Don Campbell
Stephen Brobst
What is “Business intelligence” for, if not to give a company a competitive edge? What’s the state of competitive BI today? Have enterprises been able to fully exploit BI for all its promised potential or has it turned out to be more difficult in execution then expected? Have there been any breakthrough innovations to better identify and develop opportunities for a competitive edge? Besides the tools, what else is needed for BI to become “competitive” BI?
Contributors
Transcript (Delivering Smarter, More Competitive BI)
Speaker A [00:00:01]:
The following program is being brought to you on the Voice America Business Channel. For more information about our network and to check our additional show hosts and topics of interest, please visit voiceamericabusiness.com the Voice America Talk Radio Network is the worldwide leader in live Internet Talk Radio. Visit voiceamerica.com the views and ideas expressed on the following program are strictly those of the host or guests and do not necessarily reflect the views and ideas held by the Voice America Talk Radio Network, its staff and management.
Speaker B [00:00:33]:
Welcome to CIO Talk Radio with your host, Sanjog Aul. All comments, views and opinions expressed on this show are strictly those of the host, guests and callers. Here’s Sonjog Aul.
Sanjog Aul [00:00:46]:
Good morning and welcome to CIO Talk Radio. To learn more about the show, please visit www.ciotalkradio.com. Today’s topic is Delivering Smarter, More Competitive BI.. Our guests for today’s show are Don Campbell, who is the Chief Technology Officer for analytics and performance management with IBM, and we have Stephen Brobst, who is the CTO with Teradata Corporation. Good morning, guys.
Stephen Brobst [00:01:12]:
Good morning.
Don Campbell [00:01:13]:
Hi Sanjog.
Sanjog Aul [00:01:16]:
How’s life? Where are both of you located, respectively?
Don Campbell [00:01:19]:
I’m located in Ottawa, Canada, where we’re just going to start our winter season and we’ll be snowed in for months at Teradata.
Stephen Brobst [00:01:29]:
At Teradata, Our main development lab is in San Diego, California, although I’m in Chicago at the moment.
Sanjog Aul [00:01:34]:
Okay, so you’re not that far off when it comes to temperatures, but I envy you for the San Diego weather for sure. So do you happen to actually go there and are there or you’re mostly traveling?
Stephen Brobst [00:01:46]:
I spend a fair amount of time in San Diego as well as in LA, where the second lab is but we also have labs in India and China and Toronto and Raleigh, Durham. So I’m all over the place.
Don Campbell [00:01:58]:
But let’s be real, Stephen, whenever I want to get ahold of you, if I want to send you a Christmas card, I usually send it to United Airlines because they’re likely to see you more frequently than your colleagues do.
Sanjog Aul [00:02:09]:
Very good. So thank you so much again for accepting the invitation. This topic is BI people. For some it is a sacred cal For others it is a must have and then there have been situations that have resulted with lots of investment being made and then they did not see. They say, oh, you gave me a bunch of reports. It doesn’t do me any good, or I don’t know how to use them.
Sanjog Aul [00:02:35]:
So which basically is a thumbs down to such initiatives. So if you have to spend money where money is so tight and if you see such thumbs down type of environments, how does one say that? Okay, I will give you a competitive BI let alone more analytical and just transactional insight using this methodology or paradigm. So Don, let’s start with you.
Don Campbell [00:03:03]:
Well, you know, we’ve seen customers get very good return on their investment in their BI and you know, it can be in a small, short term kind of initiative. That’s usually where we ask the customers to start to get some good feeling for the technology, get a good return on investment fairly quickly. We’ve seen the big wins too. You can get a return on investment in a month’s kind of implementation as well too. BI continues to be something that customers look at as a must have. It continues to top lists of investments for CIOs and often that’s just having the company run better and faster and smarter
Don Campbell [00:03:52]:
but in this competitive world, there’s a big need to be able to do it at least as fast, if not faster than your competition as well too and I think with the right kind of planning and the right kind of technology behind it all, it can really return the investments that you’re looking for.
