offering.
The three different types of analytics that can be used to drive a business are descriptive, predictive, and prescriptive. As Dan explains in more detail in today’s episode, his approach to business intelligence is to first identify the business problems that need solving and then work backward towards the data. Artificial intelligence becomes useful at the prescriptive analytics stage and we’re only just scratching the surface of the potential of this tool to drive value.
In big companies, there is more often than not a gap between the people who are driving the analytics and the people who are creating the analytics. However, Dan believes that in the near future it will be essential to have a greater degree of collaboration throughout the company, and for business leaders to adopt business intelligence tools in their daily workflows as opposed to relying on operational teams to present them with data. To enable true transformation in a business setting, people and processes deserve equal attention.
Dan’s wide range of experience working in different companies across many industries has allowed him to witness the trends that are taking place in the business world. Based on these trends, Dan explains the importance of focusing on seamlessness workflow to attract customers.
The world we live in is constantly changing, and changing fast. Data analytics has the potential to drive enormous value for an organization and keep it relevant in an ever-evolving environment. If you’re interested in hearing about the transformational power of data, you’ve come to the right place!
Listen to this episode of More Intelligent Tomorrow to learn:
[INTRODUCTION]
[00:00:00] DM: The way I think about BI and AI is that they’re enablement tools to solving really tough business challenges. From my perspective, you should always start with the business problem you're trying to solve and the stakeholders you’re trying to influence and get the problem structured. Then work backwards to the data and the techniques you will use on the data to drive value from it.
[00:00:30] ANNOUNCER: Welcome to More Intelligent Tomorrow, a podcast about our emerging AI-driven world. Critical conversations about tomorrow's technology today. On today's episode, host Ari Kaplan sits down with Dan Merzlyak, Head of Business Intelligence at BlackRock.
[INTERVIEW]
[00:00:54] AK: Dan, welcome to the podcast.
[00:00:57] DM: Thanks, Ari. Happy to be here.
[00:00:59] AK: So excited for you to be here as well. I heard you just recently joined BlackRock just a couple of months ago as their BI lead and head of alternatives data migration practice. That's super impressive. For the listeners, they are the largest asset management firm in the world. 10 trillion or so in assets. I was looking at LinkedIn, over a million followers. How did you get there? What do you do there?
[00:01:24] DM: Thanks, Ari. Sure. I'll start from the beginning. My name is Dan Merzlyak. I'm based out of New York City, talking to you from Brooklyn today. I began my career in analytics management consulting with the big four, where I worked in a lot of different industries and spent time creating analytics strategies and tools. I then went on to work for a private equity company based out of New York City called Cerberus Capital, and their commercial operations group. There, I worked within portfolio companies serving in interim roles, in operations, in analytics functions.
Then finally, right before I joined BlackRock about a month ago, I spent time at the London Stock Exchange Group as the global head of performance reporting in analytics, where I ran a global team of BI directors and data scientists responsible for reports, dashboards and analytics for our division globally.
Recently, at my role at BlackRock, I'm focused on building out a new data conversion and business intelligence strategy for our core alternatives platform offering, as you mentioned. As you could probably tell from my background, I think I bring a unique perspective to this strategy build out, because I spend a lot of time being in the weeds, creating the data flows and the processes. I know how the sausage is made, so to speak. I also focus on driving value with the business end goal in mind. I'm now in a position where I pivoted from the operational space to driving data products for our customers in a more external facing role. Looking forward to our conversation so we can dig into some of this.
[00:03:15] AK: In terms of process, couldn’t be in some ways a big difference at the London Stock Exchange, 300-year-old organization. BlackRock, what is it, 30, 40 years-old? What was it like in terms of process? When I think 300-year-old company, I just think set in their ways in process. I imagine, everything has to adapt and become innovative. Why don’t we start with what's it like managing processes, or working on processes at a very long-standing company?
[00:03:48] DM: It's a great question. The London Stock Exchange group is a very old company, but the beautiful part about that firm is that it's very recently gone through a ton of different acquisitions. As we went through the integration, there were a lot of opportunities for a more entrepreneurial environment, where we got to actually build new things from the ground up and integrate a lot of our core systems based on how we want to shape the strategy of the company moving forward with a lot of new people in place as well to drive that strategy.
I actually came to the London Stock Exchange Group through an acquisition, and I actually started at a company called Refinitiv, which was a Blackstone carve out of Thomson Reuters, and we stayed at Blackstone-owned company for about a two years, until ELSAG acquired us. When they had acquired us, we were actually bigger than the acquiring entity. We had, I believe over 10,000 people at the time and ELSAG had considerably less. We were combining two different core sets of products.
