Randy Bean is the CEO, co-founder and managing partner of NewVantage Partners, a strategy firm that delivers world-class expertise in data strategy for Fortune 1000 clients. We spoke with Randy about the evolution of data-driven roles, data architectural revolution, and how to win the “race to insights.”
1. The fundamental data challenges for organizations have remained the same: learning from the data they have; leveraging data as an asset to gain insight; and putting information into the hands of business users more rapidly to help them stay competitive.
2. Due to regulatory demands, many chief data officers find that 75% of their time is spent on defense (meeting compliance demands), leaving only a quarter of their time to be spent on the offensive opportunities, like creating new products and services and being able to quickly get data into the hands of the end business users.
3. Four years ago, only 12% of firms had a chief data officer in place; now, it’s just over 50%–there’s been a rapid adoption of the chief data officer role.
4. One of the immediate challenges for many organizations is how to define the metrics that will quantify the specific benefits that big data can provide for them.
5. It’s important for chief data officers to frame a data-driven story to create a vision for the longer-term opportunity and potential for the organization.
6. Big data has enabled a cultural shift that encourages us to “fail fast, fail early, fail often.” In other words, iterate so rapidly that you have no choice but to learn from your mistakes; you can find out what’s most relevant, get rid of the stuff that’s not working and begin to derive value from data instantaneously.
7. The more data you have at your disposal, the more informed you’ll be about your business challenges and successes, and the more insights you’ll have to help change your company for the better.
8. It’s not good enough just to accelerate the speed to analytics or insight. You need to get to the point where you can accelerate your decision-making and ultimately be able to quickly adapt and modify your capabilities.
9. We’re not so much bound by the data that we have; we’re bound by the way we use our imaginations. Our capabilities are ultimately determined by whether we’re asking the right questions.
ML: Thanks for being here with us today, Randy. To get things started, you’ve been in the data analytics industry for many years now; you started NewVantage Partners in 2001 and write about digital analytics and big data for The Wall Street Journal. What do you love about your job?
RB: I’ve been in the data space for about 35 years, and the fundamental challenges for organizations have remained the same: learning from the data they have; leveraging data as an asset to gain insight; and putting information into the hands of business users more rapidly to help them stay competitive. But although those challenges remain the same, technology continues to evolve dramatically. And that’s the challenge, and the most exciting part about this line of work.
I’m most excited by how the new capabilities and technologies of big data are really pushing things forward. Along those same lines, there is also the proliferation of data. Organizations were making progress on cracking this problem 30-35 years ago. But what’s happened is that there is so much more volume and variety of data. So, the closer you get to solving the problem, the further you get in other respects because the rate of data proliferation makes it such a challenge to stay on top and get ahead of things.
The biggest challenge in this line of work is to stay fresh, stay relevant and keep in tune with developing capabilities.
ML: How would you define the job of a digital analyst today?
RB: We’re more in line with the new role of chief data officers now. The chief data officer role originated from the regulatory demands that have been placed on organizations and industries like financial services, pharmaceuticals and healthcare. There’s both a defensive and an offensive aspect to the chief data officer role.
Organizations obviously want to be able to leverage the data they have to gain competitive advantage. And many of the individuals that have being hired into the chief data officer role have the skill set to make that happen. However, due to regulatory demands, many chief data officers find that 75% of their time is spent on defense (meeting compliance demands), leaving only a quarter of their time to be spent on the offensive opportunities, like creating new products and services and being able to quickly get data into the hands of the end business users.
So, it’s a new, exciting role, but it’s very much in an evolutionary stage. Over the past several years, my company has been doing a small, but very high-level survey of executives, including chief data officers, CEOs and heads of Fortune 1000 companies. And when we first did this survey four years ago, only 12% of firms had a chief data officer in place. And now, it’s just over 50%. So, there’s been a pretty rapid adoption of the chief data officer role.
ML: What are some of the KPIs that organizations can use to measure the success of a chief data officer?