Stephen Brobst [00:04:12]:
But I would suggest that real competitive advantage absolutely doesn’t come from just reporting, right? I mean, you have to build the business intelligence into business processes. You have to go into advanced analytic capabilities because reporting, everybody has reports, right? You’re not going to get competitive advantage from yet another key performance indicator report unless you really start building your operations and your processes around what you learn from the business intelligence. So my dream of business intelligence is reports go away and you’re delivering business intelligence embedded into process. Truly pervasive BI, not this kind of traditional file cabinet type of BI
Sanjog Aul [00:04:57]:
So when you mentioned that there are tools and technologies and you can use always to do your due diligence to find the best tool that’s out there and there could be some either gaps or lack of understanding or expertise in the way you bring about BI into an organization and then get it done right, and then finally get it adopted till all these three things happen. The end results are not seen and are not measured and are not. And this whole thing is not blessed, if you will. So what do you think are organizations doing who are not as successful and are satisfied with BI in the first place before it even turns into a competitive BI, Stephen?
Stephen Brobst [00:05:43]:
I think those that are not as successful are typically starting from an IT perspective rather than a business perspective. There’s this kind of myth that, you know, we will build it and they will come data is strategic. Therefore we need to get lots of it and put it in a big database. And organizations aren’t successful when they take that approach. So there really needs to be a partnership between IT and business in order to be successful. And if we treat this as IT driven initiatives or data for its own sake, you’re certainly not going to get the return on investment. I do believe that successful organizations absolutely measure return on investments and get into a self funding model rather than a BI as a cost center type of model.
Don Campbell [00:06:34]:
BI is a tool to help you implement and understand your strategy and if you don’t have your strategy in place in a way that you can turn it into discernible trackable metrics, then the BI is not going to give you any valuable insight. We had in our own case, we used to track things like the tenure of employees in the company and we would watch that number change, but we didn’t really know whether the fact that it’s going up meant that we were getting old and outdated or that we were getting a lot of good corporate memory. And so you still have to understand what the metric is that you’re tracking and how it really applies to your strategy. The BI can’t help you do that, but it can certainly expose it for you.
Sanjog Aul [00:07:21]:
Now how deer in headlights type of response or expression that you see from business when you go and champion it, or do you think mostly the companies where it is successful business is the one who brings it because they heard about it in a conference?
Don Campbell [00:07:38]:
Well, I think we were in a state years ago where we had to sort of sell the value of business intelligence. That’s not the case anymore and I think corporations understand the value of it and understand that they need to have it implemented, even if they’re not sure exactly how to do it holistically and they look for help to do that,tThey look for some guidance and some best practice on how to implement it successfully. We’ve seen various tactics, we’ve seen the tactic of trying to implement the entire system across the whole organization with support from the CFO on down and we’ve seen the departmental installations that that have a success in one department and then starts spreading to others and so the tack to it can be variable
Don Campbell [00:08:35]:
but in the end, the more they get success, the more they breed success and the more knowledge they gain about how to do it inside of their organizations and how to overcome those cultural boundaries as well.
Stephen Brobst [00:08:49]:
But it’s unlikely that that success is going to come from someone going to a conference and seeing the latest vendor demo shiny new object that attracts their attention for some short period of time. You have to have a real business problem that you’re solving and business intelligence and data warehousing is only successful as a program of ongoing exploitation of the information assets, not as a collection of point projects per se. I mean, you have to do projects of course in the context of the overall delivery of the capability but the best practices clearly are going to involve a program of continued investment, measurement of return on investment, and improvement on the capability of delivering information to knowledge workers and organizations.
Sanjog Aul [00:09:40]:
Now would you say that the BI initiatives usually are started by organizations, especially when they don’t have much in there already? Is it to come to know what they don’t know because they feel that it would hurt them if they don’t know?
Stephen Brobst [00:09:57]:
I think there’s usually some particular source of pain that they’re experiencing and they’re looking for a solution to the problem. So if I’m a retailer and I’ve got a customer dissatisfaction related to out of stock on my store sales, I’m going to look for solutions to try and figure out how to get my in stock performance better and you know, business intelligence is an important way of solving that problem. Or if I’ve got, you know, a significant amount of fraud going on, then I’m going to use business intelligence to understand why and figure out how to prevent the fraud. So there’s some pain point, there’s some opportunity that intelligence can help knowledge workers make better decisions on how to change processes or how to execute at the point of interaction during business events.
