FEI Daily: What were the survey's key findings?
Wilds Ross: From our perspective, the key findings out of this were, number one, we found that only about a third of the CEOs that we surveyed have a high level of trust in the accuracy of their data and analytics work, which was really surprising to us because we thought that there would be a much greater confidence expressed by the CEOs. Only about a third of them felt good about it. Another third had either limited or active distrust of their D&A work, which is also a serious finding. If a third of the CEOs out there just don't believe what they're seeing from their own organization in terms of the analytics that they're doing, how could they base their decisions on that?
Even separate from actively distrusting the analytics work, more than two thirds believe that the data that they're basing their analytics on is also not trustworthy. This all comes together for us to tell us that there's a trust deficit out there in the market.
CEOs say that they're using D&A to develop new products. And about half of them said that they're using D&A as ways to find new customers. Developing products and finding new customers are both around growth. This tells us that where they're focusing their attention is growth of the business, but they don't yet trust these results enough to be able to capitalize that. And that's why, I think, you see a lot of other articles coming around that are saying, "Why hasn't D&A lived up to its full promise?" And we believe that that really has to do with trusting the results that you're producing.
FEI Daily: How are companies using data and analytics for their businesses today?
Wilds Ross: The first is growth, followed by compliance, and cost control. D&A is turning out to be one of the most important tools that banks and big financial institutions can use to comply with this ever-increasing complexity in the regulatory environment. They're being asked to provide huge amounts of information and to accurately represent how their business is interacting with their customers. And that requires pulling together lots of different data sources from different systems and putting them together as well as being able to accurately model on the impact of certain conditions within the economy.
FEI Daily: What is the biggest concern executives have about their data and analytics?
Wilds Ross: Quality of the data is the primary concern according to this survey, and then the second is the reliability of the analytics work. The reason that this is so important is because the people that are doing this analytics work, these data scientists, they might understand what they're talking about, but for the CEOs, there's quite a barrier to understanding the detailed work that these people are doing. So, some of the CEOs are just saying, "I don't get it. I just don't trust it. They haven't been able to convince me that what they're doing is correct and accurate." There's a lot to be done there around building trust in the C-Suite, and that involves things like really good peer reviews and quality assurance work. It involves getting companies to come in and provide an independent third party review of some of that work. On the data side, it involves really getting serious about tackling the data quality problem, which has been talked about for decades, but there hasn't been this real concerted effort to resolve it, and that has to be overcome before the results of the analytics are going to be useful.
FEI Daily: Do executives have a real good reason to be distrustful of their data?
Wilds Ross: Well, that's a fascinating question. I am a data scientist, so I do this work. And from my perspective, there's a lot of good analytics that you can do on data that's not 100% correct. In fact, it's never going to be 100% correct.
Because in many cases these are statistical processes, you can accommodate some level of discrepancy in your data, but not a high degree of discrepancy. It has to be mostly right, but that's a really difficult concept to explain to someone who has come up in a business world where accuracy is really, really important. So, when you come back and say, "The data is not 100% correct, but we still have a high degree of confidence in our results," that just doesn't make any sense to them. Executives need to recognize that in some cases their data may be good enough.
On the other hand, you do need to enter this with a healthy dose of skepticism. I've seen a lot of academic data scientists come into the business community, unaccustomed to having their work questioned. They may think you're supposed to trust them based on their reputation, and that's insufficient. You can't trust your business analytics based on the reputation of your data scientists. You have to trust it based on the integrity of their work.
FEI Daily: What findings from the survey are most important to a financial executive?
Wilds Ross: The fact that the CEOs look at data analytics as one of their top investment priorities for the next three years is significant. The CEOs know that they have to get on top of this. Even though they don't trust the results all the time, the possibility that they would fall behind in this area could be catastrophic.
We see companies like Uber, Amazon, and Pandora that were built from the ground up on analytics principles. More mature organizations are rightly terrified that their position in the market is going to be eroded by companies that are making better use of the data and analytics than they are. So they're putting a lot of emphasis on investing in this area.
FEI Daily: How can CFOs work with their CEOs to guide the right spending on data analysis?
Wilds Ross: They need to understand where the business currently is in terms of analytical maturity. I don't have a high degree of confidence that CEOs really understand what the potential out there really is. The number of concerns they have is immense. They’re not going to sit down and roll up their sleeves and dig into what this means, but I do think they need to think about how are they are going to assess their current level of maturity, and how are they are going to build a road map for improving that maturity.
They have to understand where they sit and also have to articulate a clear business strategy that takes them forward into the mature state that maintains their competitive advantage over the coming years, and prevents their position from being eroded by these other comers. I think they also need to think clearly about what the value of this is. There's been a lot of talk over the last couple of years about how to determine what the value of D&A is in the business. Is it that the data itself has value? Is it that our ability to address the market more efficiently? Is it the way that we're negotiating with our investment community? I think the CFOs and the CEOs need to have really forthright discussions about what they're trying to accomplish and determine what the real value of these activities in their particular business sector and their particular business environment is.
FEI Daily: Who can a CFO enlist to help them understand where the business currently is in terms of analytics?
Wilds Ross: I think it's important to get a third party to come in and have an objective look. I feel that that's important because if you ask someone internal to the business, and this would be true for any business process, "Go and assess the efficacy of our IP operations," they're going to give themselves a glowing review.
Another benefit to getting a third party is that they have experience with other organizations. They've seen it across different industries, they've seen it across the different sizes of different organizations, and all of that helps to give context around a way a business is performing. A business looking at itself often has difficulty being honest with their capabilities.
Wilds Ross is Principal of Data & Analytics, KPMG. A data scientist by training, Wilds works closely with service line leaders across the firm to bring Data & Analytics capabilities into KPMG’s service offerings.