RB: Well, that’s a very good and timely question. Many organizations say that they’ve established a big data environment, but when asked about specific benefits and quantifying ROI, it’s been much harder for them to clearly articulate those things. It really depends on where the chief data officer role sits in relation to the big data function in an organization.
There are a few organizations that can precisely quantify the benefits, but those are organizations where big data tends to be closely aligned with the CFO. CFOs are measurement-oriented and driven by numbers, so they pay particular attention to the ROI they’re getting from big data, as opposed to many other organizations that section off big data in other parts of the company. One of the immediate challenges for many organizations is how to define the metrics that will quantify the specific benefits that big data can provide for them.
Today, most of the applications of big data have been operation-oriented, and there hasn’t been much focus on using it to create new products, services and capabilities. Hopefully, we’ll see more organizations taking advantage of big data to do that. But so far, it’s been hard to get the technology to stand up within various environments.
ML: You’ve talked about measuring the performance of chief data officers. What would you suggest that chief data officers focus on in the first 90 days? Where can they make the most impact immediately to demonstrate their value to the company?
RB: Well, I’m going to back up a little bit in answering that question. So, chief data officer roles have become more prevalent over the past several years; some of the earlier chief data officers have barely been in their roles for three or four years. And when I’ve asked organizations about whether their chief data officer has been successful, the responses have been mixed. Companies don’t know the answer to that question yet since it’s such a young role. This is a completely new domain, and there are varying expectations among different organizations in terms of what the chief data officer could and should be.
Chief data officers are able to define what the data assets are across an entire organization. They have the opportunity to look at data across the enterprise and to think about how data from different parts of the business can be organized to create maximum value. But most chief data officers have not really been able to pursue the offensive part of the job. Instead, they’ve been used to plug the hole in terms of meeting regulatory needs and ensuring data quality standards. I’ve even heard chief data officers referred to as “chief diplomatic officers” because they often have to make a case for why data is a critical asset for the organization. They have to explain why there has to be cooperation and participation across multiple lines of business to derive the full value of data. So, part of the chief data officer’s job is really creating the story for data and selling the story within the organization.
I think it’s really important for the chief data officer to frame a story that is relevant for that organization. Even if their energies are deflected by regulatory activities, they’re creating a vision for the possibilities that data can create in terms of longer-term opportunity and potential.
ML: Excellent; let me shift focus and talk about a couple of your articles from The Wall Street Journal. One that I think is very timely is “Big Data Fuels a ‘Data First’ Movement.” In that article, you talk about cultural shift. Can you tell us a little bit more about what that cultural shift is and how organizations go about it?
RB: Big data enables organizations to move to a more agile environment where the data tells the story; that’s really the idea behind “data first.” Historically, you had to come up with a hypothesis. Then, you had to go to your IT folks and ask them for the data that you needed (or thought you needed), which they would eventually get to you from a big data warehouse.
Big data environments can develop analytical sandboxes in about six weeks or so, wherein they just load the data into that sandbox and let the end users start to play with it immediately. In other words, you have a subset of data. If you want, you can have all the data, or you can have a piece of the data, or you can have many pieces of data. But you can basically load the data you have into an environment and let the data tell the story in terms of where the correlations are; the patterns; what’s the most relevant data. And as you see what data is most meaningful, you can load more and more data. So, it can be a very iterative, test-and-learn type of environment. We sometimes use the expression of “fail fast, fail early, fail often.” In other words, iterate so rapidly that you have no choice but to learn from your mistakes; you can find out what’s most relevant, and you can get rid of the stuff that’s not working.
ML: You’ve written articles before on the shift in data architecture. What changes do you see organizations making to their data architecture?