Don Campbell [00:10:50]:
And I think the more mature organizations are starting to realize that they’re collecting a lot of data and they now need to look at their business from an information centric approach and looking at the information asset as a strategic advantage for the company and that information led agenda is really transforming sort of how they look at their data, how they look at themselves and how they manage their information and some of them that have been doing BI for a while are now allowing that information asset to really lead the company into that next level of performance.
Sanjog Aul [00:11:32]:
When you use the word holistic, do you think that is a practical approach? Would you lose? Basically what we are seeing is that if on one hand, if you try to do it in a siloed mode, or it is rightfully so, because each division of a company may have their own business intelligence related needs where they could say, okay, this is where if we find out more information, we might be more competitive in this area versus another. But then if you do it Such a siloed mode, then you would lose the opportunity for cross departmental BI if you will, which will bring even some of the other connections which were not even obvious to people earlier and then that could take them to the next level but then if you do that, you might also start doing too much in one shot and start losing the individual’s picture. So what is that balance in terms of the way BI is actually most effective?
Stephen Brobst [00:12:27]:
Don I think there’s a great paper written by Mohan Sani that was published in Hybrid Business Review a couple years back where he points out that the best, the most successful organizations centralize data in order to enable decentralized decision making. It’s a little bit counterintuitive, but by bringing the data together, you’re able to allow decisions to get pushed out further out in an organization and have more agility in in decision making versus if you decentralize the data, that actually forces more centralization of decision making because then people have to come together to decide whose numbers are correct and how to put the numbers all together and get agreement on the performance of the organization. So centralizing the data and decentralizing the decision making, pushing the business intelligence as far out in the organization, very, powerful and so you can have different parts of the organization have different BI capabilities, but you’ve got one single source of truth to drive the analytics within the organization.
Don Campbell [00:13:42]:
The definition of the customer, from the sales department to the marketing department to the support department is all different. And how they use that information may be very unique and very different. But if you lose track of the combination and the knowledge that you can have by sharing that information across departments, then you really lose the opportunity to drive the whole company forward as one. So implementing smaller siloed kind of systems isn’t bad on its own, but it doesn’t support sort of the general needs of the business to move forward so you certainly have to be able to tie those capabilities and those information sources together in order to see the business across all of its aspects.
Stephen Brobst [00:14:30]:
That’s a good example of organizations looking at the same data in different ways. I was working with an insurance company and in their case, some people would look at revenue based on policy sold in the sales organization but in the finance organization, until I collected the premium and put it in the bank, that didn’t count as revenue. So they’re sold premium and billed premium and earned premium, all kinds of different variations and they’re all valid definitions for different parts of the organization. The key is that you can have one place to go where they all reconcile with each other so having different views of the same data is absolutely appropriate. Just like in your example, different views of what a customer means and that’s important for the agility of the decision making in the organization
Stephen Brobst [00:15:22]:
but it’s also important to have coherence and the ability to share information in a consistent way, even if you take different views of that data.
Sanjog Aul [00:15:32]:
Let’s take a quick break listeners. We’ll be right back after these messages and look at what could BI do or is truly this is a given capability of BI discipline that it could allow us to even identify the actionable steps after we have sliced and diced the data and we found some, some areas of improvement because at the end of the day people might be looking at it, they might find this is what they are supposed to do but then how much, to what degree and how will that impact the future data? So is there some forecasting, some recommendation or outlining of actionable steps? That is possibility by using BI, whether BI tools or a BI group or set of experts who could make this happen and this becomes part of the BI department. Please stay tuned. We’ll be right back.
Sanjog Aul [00:18:57]:
Welcome back. So yes, we could analyze the data all day and we eventually have to take some steps, whether it’s based on probability or some empirical data and slicing and dicing that we have done. Has BI reached a point as a discipline where it could, along with the slicing that you did, you put in those factors, those equations and here comes a set of actionable steps given the situation or some outlining of tasks which could be done at different levels, whether strategic, operational or tactical. Is that what in the you can say scope of BI or should it be?
Stephen Brobst [00:19:39]:
Absolutely, I believe it is in the scope and I believe that the industry has made some very interesting steps towards the evolution from a data warehouse and business intelligence environment as being focused mainly on strategic long term decision making to the point where now a significant part of the investment and deployment of BI is about operational business intelligence, which I think is along the lines that you are speaking, where you are translating the decision making into specific actions that’s going to help me execute more effectively. Because in the end your strategy is only as good as the execution. So you need the knowledge workers on the front lines of the organization who are taking those actions to have business intelligence. Just like you have the corporate IV power decision makers requiring data for strategic decisions.
