BI Radio
Episode 11: The Analytics Episode
(00:00)
Station ID: This is BI Radio.
(00:10)
Montage:
- There are many and varied issues.
- Build up some analytical capabilities.
- Actually with analytical work you can deal with missing data more effectively than you can in traditional business intelligence.
- Balancing all of that together, they're using analysis to drive performance.
- Everybody can get benefit from analytics.
- A significant increase and agility.
- Ten or twenty years ago such a thing would have been unheard of.
(00:35)
Ken Seeley: Hi there. And welcome to BI Radio. I'm Ken Seeley. On the show today, competing on analytics. We hear from best selling authors Tom Davenport and Jeanne Harris about crunching the numbers and the new science of winning. Plus, Kelsey Howarth talks to our own Laurence Trigwell about competing on analytics in the financial services industry. But first up, analytics, part one. Delaney Turner talks to Jeanne Harris about the new breed of analytical executive and how to catch up if you're among the analytically impaired.
Delaney Turner: Hi there. I'm Delaney Turner editor of Cognos Performance Perspectives. I'm speaking today with one of the co-authors of Competing On Analytics, Jeanne Harris. Before we get too deep into analytics themselves I would like to discuss a sentence that struck me right off the top in Gary Loveman's introduction. He said it's his job not to have all the answers but rather to ask penetrating and occasionally offensive questions that lead to insight. How different is that from the traditional approach of the CEO, and how common is it to find CEOs that take that approach?
Jeanne Harris: Well, certainly, Gary is not the typical CEO. In some ways I think he is on the vanguard, or a reflection of how executives are changing. We have a new generation of leaders coming into the C-suite who have grown up taking MBA programs, who are used to fact-based decision making and a more analytical approach to decision making.
Delaney Turner: The book is called Competing On Analytics. And you do provide some very competitive ROI studies to support your argument. But you also write that analytics alone are not enough. You've outlined that companies must apply them to a distinctive capability. They need executive level commitment and an enterprise level approach to managing them all. How difficult is it for a company to put all of these pieces together? And can it still compete if one of these pieces is missing?
Jeanne Harris: Well of course many companies compete without any of these pieces. They just don't compete on analytics. Those companies that compete on analytics, however, do generally take some time in order to pull all the pieces together. This is a somewhat complex process. However, whether you compete on analytics, that is, analytics really are your differentiating capability, or whether you simply use analytics to manage your business more responsibly and effectively, everybody can get benefits from analytics, and the kind of business intelligence tools and capabilities that they require. The catch is: to what extent is the impact going to be on your business? Most companies start by using analytics in a fairly tactical or mundane kind of way. They do section reporting, or they look for a small kind of incremental improvement. Companies that compete on analytics are really differentiated by the fact that their use of analytics, not just in one aspect of their business, but in several or many, really provide them a competitive differentiation.
Delaney Turner: Could you give me an example of a major business decision that a company would made differently based on analytics?
Jeanne Harris: Well, you know, just a few years ago companies were making major decisions about what companies to acquire based on nothing more than a few conversations and a handshake. Today, it's a much more analytical kind of process. In fact, one company that we profile in our book is called Cemex. They are one of the worlds largest distributors of cement. And they actually use analytics very effectively to identify and evaluate their acquisition target. But they don't really just stop there either. They choose analytics for something consider to be very strategic which is determining how to staff that acquisition in order to best integrate them into their organization. So analytics are used in every step of the merger acquisition and integration process, whereas I think probably ten or twenty years ago, such a thing would have been unheard of.
Delaney Turner: Provided that companies want to make the leap to get better at analytics, what would you recommend to them as the next step?
Jeanne Harris: Well you know, it really does depend where you are in this journey. A company that's analytically impaired doesn't even have the basic underlying data they need to use to make the most fundamental kinds of management decision. They don't have executive dashboards. They probably don't even have the data to produce those dashboards. So, if you think about a company that's really just struggling to find the data, you know, the first step really is to stop and go out and pull that information together. Assuming they have that then they really move into our next stage. They probably have started to use analytics in some basic kind of way. And in many companies, really, as you alluded to earlier, actually at that stage they are companies that probably are doing a couple of things pretty analytically. After all, lots of people have access to Excel as spreadsheets. So there certainly are a lot of people using and manipulating data. But they're not necessarily doing that in a way that helps create any business value. What we start to see in the next stage of localized analytics are companies who start to pull off some of the low hanging fruit and do small kind of functional applications that demonstrate and prove their worth as they go. And really, for most companies, that's the way to start to build up some analytical capabilities.
