The Biggest Picture: Knowing Your Customer Across Multiple Platforms

Eric Kavanagh: Alright, ladies and gentlemen, summertime approaches, it’s getting hot in here. Why’s that? Because it’s time for Hot Technologies. Yes indeed, my name is Eric Kavanagh. I will be your moderator for the show that is designed to – we should talk about what’s hot, what’s happening, what is the cool stuff out there in the marketplace. This is our partnership with Techopedia. We love these guys. We’ve been working with them for several years now. They have a fantastic site. If you want to know anything in the tech world, what its definition might be, go to techopedia.com. And today we’re talking about MDM, master data management. The exact title is “The Biggest Picture: Knowing Your Customer Across Multiple Platforms.” And this game is changing, folks, I can tell you right now.

So there’s a spot about yours truly, hit me up on Twitter @eric_kavanagh. I try to reply to anyone who replies to me. So the year is hot. It’s sure hot for MDM. And I tell you what, it’s hot, not just hot for the big enterprises but also for the small to medium size businesses who, guess what, have lots of different systems. CRM systems, email marketing systems, ERP systems, web analytic systems, eBusiness suites, etc. There are lots of different points of access to information about customers and the better a job that companies can do of weaving that all together, the better they’re going to be able to serve the customer, not tick off the customer and keep those customers around. Keep them buying some more stuff.

I’ve actually tracked MDM personally since about 2003 which is around when the term was really coined. Frankly there was a bank, Chase Bank in fact, I think it was Bank One way back then, and one of my good friends now, a guy named Joe Northern worked with a company called Razza Solutions, and they had what became the DRM tool of Oracle. So they actually rolled up accounts and did hierarchy management for the bank way back then and that’s some of the early days of master data management.

So these days we talk about both analytical and operational MDMs. We’re going to talk a lot about that stuff today and really help you understand how you can leverage this technology to get that complete view of your customer, to understand who they are and to make sure that you can take care of their needs in what is a frankly very competitive environment worldwide. We’re seeing that all over the place.

So, rock star cast of characters here: Dez Blanchfield, Robin Bloor, John Evans, Diana Collins. Calling in from four different locations all around the planet. We start off with Dez Blanchfield and with that I’m going to hand the keys to you, Dez, and I’ll start tweeting. Take it away.

Dez Blanchfield: Thank you Eric. I just had to remind myself to get off mute. I apologize for that. Thank you for the opportunity to present on this. So, I’m going to come at this from the point of view of a real-world example of an organizational challenge to deal with what I’ve referred to as one of the biggest disruptions to organizations they’re going to see for some time. We’ve seen a number of challenges. The GFC hit companies had to deal with it. We get regular changes in law around privacy we have to deal with.

One of the things that I think organizations are being caught out with that they didn’t see coming was the impact of this whole celebrity experience issue. And essentially people running around with mobile phones wanting instant gratification in some ways. But, instant gratification in a good way, not a petulant childish way. Just a realization that they’re the customer, they’re paying the money and they should be getting the value for it. And so there’s been this coinage of customer centricity or becoming a customer-centric organization. So I’m going to quickly walk through what that means and lead into a slightly more technical part of our discussions soon.

I’m just going to put it out there and say that firstly, being a customer-centric organization comes down to one simple thing: You need a complete view of your customer and your customer data. You may have different systems. You may have plenty different products. You might have fifty different departments in your organization, but no matter where you are in the organization, no matter what your job function is, you should be able to get a complete view of all your customers, or the customers that are in context to what your job function is. And every part of the data set that you have, or all parts of the data sets you have that tell you what the state of the nation for that customer is.

I put this line that, a complete picture of customers across all of your systems isn’t just a nicety! Nowadays it’s a necessity. And the first time that you get caught out in a scenario where you’re dealing with anything to do with a customer, particularly if it’s live, on the phone, or in a webchat or in person which is even more frightening, and you can’t tell them everything that you should know about them, it becomes very obvious and it’s a very unfortunate situation to be in.

I’m going to start out with a very quick anecdote around real-world scenarios. This is a photo of a whiteboard and this is less than five days old. This is an actual scenario in a whiteboard in a room recently, a couple of days ago, talking about the very topic of how we go from being a very large organization with something like ninety different parts of our business. It’s an Asian bank, they have ninety different business units. They do everything from society loans and peer-to-peer and micro loans all the way through to financing putting satellites in space. So they’re a monster. They have tens of millions of clients. I think they’ve got just under fifty million clients. And they are faced with this typical challenge of how do we approach not just master data management but customer data in particular and a single-unit client.