Don Campbell [00:20:32]:
And as BI gets used by more and more of the population in the business, not just by the the business analysts, then you have to distill down the knowledge of that system to something that is actionable by that individual. If Stephen and I both walk up to the counter at the exact same time looking for the upgrade on that flight, the system has to determine who is that most valuable customer and give the attendant somebody some answer that they can act on immediately in a real time kind of situation and not just just provide a tool that they can do the analysis for themselves. So I think we’re seeing BI becoming much more of an ability to act inside of a process, as part of a process, and be able to deliver actionable insight in real time.
Sanjog Aul [00:21:24]:
So you basically are saying that. Go ahead.
Stephen Brobst [00:21:27]:
I think that’s when BI becomes truly pervasive, is when it’s embedded in the process as you described, Don, because when it becomes invisible, then you can really get it to everybody in the organization who needs it, not just someone who knows how to work up the itool and those kinds of things. It becomes really truly pervasive at that point.
Sanjog Aul [00:21:46]:
Now when you do this, do you actually look at these actionable steps, etc, as something which is going to automatically come from the crunching? Or will this be something you will wrap around a BI with a decision support and or export system which of course leverages the slicing that you have done.
Stephen Brobst [00:22:07]:
I think that there are very frequently rule based systems involved in helping drive decisions to automate things. You can call those expert systems if you want, of sorts in that direction but the key point is that you don’t want to have to require IT professionals in order to be involved in sort of creating the business rules. You want IT professionals to create the infrastructure so that the business can be successful in owning the business rules and that you can evolve the business rules without requiring six month reprogramming efforts in order to do that.
Don Campbell [00:22:54]:
And BI started out as a report. There’s report authoring capabilities and report generation capabilities and the ability to do maybe some slicing and dicing multidimensionally or something. And it’s matured a lot since then. And producing scorecards and dashboards and all sorts of application faces on top of that intelligence and where the BI system stops and some other kind of expert system or process or application begins somewhat blurred and maybe the distinction is unnecessary at this point.
Sanjog Aul [00:23:33]:
Now let’s take an example of say two mature retail companies and both of them have of course a ton of data that is getting generated and both of them have some BI capabilities and they are slicing and dicing data and looking into it and thinking and or applying some of those insights that they are getting. How big of a quantum leap, if you will, would be achieved by one over the other? Because the difference in the way they implement would not be as significantly. There would not be as much delta, if you will, in the way they both operate and or apply bi. So is there a magic in BI which will give a quantum leap or we are just saying that then you have to find another way of doing it, not Just bank on BI to get you competitive advantage.
Don Campbell [00:24:27]:
Well, I don’t think you buy a BI tool and you stick it into your IT infrastructure and all of a sudden you get a certain amount of impact from that and everybody gets the same amount of impact. I think it is now still a tool and how that organization leverages that tool is how they receive their value from it. A lot of BI implementations are require cultural change, they require operational change for an organization to, you know, to really do that transformation to an information kind of LED transformation for their business and as they put that into the systems of the business, as they allow more and more users to get access to IT and to do more and more valuable work with the BI and get more and more insights out of it, it will transform their business more than maybe perhaps another one as they take advantage of not only the history, but more the real time and then they go into the future and start leveraging it for predictive analytics, which is another big trend as they bring in unstructured capabilities as well as structured capabilities. So I think it’s a tool and the people that get the most value out of it are the ones that put the most effort into the implementation.
Stephen Brobst [00:25:45]:
Yeah, I agree 100% with Don on this. There’s no silver bullet out there. You can have the best BI tool in the world, the best data warehouse infrastructure in the world, but if you don’t do the work on the business process side, if you don’t drive the adoption of the knowledge workers, if you don’t have the knowledge workers involved in how that information is going to be used to change the way decisions get made, it doesn’t matter. A successful organization will invest as much in the sort of business side in terms of training and process design as they do in the technology infrastructure, if not more.
Sanjog Aul [00:26:25]:
Now so if you were to go back and talk to these people and as like various companies who are saying that BI is going to be, I mean, they are of course not assuming that it’s going to be silver bullet. But if they start investing way too much, then there would be a lot of diminishing returns that would kick in even with respect to bi, because it’s only so much it could do, right?