Our space was delivering financial data and analytics to institutions, so they can make actionable decisions off of that data. We really led the horse to water as it relates to how we would carve out the next iteration of building out our financial platform offering for this now combined set of customers, that there were significant synergies with ELSAG in terms of driving that revenue growth for.
For me, although as you can imagine, like any large financial institution, there were definitely significant hurdles that you had to jump through in order to get things to the finish line, I actually think that because of the different M&A activity that was happening at the time, a lot of the stakeholders that I worked with were very open to new ideas and subsequently carved the path for us to get those ideas across the finish line.
There was a lot of impact that you could drive, even in an older organization. Even at BlackRock, similar, because there are some acquisitions that we've made as a company. But as you alluded to, it's a little bit of a newer company. The leadership team here, we're all very focused on driving new ideas and looking at challenges with a fresh perspective.
[00:06:40] AK: It was incredibly impressive how many mergers, acquisitions and then also, large investments that they made, so they put their investment money to work. Hundreds of organizations, like Merrill Lynch Investment, Barclays, Global Investor, State Street Research, a lot of firms that are extremely well-respected around the world, but then investments in pretty much every single industry that you can imagine.
Much younger company than the London Stock Exchange Group, but the speed of growth was dizzying. It was doubling in asset money under management in about eight years. Quite impressive.
[00:07:24] DM: Yeah. BlackRock really, as you mentioned, grew out of Blackstone in the late 80s. It started as an asset manager and it also, over the years, grew to being the largest asset manager with now, I believe last I checked, over 10 trillion dollars under management. It's definitely a behemoth in that space. There's two, I would say, aspects of the company. One is the asset manager. That's really the company as it invests institutional money to drive profits for its investors.
The other great thing that BlackRock recognized early on in its journey is that data and technology enables the profits that BlackRock can bring to its customers. Then it realized that it could sell those services to other institutions that want to leverage a similar technological ecosystem to drive profits for their investors. Really, the whole company is very interested in offering a broad range of its products as a suite. That's what really makes it so interesting and be power player in this environment.
[00:10:04] AK: Yeah. Very interesting. Reminds me in one aspect of the Bloomberg terminal way back in the day, where they had some unique insights. They had data. Yeah. Well, one of the things I was interested in, in your title, there's business intelligence and there's also now, artificial intelligence. What do you see BI as the role in the industry and what do you think of artificial intelligence playing a role?
[00:11:15] DM: Yeah, that's a great question. I think before I answer that, I want to take a step back, because these two terms, BI and AI, they're used so often as buzzwords. I think, a lot of the time in the industry, people start with the data first. I don't think that's always the best approach. I think that’s largely shaped by my prior experience as being focused on driving business value first.
The way I think about BI and AI is that they’re enablement tools to solving really tough business challenges. From my perspective, you should always start with the business problem you're trying to solve and the stakeholders you’re trying to influence and get the problem structured and then work backwards to the data and the techniques you will use on the data to drive value from it to make sure that it's created the right way and the analytics you're creating are crafted in a way that they're able to service the use case at hand. For me, data is only as good as its ability to drive value. Usually, that means operationalizing it into workflows and then through things like BI, business intelligence, enabling self-service and really making sure business users understand both how and why they're doing something.
Business intelligence recently has taken off, and not so recent now, but there's a lot of modern platforms in the space, like Power BI in the Microsoft Suite, or Tableau that really help bring data discovery and self-service to the business user. That's what I think business intelligence will help provide moving forward is really bringing data tools and the ability to use data to drive operational decisions to the business user, without having to do a lot of the technical backend themselves.
That business user can be an analyst, helping leadership understand the insights to drive the next best action, or can be the leader themselves. That's why I think that in the future, I see leadership adopting business intelligence tools in their own daily workflows. Rather than receiving insights and PowerPoints from their operational teams, being able to draw those insights from their own self-discovery in some of these different BI tools that I mentioned.
I think, ultimately, being data driven and the ability to do that analysis is going to be crucial for a leader moving forward 10 years from now, if it not already is. I think that a leader will have to be able to drive some of those decisions themselves off of these tools. Artificial intelligence, I think you mentioned that as well. I think of artificial intelligence and machine learning, I think of that space as another technique to drive insights.