Delaney Turner: Now let's look at the other end of the spectrum for a minute. One commonality of the best analytical competitors is the close alignment of business and IT. What does an IT business conversation look like in these companies, and what are they talking about that's different from their competitors?
Jeanne Harris: Well, excellent point, because there's a huge diversity in the way CIOs approach analytics in their organization. I think in the companies that truly compete on analytics the CIO has a very high profile role, sometimes, actually, as the business owner of the analytical capability in the organization. In many of the other organizations you think of, their role really is to be a change agent. They help introduce some of these concepts and connect them to the formal value proposition. And of course, another obvious thing that they do is they're responsible for building the overall business intelligence architecture as an element of the overall IT architecture to make sure that the organization has the right tools and infrastructure that is going to enable analytical competition.
Delaney Turner: It's interesting you mention business intelligence. You've dedicated an entire chapter to the architecture of business intelligence and you say that one of the sign posts of an effective IT organization is that managers never argue over whose numbers are accurate. Could you give us some insight into how analytical competitors use BI and how it differs from companies that compete in other ways?
Jeanne Harris: Well, you know, one of the things that was I think somewhat surprising to us as we were researching this book was just to what extent business intelligence was really viewed as an enterprise wide capability, and an enterprise wide resource amongst companies that truly compete on analytics. That's not true for companies, typically, that are at earlier stages in their development. Typically, as you may know, business intelligence often develops as a result of an initiative often outside the IT organization. So it may come from the controller's office, or from the CFO. Or from somebody from marketing who has a very specific kind of analytical project in mind. But as companies develop analytics more broadly across the business enterprise, as they start to view it more strategically, they realize their synergies across all these different prophecies. In fact, some people have almost defined it akin to analytical meta-process that overlays across all their other business processes. And so effectively to address all of those concerns, you really do need to take more of an enterprise wide approach to building your business intelligence architecture.
Delaney Turner: Competing On Analytics: The New Science Of Winning, certainly is a compelling read and almost a call to action. Jeanne Harris, thank you for speaking with us today. Thank you very much Delaney. It's been a pleasure.
(09:50)
Advertisement: Got perspective? Cognos Performance Perspectives is the Cognos e-newsletter for business intelligence, enterprise planning, and performance management. It's your source for exclusive interviews, insights, and opinion on industry trends. It's a perspective unique to Cognos, one we think you'll enjoy. Get the newsletter and get a new perspective. Subscribe now at cognos dot com slash newsletter.
Station ID: Interviews, insights, and opinions on performance management. You're listening to BI Radio.
(10:30)
Delaney Turner: Hi. I'm Delaney Turner. Today, I'm talking to Tom Davenport. Tom is the co-author of Competing on Analytics, and the President's Distinguished Professor of Information Technology and Management at Babson College. Tom, welcome to the show.
Tom Davenport: Glad to be here.
Delaney Turner: Jeanne Harris mentioned that most companies are in the early stages of competing on analytics. If they're not outright analytically impaired, they're certainly not all analytic competitors. And if we look at the progression of performance management initiatives most industry analysts would say that companies are also in the early stages. What link do you see between analytical competition and performance management?
Tom Davenport: Well I think they are very related. And I think of performance management as something that you can do in a couple of different ways, to put it bluntly. You could do it with reporting, which is certainly the vast majority of what organizations have done in performance management. They're reporting about the past. Mostly in standard reports, sometimes in ad hoc reports and queries and alerts. But it's about something that already happened in your business. And not much of a relationship between the different things you're reporting on. The other thing that one can do in performance management is to do it analytically. And that means to start using more quantitatively sophisticated approaches to looking at trends. Looking at drivers of performance. And looking at relationships between non-financial performance measures and financial performance measures. And I think this is going to happen. Although we're in an earlier stage relative to performance management analytics than we might be, say, in marketing analytics, or supply chain analytics, or some of these more operational forms on it.
Delaney Turner: Now at Cognos we talk a lot about the different personas or users of performance manager within a typical company. Business managers, executives, frontline workers, and so on. You've mapped out different analytical personas. Could you outline what those personas are, and then how they work together to drive an analytics based strategy?