And as we mapped it out the thing that jumped off us from this whiteboard was they had not just a problem, they had a nightmare because none of their systems talked to each other. I could go into any part of the bank or any part of the business and ask for a loan, it might be a car loan, a house loan, a small business loan, and they couldn’t tell themselves anything, or they couldn’t discover anything about any other relationship I had with the bank. And it was absolutely scaring the daylights out of them because they realized that the banks down the road can already do this and they’re potentially 12, 15 years behind the eight ball. And it comes down to these key value propositions that clients are just looking for which is just a consistent view of me as a customer, and you need to figure out how you’re going to deliver that. Particularly now that I’m dealing with you on the web, more likely to be the case via app these days.

It’s come down to this key thing of “it’s all about me, the customer.” And so when we map out what the customer-centric culture looks like, it’s about incorporating everything we’ve got from core systems that capture things like your first name, last name and other details when you fill in a form or fill it up online or come to us at a counter somewhere at an outlet, and we get to know you initially through the entire journey of us delivering the products themselves or service to you. And mapping that out from top to bottom. Continually refining the data and the data models we use to understand that. Aligning how those technologies and processes in the business, work flows, continually tighten up our view of you. The ongoing engagement that we have with you. How we continually focus around you the client and how we communicate with you. If I’m selling you three services I don’t want to be sending you three different pieces of paper every month or three statements or bills, and so on.

The customer-centric story is getting some real traction now and organizations are seeing the value of it. It’s still a real challenge in that it’s, “Okay, well I’ve got ten different systems and they don’t talk to each other. I don’t have a tool or a system or a platform to pull it all together.” And invariably people end up in a room doing whiteboard sessions like the one I just showed you. But it all comes down to one core thing in the left-hand corner there of a transformation. And transformation from culture of the organization, and the people, and the staff and the operational model, all the way down to the technology stack that supports them. So there’s a fairly common checklist that organizations go through to get to this point where they even understand the challenge of what it means to be customer centric and the need to build a system and get access to tools that can help them do this.

It’s things like mapping the customer journey through the full lifecycle and the experience that they have with you as an organization. Refining your operating models and how you’re organizing yourselves to be focused on the customer and the value proposition you’ve provided the customer. And then of course aligning your technologies and your technology stacks and processes around them to make sure that you are actually continually driving further engagement, and better and tighter engagement with your clients. And the actual engagement process itself from the executives downwards.

If you haven’t changed your view of the world from the top of your food chain, from the boardroom downwards, then there’s little chance that your depo level or your day-to-day finance staff are going to change their behavior. You have to lead from the top. You have to continually refresh and redefine and redevelop how you address targeting the actual focus on the client. So, how are you bringing about not just a cultural shift at the top end but behavioral change at the bottom end of the organization, and the tools that you’re making available to do that?

It’s one thing to say you’re a customer-centric organization and you want people to behave one way, but you haven’t given them the means and the tools and the capability of doing that, you’re not going to get a behavioral shift because people will just keep falling back into the habits they’ve known prior to thinking they were customer-centric organizations. And then the overall integration of the disparate parts of the organization and the culture that lived inside that and obviously underpinned by the tools and the platform.

So how do you take these disparate business units or businesses or parts of your organization and have them behave differently from a cultural point of view and downwards? Well, you provide them the appropriate tools and ways and means to get that complete and single view of the client and the client experience. And then how do you put some KPIs and measure it against that and track it and put some metrics against those and measure those KPIs and provide value to that? Business value to yourselves and obviously value in some form in the value chain to the customer and keep them coming back. And then incorporate all of the communication you’ve got with your clients from feedback and real time or iteratively processed so that your behavior and your cultural shift hopefully gets captured in some sort of feedback cycle and feedback loop and you can figure out whether you’re actually hitting the mark or not.

We get to the scenario where you know eventually organizations are going to find themselves effectively drowning in disparate data and we’ve seen some sorts of that here, some internal, some external. Historically we’ve had a customer relationship management platforms and advertising platforms and marketing platforms. We’ve had all kinds of different systems that run independently and then hopefully they do talk to each other in some form. We’ve had in the last couple of weeks an explosion of interactions with you now, so we talk to you via social media, we talk to you via our website, we’re getting emails from you.