Stephen Brobst [00:26:47]:
Well, yeah, you have to invest in sort of proportional way. There’s this concept called the synchronization index that says I need to invest, yes, in the technology and the BI tools and the data infrastructure and so on, but I also need to invest in my organizational capabilities, my skill sets, my business processes and so on. And I also need to invest in the customer interfaces for how I’m going to translate that knowledge, that business intelligence, into actions that are going to benefit my customers. And if any one of these three dimensions are underinvested, then you’re not going to get the full value from the investment that you’ve made. So over investing in one area at the expense of another is in the end going to hurt you.
Don Campbell [00:27:36]:
And nowadays the opportunity to invest is maybe less than it has been at other times and it’s a big prioritization issue and you have to understand where to get the big bang for the buck and invest in those systems and those processes first before you feed the other ones. And we consistently see that customers are investing in that BI layer in order to provide that valuable insight. But as Stephen says, it doesn’t stop at the IT department and it flows into all the operations of the business and it really needs to be fed right through the value chain in order to realize that end value or it’s just going to be a tool that’s implemented and not supported downstream.
Stephen Brobst [00:28:26]:
In fact, I would suggest that as your data warehouse and BI infrastructure becomes more and more successful, you will put less investment in the infrastructure side and more investment in the organizational chain side because as the data warehouse matures, you’ll find that you have most of the data that you need and you’re reusing the data rather than going back and getting more for each new analytic capability that you want to deliver to the organization. So if you’ve done a good job and centralizing the data and creating the capability for reuse of the data, I can deliver new analytic capability without the same amount of infrastructure work that I had to do upfront and that gives me the ability to invest more on the organizational side and on the BI capability side as opposed to the sort of low level infrastructure side.
Sanjog Aul [00:29:21]:
Let’s take a quick break listeners. We’ll be right back after these messages and and let’s see if we can take an inventory of the typical low hanging fruits when it comes to leveraging BI to get some competitive advantage and then what are some of those unique examples where people really went out of their scope, if you will, and then saw deep down somewhere opportunities where BI actually was successful. So some of those examples could possibly help. Please stay tuned. We’ll be right back and share those.
Sanjog Aul [00:32:42]:
Welcome back. So we have to take some inventory, get some examples of suppose an organization wants to use BI. What are the low hanging fruits where they can hit first and get the most and then where they could potentially look under the hood and see deep down where the opportunities may be. And some examples would definitely so Don, you want to take a shot?
Don Campbell [00:33:04]:
Well, you know, the low hanging fruit. You really have to look at a pain point in the business and that could be in a departmental implementation. While you may have big plans for organizational change that’s fine, and those can be supported you need to get some quick wins, you need to start where that pain point is, get some internal champions and that’s where a business intelligence competency center can play a big role as well too, as they help to navigate that organizational change that’s necessary to implement BI and you know, for some, it’s a cultural issue when you’re trying to share information across departments, and some departments aren’t willing to do that right off the bat and so the internal champions can help share that value, share those quick wins, start getting the rest of the organization more excited about it, get some senior management support, etc and it can really build from there, so it doesn’t have to be eat the whole elephant in one bite kind of a problem.
Don Campbell [00:34:14]:
You can really start small and grow the success.