From a business perspective, there's really three different types of analytics that you can use to drive your business. There's descriptive, predictive and prescriptive. I think that AI will really help enable leaders to drive the next best action, or the prescriptive layer. Right now, I think most organizations that are not as mature in their data ecosystem are still very much focused on descriptive and predictive analytics when it comes to using data to drive operational decisions. I think that AI will help organizations ultimately get to the next level in how they're thinking about using data to drive their businesses forward.
[00:15:47] AK: Very well said. I love the concept of, whether you call it citizen data scientist, or citizen business analyst, many years ago you needed a special group of people to pretty much do everything from descriptive to predictive insights. Now that's a rising trend. If you have 30 times the number of people in an organization that are business people that understand the data, or understand the business, or can ask the question compared to the technical people, you've just unleashed a whole other level of the value of your data assets.
You're saying in the next 10 years, you're seeing that as a trend where more and more people will be able to ask more and more predictive type of questions? Where do you see that as of now? Do you think it's at a place now where people who just know the data and just know that the business are able to ask more and more complex questions?
[00:16:46] DM: Yeah. I think, based on the companies that I've been in, I cannot speak for all organizations out there, but I have seen that the ability to drive analytics is often segregated from those that are creating the analytics, very often. It's rare that you have a dual mandate in your role in order to be doing both, when it comes to a big organization. I think that startups are a little bit more nimble and they hire data-driven talent to be able to not only analyze the data such that is meaningful, but also draw conclusions off of that data.
Ultimately, in my opinion, if you don't have the right people behind your technology, you'll be dead in the water. The biggest challenge is finding and hiring talent and then enabling the talent the right way. For employees that are already in the building, it's important to invest in upskilling and talent development to make sure that people have the right skills to deliver on the data initiatives in flight. You could think of these as formal training programs, or just on the job development, actually practicing some of this.
You also want to make sure that people are aligned to your vision as a leader. Often, if you're a leader practicing what you preach in terms of driving strategy off of the data and using these tools, it's much more likely then that the people in your team will try to do the same. I feel that it's not as effective talking at someone as it is to actually doing it. It's important that people believe in what you're doing, not because they must, but because it's the right thing to drive the business forward. I think, ultimately, by going through iterations of using data to drive strategy and being successful, it'll show people that really, data informed decisions are the best way to craft a strategy and also implement it.
Clarifying the company and team objectives through transparent plans and making progress towards some of these milestones is really, from my perspective, instrumental to execution of it. The other thing that I'll say is the people behind the wheel need to really be curious and data-driven individuals that operate with the business value in mind, first and foremost. As I mentioned in the beginning of our call, they need to understand how to effectively communicate their tools to a business audience and take a business idea end to end. It's not just about creating the tool and dropping it on someone's desk, or creating a PowerPoint and dropping it on someone's metaphorical desk. It's really driving the how you would use that data to create the next prescriptive action for the business, and taking the idea’s inception from the data requirements to the front-end buildout and finally rollout maintenance and enhancement.
[00:20:18] AK: How can you incorporate explainability and trust in the results, whether it's BI or AI?
[00:20:25] DM: In my experience, I think that model explainability and transparency is extremely important when socializing the output of your analysis to a business stakeholder. What I mean by that is not only strategically explaining how your model was built, but also looking at what factors contributed to the scoring of your variable when talking about your results. I think we're just scratching the surface as it relates to using AI and explainable machine learning techniques to drive value in an operational space.
[00:21:09] AK: Yeah, wonderful. I like that story, since salespeople could be, but not necessarily pure data scientists. They’re probably not hacking away in Python or R, but you and your group were able to give them what features, or variables are most important for some action. It was great to hear that there was uncovered things that were not obvious to business users. To me, that's one of the big advantages of AI, or even BI, is you can get a lot of variables and it could uncover some subtle patterns, or trends or how information interrelates. Then it's still up to the salespeople, or the non-technical people to be part of the conversation and say, “That makes sense. I didn't think of that.” Or, “You know what, data science team or BI team, you didn't factor in X, Y, Z condition,” and then you collaborate up. Just hearing that you were able to uncover non-obvious points that seemed useful, that's a great story to tell.
[00:22:10] DM: Yeah, absolutely. I think, ultimately, you really need to be able to, again, focus on how you're going to drive the business forward. All of these tools that we talk about, Data Robot, or in general, BI platforms, like the ones I've mentioned, they're all enablement tools to really get the business to use data in daily workflows.