Our IVR systems that talk to you via the phone are now having to map that data back and tell us how you dealt with our phone system and interact with our databases and if you’d been on a phone call with us, all that’s got to be captured in real time and we need to be able to make sure that we can get a common view of it, which hopefully is that common data management platform in the center of that diagram there.

There’s the phrase that’s been coined recently that’s of a “celebrity customer experience.” Well what does that really mean? It’s not that we think our end users or consumers are badly behaved celebrities and that they feel different in any way. What it means is that we’ve woken up to the fact that we should be treating every one of our customers as a celebrity. They should be getting the VIP treatment from the very moment we meet them through the entire lifecycle of our having the pleasure of having them as customers.

And so the question that I get asked on a regular basis – bringing this back to a slightly more anecdotal real story of a client – is how do we enable our organization to deliver on that increasing demand for a celebrity customer experience? Because the thing we’re seeing now is one of the greatest disruptions to organizations is that requirement to deliver on that promise to clients. To give them the celebrity customer experience. Organizations, from my experience, and certainly around the world that I’m seeing, are being disrupted without realizing it with the shift from other influences that they might have already known about or seen coming to their actual customers. Their customers are disrupting them and disrupting them in a very serious way. And then if you can’t provide this celebrity experience and provide the tools and the ways and the means for your organization to get that single view of the client then you’re going to miss by a mile, a country mile at least, the capability and the capacity to deliver on that promise.

There’s some key points I’m going to throw out here, and then hand over to Robin to get into slightly more technical details, that I recommend that all organizations think very hard and fast about if they’re even remotely close to this delivering on the promise to their staff and their organization to become a customer-centric entity. And that is focus on the basic components and create a single customer view. That sounds very simple, but what does it mean? Well it means making sure you’ve got the right data from the right data sources all the time, and at the right time. Making sure that the data is available in the right place all the time. Not just some of the time.

And it has to be tightly integrated. And it has to be natively built into your platform. It can’t just be something you think you do. A single marketing campaign. Every time you look at a customer you need to be able to get this all the time. It needs to be available to all the right people all the time. So I don’t want to be running around the hallways looking for tribal knowledge. I need to be able to get this at a moment’s notice just by getting to one tool. And you need to provide it in the right platform with the right tool. So it has to be built into the existing systems you’re already using.

Your CRM needs to be able to see everything from when I’m visiting you from my mobile app, from the website, from talking to your IVR, interactive voice recording, to go through your phone help desk myself as a self-service. Or if I push star-nine and I get to the human being then I ask the slightly more challenging question that the IVR isn’t programmed to deal with. If I tweet something happy, if I’ve written an article on LinkedIn. These all need to eventually feed back into the CRM so that if I’m managing anything to do with the customer I’m able to see that. We need to make it the default and not the exception.

It’s still very much the exception that people want to run a campaign, they want to run a sales and marketing effort, or they’re looking to solve some problem or deal with a pricing issue. We run a one-off campaign and try and get a single view of a particular segment of our client and start running reports and printing things off and handing them around in bound printed copy format. That’s an exception. That needs to be the default. Your systems have to, all the time, provide this single view of the client. And in any way we come at it – whether it’s a sales and marketing, or just an operational, or manufacturing, or logistics, or whatever it might be, point of view – the reality is you’re going to have to do all that before you can see a solid ROI on your investment in this transition to becoming a customer-centric organization. You’re going to get some quick wins. There are definitely going to be quick wins. So there’s some good news on that front. But the reality is that until you complete a transition to becoming a full single view of your client customer-centric organization that ROI isn’t going to jump off the screen at you. And it’s a fun journey. It’s a worthwhile journey. And it’s all underpinned by having the right tools, the right platforms, and making it available to your organization at the earliest possible time, in a sensible, technically and commercially viable form. With that in mind I’m going to hand over to Robin. Robin?

Robin Bloor: Thank you, Dez. I had to do the same as you, I had to un-mute myself. Okay, I was going to approach this more from a conceptual point of view than the kind of practical scenario that Dez went through. We’re really talking about a very specific set of activities within an organization when we get into the area of MDM and of course customer is the big deal. The entity identity of customer is much more difficult to get at for a whole host of reasons than anything else. It’s likely to be the most important entity. There are some businesses where they may only have one customer and they may have all the information that they could get about that customer. Very rare. Mostly organizations have multiple customers and the customers have multiple facets. And the data is pretty much spread all over the place. I’ve been working with this idea fairly recently, the idea of a data pyramid. That there’s a distinct difference between data and information and knowledge, and actually understanding. But data, information and knowledge can live in computers. Data at the lowest level is just signals and measurements. And information you can get your hands on which is what—

Eric Kavanagh: Your audio is starting to fade out a bit, Robin. Just so you know.