Stephen Brobst [00:34:19]:
As Don mentioned, there’s, you know, the opportunities, the points of pain are going to be different for different organizations, but the ones that really get a big bang are the ones that use business intelligence to change the way they do business in some significant way. A good example that comes to mind, in fact, a joint teradata cognos customer is Hera’s Entertainment. They’re a large casino hotel chain, and back in the day, the casinos were competing by building bigger and bigger casinos, Egyptian pyramids and Roman cities, etc. and Harris made the decision that this was not an effective way to compete by sinking more and more capital into bigger and bigger casinos. So they decided to invest in their data assets, and they built the most successful loyalty program in the industry, probably one of the most successful loyalty programs across any industry. And they’re able to understand detailed customer behaviors using analytics in very, aggressive and interesting ways. So they create this loyalty program. They analytically predict maximum threshold of loss for each one of their customers
Stephen Brobst [00:35:42]:
and they have created processes whereby if you’re going to one of their properties, you want to have an entertainment experience, but you don’t want to lose your shirt in the process and each person has a different threshold of loss beyond which it becomes painful. Hara’s will calculate that threshold of loss on an individual customer basis. And then they have processes in place in order to manage that experience so that you don’t lose more than what would be comfortable. And therefore they keep you as a customer for life. So a big part of this is analytics, but also a big part of it is the business process change. For how do you manage that customer experience and keep the customers happy? It’s a very interesting case study of how an organization used analytics to not follow the pack and be like all the rest, but invest in a different way. On this basis they were able in the first year after they delivered this analytic capability, they increased the frequency of visit for their regular players by more than 50% and they increased cross property revenue by 72%
Stephen Brobst [00:36:54]:
this was huge return on investment from their data warehouse and it wasn’t a huge data warehouse, right? I mean they implemented in phases, just like Don said, one byte at a time. Their initial data warehouse was 200 gigabytes, practically nothing by today’s standards and of course they’ve grown it as successful help fund more sophisticated initiatives but they changed the way they do business and they changed the business model for an industry really.
Don Campbell [00:37:20]:
They would go as far as to map out speacially what the gaming floor looked like and what the successes were at certain machines, even in the slots and the tour that people took as they went from the Czechia into the restaurant, etc and they would visualize even all the return on investment they were getting and adjust the payouts, adjust where they flow the traffic and really quite an extensive use of analytics.
Sanjog Aul [00:37:55]:
Now would you say that this is, I would say an epitome of what a company can do with BI? The companies who do not at least know about such beautiful opportunities that might exist for them to take it to the next level. What is supposed to be their approach to identifying these cool ways of using BI? I mean people sometimes could be lost because they don’t know where to look. Is there something.
Stephen Brobst [00:38:22]:
I think there’s a couple of different approaches that are successful. One is first of all it has to be business driven. So don’t, have a solution, go looking for a problem to solve. There is a process that we use called business and information discovery. Where we look for where are the opportunities for making better decisions? Forget about technology, forget about implementation. Where are those opportunities for better decision making in an organization and tie those to specific business impact models.
Stephen Brobst [00:38:54]:
The business, they may not express it.
Stephen Brobst [00:38:55]:
In the terms of business intelligence that they will know. Good business leaders will know where opportunities are for improvement and then it’s a BI professional’s job to translate that into how will knowledge, how will better business intelligence help translate this opportunity into something that makes money for the organization? The second thing that I would point out is organizations should benchmark themselves against the capability of other organizations, both in their industry and outside their industry. There’s a customer that we have, Commonwealth bank of Australia every year they go on, they call it a field trip and they take two weeks out of their busy schedule ao this is a big commitment. It’s the business people and IT people together and they will tour what they consider best practice organizations around the world. They’re in Australia, so they sometimes go a bit isolated way down there
Stephen Brobst [00:39:55]:
and so they’ll take their thought leaders and they’ll bring them to organizations like Walmart or Harrods that we just mentioned or maybe Tesco’s in the UK and they’re pointing out they might be financial organizations, but they also might be non financial. There’s a lot to learn both within your industry and outside your industry about what best practices are and what the capabilities are for business intelligence within an organization.
Don Campbell [00:40:27]:
One of the things to do is to talk to the users. One out of three business leaders will tell you that they don’t have the information that they need to make critical business decisions and so finding out what kind of decisions they really need and, and putting a BI system in place that will answer those questions is critical. More than half of them don’t have information across departments that they need in order to do their job because they only have access to interdepartmental kind of information as well and so really asking those key questions, asking what is the cost of not making a decision? Because that can be an extremely costly thing for a business as well too when you are stalled because you don’t have the information you need to make a decision and then putting those kind of systems in place to enable those decisions going forward. So you know the problems are out there and people in the business tend to know what those problems are and really we should be asking them those kinds of questions and then putting together the system to be able to make them successful.
Sanjog Aul [00:41:32]:
Now in this type of world that we live in where we are so connected to each other, that also means that we’re connected to our partners along the value chain and they may all do their own respective BI and try to get some competitive advantage out of it. Have we looked at a model or is this something that is actually being practiced today where we open up our respective sensitive data and look at holistic being holistic across the value chain.