The other thing that I think is super important is the concept of data literacy in a business setting. Not only do you need the people to be able to create the tools and drive the insights, but you also need folks to be able to use the tools. I think data culture is another core business responsibility that needs to be driven in an organization in close partnership with technology.
For me, as someone on the business side, when I build new tools and try to guide people to a single source of truth for their use case, I always try to show why the analytics I am trying to evangelize are better than the current state. What value add am I bringing to what's already in the ecosystem? Often, it's a combination of a few things. Standardization across enterprise, data integrity, different views and abilities to slice and dice the data. Then ultimately, prescriptive insights into how to use the data to drive the next best action.
A couple of the core concepts that we just mentioned, explainability and transparency to drive people towards these new changes is really important to making people – making sure that they're not changing because they just have to, but because they want to. Technology for me is really an enabler. There are many tools that do the same thing with different strengths and weaknesses. We mentioned a few tools. Again, Tableau, Power BI, Data Robot, that are all relevant within their own specific use cases.
It's always important to choose the right technology for the right use case. Of course, newer tools, like cloud tools, like Snowflake over, say, Microsoft SQL Server for databases can shorten query times but rarely are the existing tools at any mature company nowadays a deal breaker to accomplishing initiatives. From my perspective, I try to go into every situation taking into account of what are the resources in the ecosystem? What technology is already available to me? Then map out a plan using that while in parallel thinking about how to optimize with new technologies, or techniques in a future state.
[00:25:20] AK: I was thinking of the single source of truth and standardization, and then I was thinking BlackRock was acquiring or investing in a ton of companies that you probably are acquiring new data sources. You get a new company, something new changes in the world. We're going through a period of stock market changes and inflation. That's under the umbrella of data drift, where the data that you've been doing for your interpretations and recommendations has suddenly shifted. What are some of your thoughts on the challenges, or how you handle a world where data is changing quickly and you want to recalibrate your insights?
[00:26:05] DM: Yeah. That's definitely a big challenge. The machine learning process and analytics process in general, the concept of recalibrating your models to changing external variables to general trends shifting over time. Obviously, one of the things that a modern organization needs to do with any models that it has implemented is monitor those models, and analyze that drift that you mentioned over time to recalibrate the models when the drift has gone too far.
I think that looking at new variables in the ecosystem is super important in making sure that your model is accurate and also making sure that the model is providing the right data outputs to be taken into the next step in the operational process. I think that by definition, no model is perfect. Models will be used for the best approximation, your target variable, but you will have to certainly make sure that the way you use the model and the way you use the model over time is purpose-built for the use case that you're trying to solve.
[00:27:36] AK: One of the things we were talking about when we last talked was different metrics that you were using, things like growth and culture and efficiency metrics. Well, I love to hear more about those.
[00:27:48] DM: Yeah, sure. I think there's a couple ways that I'd like to take this question. First, I think everyone always wants to measure ROI on their investments and initiatives. That can help prioritize how we tackle strategic and technical priorities. For value delivery, when an organization thinks about what are the initiatives that we should tackle in our priorities that will drive the needle forward. There's a number of both operational and financial metrics that we track on an ongoing basis and executive reports and scorecards.
Previously, I'll talk about another employer as an example, we tracked four key groups of KPIs related to people, or customer success, pipeline management and forecasting. Collectively, we measured how the KPIs are tracking on a month over month, year over year and versus benchmarks basis that the business identifies as success. I think that tracking KPIs by themselves in terms of actuals is not meaningful, unless you understand if the KPIs that you're tracking are doing well or poorly. It's really important to track them against benchmarks that the business can use to identify those two scenarios.
These are influenced by many different groups in the business. Attribution for a distinct initiative is often a difficult challenge. When you're tracking these KPIs in aggregate, it's usually a combination of a variety of different factors that are influencing the performance of these KPIs. I don't know that any company has figured out the challenge of isolating maybe unique operational initiatives to quantify the impact acts of a specific one of them very well. That's one of the challenges that I've personally encountered and tried many ways to solve it, whether it comes to pilots and quantifying the success of pilots, and then using that to extrapolate that information for a broader perspective.
When we're launching a new initiative, like scoring and identifying a high-quality pipeline for sellers, we try to pilot the workstream and measure success compared to a control group. The associated lift of the pilot group against the expected performance based on the control groups gives us total lift for criteria like pipeline hygiene, top line revenue, and enabling better performance management conversations. Data initiative impacts in general, I think can be examined from a more traditional perspective, like new business models and long-term investment for items. Like, go to market strategy and territory optimization.