Robin Bloor: Okay I’ll move the microphone. How about that?

Eric Kavanagh: There you go. That sounds much better. There you go.

Robin Bloor: Yeah, so data is made up primarily of signals, measurements, recordings and things like that. It has no specific context. It becomes information by giving it that context. Linking data together. Structuring the data. Creating visualizations, glossaries, schemas. Anything you want to create around it. It gets transferred into knowledge when in one way or another you can actually start to predict the behavior of a given entity and also implement policies and rules for handling it. Understanding lives entirely in human beings. And that’s part of the problem. When you actually look at the fragmentation that exists in terms of the customer situation you often discover that, well really sales has one view of customer, marketing has another. Sales support or actually just customer maintenance has a different view. There may be many touch points that a customer has with an organization. And none of that’s integrated into properly structured information or a lot of it isn’t integrated.

And then we have the problem that has started to become much more prevalent in the past few years of, you can gather external data on people and that’s very useful but you actually have to integrate that for that to have any real value. So in the refinement of data the big difficulties arise from fragmentation. That data’s coming from different places and it’s not well structured. And the fact that there tends to be an incessant supply of new data and this is nearly always the case when it comes to customer. And every entity is a moving target. We didn’t care, perhaps three or four years ago, about the social media profile of customers, but we care about it now. We care about it because it can be damaging to an organization or boosting to an organization, depending on what’s going on out there.

If you actually have the idea, if you sat down and did an exercise and tried to work out what was the things that you were interested about customer five years ago? And you do it again and you discover that stuff has been added. And stuff may have been taken away. I mean nobody cares anymore for instance what fax number people actually have. Some people used to have fax numbers on their business cards. But no one cares anymore because fax died. So, it’s a moving target. When you look at data modeling and MDM the first thing – well actually I have to say about this, is that this is part of data governance, if you’re not doing this then there’s a problem in the way that you’re governing data. Because if you’re not actually doing data modeling and MDM, then in one way or another you don’t actually have a very good top-down view of any given entity in actual fact.

But I’ve listed here data governance. I’ve listed lineage, data usage, quality, security, service management, recovery. You could add lifecycle and so on. There’s an awful lot to data governance and data modeling and MDM is a fundamental and perhaps central part of that. Change comes from the top down in the sense that you realize that change is occurring because people realize that it is occurring. And therefore one might think in terms of this whole stack from files and databases through data elements to beta data and business definitions.

You might think in terms of actually having to, in one way or another, to manage the whole stack and keep the whole stack up to date because knowing something at a business definition level doesn’t actually mean that you’re catching the data at file and database level. It’s a very broad picture and until you actually think about it you don’t realize how broad it is. The modeling and MDM, if you actually look, the whole big data trend isn’t simply about that – there’s a lot more data. It’s about that there’s a lot more data from a lot more sources, giving you a lot more perspectives on any given entity that you’re actually collecting information about. And the more complex that is, the more you need a model, the less easy it is to comprehend. Just by looking at let’s say a database schema what’s going on when data is actually coming from 10, 20, 30 sources.

In theory you can say MDM gives you a view of the data universe but in practice it’s actually part of it. And we actually just discussed if you’re looking at the business meaning of data then that, the information about the meaning of data, is actually part of the data universe that you’re looking at. Modeling is from top down and from bottom up. That is that you can look at things from a business perspective but you can also look at things from the perspective of what we’ve got. And you build in both directions. And this is not, and never can be, a project. To start it off is a project. It’s an ongoing activity. You may kick it off as a project because you don’t have anything coherent in place, but once you’ve kicked it off it should be an ongoing activity. And anything that’s done within the sphere of data, the MDM team if you like, should know about it.

The customer challenges, just look at focusing in on the customer entity. There’s way more data available now about the customer from far more sources than for any other entity by far. And it just seems to increase all the time. It’s often inaccurate. If you’re gathering data from me, for instance. If you’re gathering data about me you will realize that I have different identities which is just whether I use middle initials or not when I go to various websites. And I do that often just to discover where I’m going to get spam from a given identity. But a lot of people do that. And then people make accidental errors. And then information’s out of date.