Tom Davenport: Sure. We talked about three different types of analysts, if you will. There were the professional analysts who, in most organizations, represent a relatively small proportion of the total. And these would be people who have very strong quantitative backgrounds, maybe PhDs in some quantitative oriented discipline. And it would be their job to kind of come up with new algorithms, new metrics, new ways of analyzing what's happening in the business. Then you have what we call the analytical semi-professionals. A typical educational background for someone in that job would be an MBA, for example. And they can typically do some statistical work. Maybe some regression analysis, for example. A lot of people are taught that in MBA programs. And they can certainly do some visual manipulation of data and can create some pretty powerful models in a spreadsheet, if not in a BI tool or a statistical tool. And then an audience, a persona that's often overlooked is what we call the analytical amateur. These are typically people at the front lines of organizations who may not be creating the model, but have to use them say with customers. And so a call center representative who has an analytical model determining what would be the next best offer that you would offer to a customer. Say if I'm in a bank and you're talking to someone in the call center and they say, put all that money in your chequing account into a certificate of deposit. I think there's some need for those frontline people to have some sense of how those recommendations were arrived at.
Delaney Turner: If we look at the data itself what strategies does an analytical competitor put into place to be successful? And what technologies do you see them investing in?
Tom Davenport: Well initially a lot of the same kinds of strategies that you would apply to data for business intelligence are needed for analytical work. So the data needs to be, as you suggested, fairly clean. Often, actually with analytical work you can deal with missing data more effectively than you can in traditional reporting oriented business intelligence. So statisticians are quite comfortable with dealing with missing data. And if there's too much of it they can actually do estimations of missing data. So that's a little bit less of a problem. But in both cases I think you have to have data that's fairly common across the enterprise. It's hard to, for example, calculate some predictive measure of what a customer might buy when you have a different definition of a customer across different business units, or product groups, or whatever. So that consistency, that commonality across the organization. And of course there's a certain accessibility: getting data in a place where you can actually manipulate it. Traditionally we thought of that in BI in terms of enterprise data warehouses. But for analytical work, typically, we would put it in smaller more analytically oriented marts, if you will, for doing the actual analysis.
Delaney Turner: The job of managing the analytics. It seems to be similar to managing what's known as a business intelligence competency center. You outlined four models large companies have chosen. I was wondering if you could let us know or tell us how they made that choice? And how would a company decide which model is right for them?
Tom Davenport: It's sometimes hard to talk about it in the abstract. But as you suggested some companies have taken all of their analytical people and put them in one part of the organization. And in many cases, it's IT. The analytical professionals are part of the IT organization. And that seems to work pretty well. It makes a lot of sense to do that when you have analytics taking place across a whole variety of functions within your business. If your primary focus is in one functional area, say marketing, then it probably makes sense to have the analysts be dedicated within that function rather than putting them in IT. Although, as I said earlier, you do need a close relationship to the IT organization. There are some cases where analytical people might be in the strategy organization. So I think it's a function of what executives are really interested in doing analytical work. What is the particular target of the organization as far as where its analytical capabilities are focused. Is the executives who might manage these people really comfortable with that idea, and converse it with some of the basic ideas in analytics so that they can provide a sympathetic home for people who think that way?
Delaney Turner: Well if I am a CEO it's obviously some very important decisions to make. But the benefit of analytics is very clearly outlined in a very compelling argument. Tom Davenport, thank you for talking to us today.
(19:00)
Advertisement: Virtual venue, rock solid content. Cognos Virtual Finance Forum is our virtual conference and expo for finance executives. Get practical advice to solve your financial performance management challenges. You'll see effective ways to plan, forecast, and control resources. Hear from successful Cognos customers, and access on demand presentations from Tom Davenport, co-author of Competing on Analytics: The New Science of Winning, and Bryan Hall of the Hackett Group. It's all happening June 4th. Register today at cognos dot com slash virtual finance 08.
(19:35)
Station ID: Insight on performance management from the people who shape the industry.
Kelsey Howarth: Banks and financial institutions have tended to be ahead of the curve in competing on analytics. They are early adopters of technology, and as such, have made performance management a key mandate in their organizations. In this segment you'll listen in as I speak with Laurence Trigwell, Associate Vice President for Financial Services at Cognos. Laurence has more than twenty years financial services experience in helping customers leverage technology to meet performance challenges and gain competitive advantage. Here, he shares how banks are using Cognos to gain better customer and risk and insight.
Kelsey Howarth: What are some of the major challenges facing the banking industry?
Laurence Trigwell: Well I think there are many, and varied issues facing banking, obviously globally, but I think fall really into two broad categories. First is what we would refer to as governance, risk, and compliance. Responding to the regulatory theme for tracking increased transparency and disclosure, partly; driven by the high profile situations in banking for the last few years; and the ongoing situation in banking currently. And the second of those categories would be what we would consider as being the other side of the coin: Driving how, specifically, satisfying shareholders and stakeholders to meet their gross and margin objectives.