Don Campbell [00:42:02]:
One of the examples that we’ve seen is actually some of our customers making scorecards available of their suppliers and making them available to their suppliers so that they can see how their competition is doing in supporting our customer. So it becomes a little bit of competitive situation where they can see if they’re actually trailing the abilities to support this customer compared to some of their colleagues in the industry, then they need to pick up their game. So the sharing of information can actually be valuable from, from a business standpoint, not just a scary thing and letting your information outside of your firewall, but it can actually help to create that competitive ecosystem necessary in your supply chain in order to improve the situation for you.
Stephen Brobst [00:43:01]:
One of the measures of maturity that I use for a data warehouse is if a data warehouse is truly mature, then you should have more users of the data warehouse who are not your employees than those who are. There’s an interesting study done by IDC who was published I think the beginning of last year that identified that 20% of the global 5000 organizations across industry, across geography provided business intelligence capabilities to their suppliers, distributors, customers or government agencies that weren’t part of their organization. So in answer to your question, people are absolutely doing this and they are typically the leaders. Walmart was an early player in this. Their retail link application allows their suppliers like Procter and Gamble and Coke and Pepsi and so on to directly access the Walmart data warehouse in order to make decisions and they put into place a whole process called vendor managed inventory, where if I’m Proctor and Gamble, I own my
Stephen Brobst [00:44:14]:
Store shelf at a Walmart store and
Stephen Brobst [00:44:18]:
it’S my responsibility to do the analytics on a per store product basis to make sure the right inventory is on that store shelf at the right time. What they’re doing is thinking about the organization in a different way, integrating across the value chain and sharing information in order to do that. So with this vendor managed inventory, Walmart is essentially a consignment store, Procter and Gamble owns the inventory on the Walmart shelf until that inventory is sold so it is called pay for scan. The goal is not just to shift carrying cost to the supplier. The goal is to eliminate the carrying cost by having a tighter integration by sharing data on a selective basis and providing business intelligence capability. 3M is another company that does this.
Stephen Brobst [00:45:09]:
they share data from their data warehouse with their distributors but in extreme cases, you might even allow your customers to have access to decision support off of your data warehouse. A good example of this is Wells Fargo Bank. Wells Fargo Bank, if you go to their Internet banking site and you go to the my spending tab and you access for your own personal finances, you can see how
Stephen Brobst [00:45:36]:
you are spending your money.
Stephen Brobst [00:45:37]:
Are you spending money on restaurants or airfares or whatever, how you’re doing related to your savings goals and so on. You are accessing Wells Fargo’s Teradata data warehouse.
Stephen Brobst [00:45:48]:
When you do that, of course you
Stephen Brobst [00:45:49]:
don’t know that, right? But that’s what’s going on and of course I can’t see someone else’s data, so all the security and privacy controls have to be in place but what it does is it delivers this decisioning capability further out into the value chain of an organization and it creates competitive advantage.
Sanjog Aul [00:46:12]:
Let’s take a quick break, listeners. We’ll be right back after these messages. And let’s see if truly BI, especially going in the competitive BI segment, would this be an R and D function, something which is going to always require for us to be looking a needle in a haystack or try to look for the unknown or unusual? And then if you wanted to do that, if that’s what we want to achieve from BI would it be a good idea to even enforce standards or have some sort of a centrist of excellence where everything is standardized and there’s some governance? So should there be a governance on BI for it to be also delivering competitive advantage at the same time? Please stay tuned. We’ll be right back and explore.
Sanjog Aul [00:49:09]:
Welcome back. So BI, we got to go crazy with it for us to be able to get the most out of it. I guess you have to look down under somewhere to find out those competitive advantage related opportunities. If you’re doing that, you’re essentially doing some R and D and if you are doing R and D, you don’t want to be stopped and you don’t want to be standardized. So is this a catch 22 that if you don’t put BI under some sort of a center of excellence and some standards are not enforced, you could keep going on a wild goose chase But at the same time, if you just stifle it so much that it doesn’t allow you to look under those areas where nobody else has gone before, then you don’t see those areas which nobody else has that especially your competitors have not seen and you don’t get the competitive advantage. So how do you eat the cake and have it too?