RB: It’s like what I was talking about before regarding the shift from the traditional data warehouse to the analytical sandbox. Things are shifting to give users the ability to mitigate the upfront data engineering. In other words, rather than having to clearly define and engineer 1000 data variables, you load the data and find which data shows the greatest correlation between the outcomes that you’re looking for. Then, you focus on those elements of data and load more of that data, and you do the ultimate data engineering on that resulting data, which is the most critical and meaningful to your organization. It’s really a case of focusing your efforts and dollars on that substantive data that’s most relevant to you.
ML: In your articles, I’ve seen you advocate for architectural revolution, overall digital revolution and the hiring of chief data officers within organizations. But many companies trying to go this route are often shocked at the high cost. How do you justify the business case for going through this?
RB: Simply put, the more data you have at your disposal, the more informed you’ll be about your business challenges and successes, and the more insights you’ll have to help change your company for the better. It’s often useful to describe the advantages of data with examples from the worlds of sports or politics. So, for example, if a baseball pitcher looks visibly tired and is grabbing his arms, you might think that he’s done for the game. But these days, complex statistics and algorithms have been developed to suggest historical performance of pitchers based on all kinds of factors. This allows you to make a better guess of the pitcher’s performance versus relying solely on one data point (how the pitcher appears to be acting).
The same principles apply for businesses; having data and incorporating it as part of your analysis, whether it’s 50%, 80% or 30% of what informs the ultimate decision, makes any organization more informed since they have more to work with.
ML: So, let’s build on that. Let’s say you’ve already convinced the innovative thinkers in your organization to use more analytics. At what point does digital analytics move from that initial entry point to becoming a widely-adopted process that everyone in the company sees value in and wants to leverage?
RB: The goal is to push the analytics down through the company and then embed it in a broader set of business processes, particularly those that are customer-based. I think when others in the company see how much information they can gather through those customer interactions–and how easily they can access that information–they feel empowered to make decisions themselves because they have the information they need to do that.
ML: So, everyone is talking about “the race to insights.” Technology makes it fairly cheap and easy to collect relevant data, but how do we most effectively take this raw material and turn it into something that can create business value?
RB: Well, if you can figure out something faster than the other guy, you have an advantage. Organizations need to act quickly on the insights they gain from their data. It’s also important for them to integrate their data into the development of products, services or capabilities once they’ve analyzed it. It’s not good enough just to accelerate the speed to analytics or insight. You need to get to the point where you can accelerate your decision-making and ultimately be able to quickly adapt and modify your capabilities. It’s all about being able to take advantage of that insight and to put it in the hands of the person who needs it at the time.
ML: It seems like every report that I read talks about projected shortage of analytical talent. Companies are hiring or competing for talent. This seems like a bubble that is inevitably going to burst. What can we expect from this trend?
RB: I actually wrote my last Wall Street Journal article on how the data scientist role is considered one of the sexiest jobs out there. However, it’s important to keep in perspective that organizations have had these sophisticated analytics folks around for years in a variety of iterations. And historically, they have been among the first to be eliminated in economic downturns because they were not in direct revenue-producing positions. People with analytics skills will be of tremendous value to organizations, but we shouldn’t necessarily assume there’s going to be a straight outward trajectory for those types of roles.
ML: Last question–you’ve been in this field for a couple of decades now. What do you wish you had known about analytics when you were first getting started?
RB: Well, I actually started on the technology side before moving to the business side, and we were collecting large amounts of data that needed to be maintained and stored for accounting, record-keeping and operational purposes. And at the time, it struck me that this data could be organized and analyzed to understand and identify customer behaviors, but that wasn’t what it was being used for at that time. So, from the outset of my career, I had more interest in what could be learned from the data than what organizations captured, and that really started my ongoing quest. There is so much information out there to mine and to learn from. We’re not so much bound by the data that we have; we’re bound by the way we use our imaginations. What types of patterns and correlations should we be thinking about and looking for? Our capabilities are ultimately determined by whether we’re asking the right questions.
2. The Big Data Business Adoption Journey – The Wall Street Journal.
3. NewVantage Partners Big Data Executive Survey 2014.
4. NewVantage Partners articles and resources about Big Data.