Then there's cost reduction and improved customer experience, new sources of revenue, or just better decision-making in general. This can be used to identify which new whitespace initiatives an organization wants to pursue. I think that KPIs are very meaningful, but it's really important to combine them with the business context of that specific organization and really use both to drive how you would use the data to grow your business.
[00:31:43] AK: Dan, the podcast is called More Intelligent Tomorrow. What do you think either the world will be like, or technology will be like, 5, 10, 15 years into the future?
[00:31:56] DM: It's a great question. If I knew the answer to that, I would be a very rich man 10 years from now. I wish I predicted Bitcoin 10 years ago. That being said, I do think there's a couple key trends that have emerged. One is from a business perspective, automation, and this concept of folks wanting seamless customer experience that is as easy as possible.
I think people tend to get really distracted very easily. I think partially because we're on our phones every day and you see a pop-up and you want to click on that notification. I think that's contributing to that a little bit. I would say that in order to make sure you get customers’ attention, you really need to have a workflow in your product that is seamless and easy and minimizes the amount of clicks that they need to do in order to get to the next step. Whether your outcome is to get them to order something, or get them to watch something, or take a survey. I think that's really important.
Also, automation in terms of having technology. Do that, rather than someone behind a desk. I think that right now, especially the younger generations, they're very comfortable with self-serving. That doesn't mean just self-serving insights from a dashboard, but that also means doing everything themselves from managing their 401k investments, or using a robo advisor to ordering exactly what they want as toppings for their salad on an application. I really think that it's important that the technology keeps pace to have the ability for companies to take those needs of a customer and implement them through a seamless customer experience. I think that's really a place where companies need to make sure that they're digitally first and they're able to actually action on these demands.
[00:34:35] AK: Fascinating. One other topic I wanted to cover was data for transformations. We're talking about transforming businesses. How can data be helpful there?
[00:34:45] DM: Yeah. That's a great question. I think the word transformational is a word that means a lot of different things, depending on who you ask and the context. To me, being transformational in my environment right now means to implement data-driven decision-making through the right measures, which is as much a cultural change as it is the creation of the right reports and dashboards.
For me, when I build new tools and try to guide people to them, that single source of truth, I want to be able to show why these tools I'm trying to evangelize are better than the current state rate. I think that using data to guide meaningful conversations is extremely important. For example, in an old role, we expose client retention data in our executive reports and dashboards by our win-loss reason, which enabled our sales, account managers and operational teams to hone in on why someone is canceling our services, and then creating more customized strategies when talking to customers that tend to cancel this way with preemptive solutions to their needs.
Whereas before, the approaches that we customized were not feasible because the visibility of that granular data and summarized insights highlighting the top loss reasons were not available. In order to get the transformation accomplished, you need people and process in place. Both of these are really important drivers of getting transformation moving in the right direction. I try to make sure they're aligned in the right way when approaching a new business challenge.
[00:36:40] AK: Just globally speaking, what are some of the challenges that you see in the world, or see in business today?
[00:36:50] DM: I think that as it relates to the tools that we built, it's really important to make sure that you have a leadership team that is practicing, not talking, and having folks that are doing what they preach in terms of approaching a business problem in a way that uses all the resources you have at hand to guide how you would approach solving it. That's one.
I think in general, having organizations that are focused on their talent is extremely important to developing not just current workers, but the next generation. I think, especially as we get into a lot of remote work and using tools like Zoom or Teams to interact with each other, there are certain things that get lost in translation, or are not as meaningful as real life, in-person interactions that I grew up in my career having.
I think that companies need to find ways to better engage with their people, which will really help with retention and motivation and ultimately, performance. I think, as we get more into normal functioning in the post-COVID era, I don't know if we'll ever have a true post-COVID era, but at least after the first couple of years, I think employers and employees will need to come up with some middle ground that values employee flexibility to be able to function in their own time. Yet, the things that they need to do in their daily lives. Also, provide those meaningful life working sessions that will actually foster collaboration.
[00:38:51] AK: Excellent. Dan, thank you for being on More Intelligent Tomorrow. These have been fascinating topics and wishing you, and BlackRock all the best.
[00:39:02] DM: Thanks, Ari. It was a pleasure being here and it was great talking to you.
[END OF INTERVIEW]
[00:39:09] ANNOUNCER: Thank you for joining us on this More Intelligent Tomorrow journey. Discover more and join the conversation at moreintelligent.ai. The future is closer than we think.
[END]