Kelsey Howarth: How are banks using Cognos to compete on analytics more effectively? What are some of the key metrics that they're really tracking?
Laurence Trigwell: Given the two challenges facing banks at the moment around governance, risk, and compliance and growth and margin performance objectives. So banks are using analytics to fundamentally just sweat the assets that exist in the banks. So is that a class leading set of products? So they've got to make sure that they're selling the right product in an appropriate volume. Or is it about sweating a particular channel or group of channels. So they need to be able to make sure they're optimizing the capacity of those networks both in terms of setting an appropriate set of plans, but also, measuring that performance against those plans. And, of course, in a multi channel environment it's about managing both the branch networks, but also the online presence of third party and agency presence, as well as call center and telephone banking networks. Whether that asset is in fact a customer asset or a capital asset. So in a customer base, for example, banks have invested heavily. Sometimes through acquisitions, sometimes through organic development of a customer base. Often, obviously, a combination of both of those. And within that asset if you can think of customers as a core asset, it's got to be about making sure that those customers are serviced appropriately with the right product, at the right time, through the right channel. And obviously they're looking to do that as efficiently as they can in terms of their risk profile and their risk appetite setting. So they're to, again, not attract the wrong type of customer at a price that's inappropriate. So balancing all of that together they're using analysis to drive performance.
Kelsey Howarth: Can you give us some examples?
Laurence Trigwell: Sure. So one great example to think about from an analytical point of view, for example: One of the world's largest banks has a particularly large branch network. Not unlike many of the worlds largest banks. And one of the challenges it was facing in terms of meeting its particular growth objectives was balancing the compensation plans for staff in those branches with its particular branch footprint and its branch service plan, and sales performance. And they had a monthly view, at best, of how performance was happening. So one of the things they needed to do was both change the compensation plan at a branch level and simultaneously provide a mechanism to aggregate that information on branch performance. What that typically meant was that manual, untimely process meant that management reports for branch performance were taking seven weeks to collect the information, to build it into a consistent format, and then to subsequently distribute it. And what that meant of course was that people were dealing with old information and infrequently dealing with that information. Well, they've subsequently been able to replace that seven week process with a seven second process. So branch managers and relationship managers within the branch now receive daily performance reports within seven seconds of the information being available to them. What they now get of course is one decision point for every trading day. So for example 120 decision points a year. And that's a significant increase in agility.
Kelsey Howarth: Any other examples that spring to mind?
Laurence Trigwell: Another example that I wanted to talk to really was in the area of risk management. And it's obviously very topical and where rich concentrations are essential to provide to senior management for all sorts of rather topical and painful regions right now. And one great example is one of the large European banks are using Cognos to provide risk insight to senior management in a way that they simply weren't able to do before. And the ramifications of that were that it would typically take between two or three weeks of risk analysis from coordinated groups of finance, capital management, and senior management to understand the concentration of risk in a particular industry group or region or geography or legal entity or customer class or even product group. And they now are able to do is that the senior management in that group as a divisional and a group level, in fact, get a consistent view of those risk concentrations. And that's really powerful because they're now able to manage with some confidence on a daily basis the concentrations against a customer or product orthography.
Kelsey Howarth: For more information on Cognos Solutions for banking please visit our website at www dot cognos dot com slash banking.
(28:25)
Advertisement: Want better visibility into your bank's performance? Cognos is the Hubble telescope for banking's black holes. With seventeen years of proven performance management experience. Like our performance management framework that shows you the sweet spots of information to truly make a difference to your performance. Or our IBM Cognos performance blueprints that give you a head start to better decisions in key areas like customer profitability and risk. Get better visibility into your performance. Visit Cognos dot com slash telescope today. That's cognos dot com slash telescope.
(28:55)
Ken Seeley: Well that's a wrap. I would like to thank our guests today. From the Accenture Institute for high performance business, Jeanne Harris. From Babson College, Tom Davenport. And from Cognos, an IBM company, Laurence Trigwell. A special thanks as well to our contributing producers Kelsey Howarth and Delaney Turner. Thanks to producer and audio engineer Derek Schraner who composed all the original music you hear on BI Radio and for making us sound so good. A reminder to check us out online at radio cognos dot com where you can listen to previous shows, download individual segments, and view the transcript of each broadcast. If you have a question or care to comment on what you hear on BI Radio send on email to radio at cognos dot com. Thanks for listening. I'm Ken Seeley. See you in about six weeks.
(30:00)
END OF RECORDING
RSS Feeds