Don Campbell [00:49:55]:
Well, the technology world continues to change and I think you have to implement an infrastructure, as Stephen was saying earlier, spend some time and put a system into place that’s flexible and that’s agile and that can move forward with your business needs and as technology continues to grow and change, we get new opportunities to get new insight faster and more valuably then those should be able to be plugged into an infrastructure that is already in place. So creating multiple systems and then tying them all together is a very difficult thing. Laying down a platform that can be integrated on top of I think is really the way to do it and then you can take that area of insight that you really want to drive forward for your business. You want to enable your mobile workforce. You put a mobile system on top of that, you want to put some predictive analytics in there, then you put that on top of that infrastructure, etc. Once, once you’ve got that infrastructure in place, then it enables you to do sort of that area of mining that is the most important to your business.
Stephen Brobst [00:51:05]:
I think you actually have to do both, right? And it’s a very tricky balance, right?
Stephen Brobst [00:51:11]:
You need standardization of data definitions, you need to have a sort of standard for data quality. You need those things in place in order to get full value from your Information assets, but those same sort of.
Stephen Brobst [00:51:26]:
There’s a lot of effort to do that and those can become a barrier to innovation
Stephen Brobst [00:51:31]:
and so I think that you have.
Stephen Brobst [00:51:33]:
To decide at what point do you do that.
Stephen Brobst [00:51:35]:
And a good example of an organization.
Stephen Brobst [00:51:37]:
I think is really innovative is ebay. They’re a very, data driven company
Stephen Brobst [00:51:43]:
and one of the things that they do is they’ve created this agile data warehousing methodology where they have a certain
Stephen Brobst [00:51:52]:
core of data which is standardized and it’s data quality certified and they’ve got data modeled into an enterprise infrastructure and all this kind of stuff but they also have a methodology that allows a person, you can call them an R and D person if you want, but it’s very decentralized across different functional areas within the organization and they can bring any data they want into the data warehouse, the business, if users can self provision data. So they provided some tools to make that relatively easy. They self provision data into the data warehouse and they can use this, I’ll call it experimental or R and D data, together with the core of production certified reusable data in the data warehouse and they explore for new ideas and sometimes they’re successful and sometimes they’re not but the failure is okay as long as you learn from the failure and the goal is to fail fast, learn quickly and get to the successes more quickly.
Stephen Brobst [00:52:55]:
And by not imposing a lot of the standardization and overhead and processes and documentation too early on in the R and D process that allows them to be much more agile and this has been a hugely successful initiative at ebay and this technique has started to be adopted in a pretty widespread way across our customer base. The key point here to recognize is that it really is. There’s a lot of experimentation and as you said, it’s an R and D effort. I don’t believe you want to centralize that R and D effort. I think you want to have this business intelligence center of expertise or competency center, build the infrastructure and tools to allow decentralization and widespread exploitation of data
Stephen Brobst [00:53:49]:
using these agile techniques.
Stephen Brobst [00:53:51]:
And then you want to have some level of centralization of the capabilities to promote the data from these sandboxing environments into the reusable data in a standardized, productionalized way. So you got to make careful trade offs related to how much standardization to do when to have a good trade
Stephen Brobst [00:54:17]:
off between agility and reusability of the data within the organization.
Sanjog Aul [00:54:22]:
On behalf of the show and our listeners, I’d really like to thank you, Don and Stephen for sharing your thoughts about how organizations can take a second look at their culture, can take a second look at how they are leveraging data for competitive advantages.
Don Campbell [00:54:38]:
You’re welcome, Sanjog.
Stephen Brobst [00:54:39]:
Thank you very much for having us on your show.
Sanjog Aul [00:54:42]:
Thank you so much again and listeners, if you have any questions or views, please send us the email to viewsciotalkradio.com that is viewsciotalkradio.com thank you again for listening to CIO Talk Radio. This is Sanjog Aul your talk show host till next week take care and god bless.
Speaker B [00:55:04]:
Thank you for tuning in to CIO Talk Radio. To learn more about the show, please visit www.ciotalkradio.com. Join Sanjog Aul next Wednesday at 9am Central, 7am Pacific for another hour of CIO Talk Radio.
Download Podcast
Apple Podcast, Google Podcast, Spotify, Pandora, iHeartRadio, SoundCloud, TuneIn, and Stitcher. Find other syndication channels here or search CIO Talk Network podcast on any other app.
Explore More