I went to one of these data resources that claims to be able to give you lots of information about any given individual, and did the obvious thing and asked questions about myself. And half of the information they gave me was actually out of date. And some of it was wrong anyway. And you look at that and you think, if you’re going to in one way or another gather data from other sources, then there’s a huge element of cleansing the data and being able to identify whether it’s the data you’ve got. As individuals we have no unique identifier. Name and mobile phone number will probably get you close to most people, but not everybody has a mobile phone number. And it’s different in different cultures as well. And then there’s the nature of data in terms of analytics.

I’m not going to go into this in any depth, but data can be select. If you’ve got somebody’s Twitter data, there’s only a small population of individuals that actively put data on Twitter. And they’re select. They’re not randomly selected customers. They’re the ones that have decided that they want to be vociferous on Twitter. It’s notoriously difficult to get a 360-degree view of a customer. And that’s partly simply because of everybody’s technical history. It’s not unusual to discover there’s three or more customer databases, just as databases, never mind lots of other sources of information that you actually collect about the customer. And customer analytics, it’s worth saying that it’s a huge opportunity now. We used to do segmentation in churn but now it’s really, because there’s an awful lot of external data available on customers, you can do an awful lot of relationship graph analytics, which is really relatively new. You can use predictive analytics, you never knew before. You can gather fashion information and opinion information you could never gather before.

There’s a very good reason to review what you’re doing in respect of the customer and to think in terms of how you can best leverage the data you have. A practical view. The modeling of the customer entity is a necessary activity for the sake of accurate and useful BI and the refinement of knowledge. In other words, if you’ve got a reasonably large population of customers it’s not really an optional thing. You kind of have to do it. And I think that’s all I’ve got to say. Let’s pass the ball on.

Eric Kavanagh: Alright, so John, I believe you’re going next? Then Diana will do a demo. So with that, John Evans, take it away. And folks, don’t be shy, send your questions in at any time. We’ll be monitoring that for the Q&A. Take it away, John Evans.

John Evans: Alright. Thank you, Eric. And thank you Dez and Robin for that introduction and those comments. There was a lot of overlap between what you talked about there and what we’re going to talk about and show today, which is great. And that we would definitely agree that this notion of customer centricity is something that people are seeking to achieve and I think at the root of that we’d say that having good data, as good data as you can get about your customers, is the only way to have a prayer of achieving that. So what we want to do today is talk about customer-oriented master data management and share with everyone a little bit about how we approach that, solving that issue, and talk about a new offering that we’ve just introduced that’s designed to make it easy for companies of all sizes to deliver better customer data throughout their fragmented data landscape. So that landscape could look something like this.

We’ve got a variety of systems around the perimeter here, lots of fragmented applications, some of them are running in the cloud, some of them are running on premises. And within each of these, by definition, you’re going to have different ways of identifying customers and customer information. Different models of customer data with different attributes, different priorities and so forth. And even if you were an organization where you consider yourself to be, you know, an SAP shop or an Oracle shop, or you’re just running your business on SAP for example, or just on Oracle, or you’re using SalesForce, you may have multiple instances of those systems even within your own company. Maybe they’re deployed by a different location or by a region that’s set up for different reasons, different zones of the world, or you might have them set up differently by line of business. And even if you’ve got a single ERP, if you’ve done the customization across those, there’s going to be conflicts in the data.

Now the fragmentation we’re seeing is further compounded by the increase in adoption of cloud-based systems and best-of-breed applications. So while a really large, complex, convoluted environment like this used to be something that everyone thought, “Well that really only occurs in the really large companies,” because of this advent of cloud solutions and best-of-breed approach, that issue’s now becoming more prevalent even in smaller organizations. So it really runs a range from small enterprises all the way to large enterprises. Everybody’s suffering from the same problem with their customer data. And you can look at some of those problems I’ve listed here in the middle.

I kind of break them up into three types. There’s data-related problems where you’ve got duplicates, you’ve got invalid data, you have missing fields, you have inconsistent information, inconsistent hierarchies, and those things just tend to get worse as time goes on. Then you’ve got people-related challenges where people can’t access the data, they can’t answer the questions that they have, where they are seeking but they are unable to attain that 360-degree view that Robin was talking about.

And on the third area is process-related challenges where you’ve got data in multiple places and also people don’t know what changed and when because things are happening to the data all the time. So there’s no control or governance over how to keep that data clean. So as you’re attempting to deliver a more cohesive/cogent customer experience and engage in a dialogue with customers, it’s really difficult to achieve that when your own data about those individuals isn’t consistent and isn’t accurate.

Just as an aside I saw, I think it was last week or the week before, an article in “Information Management” that was talking about why personalized marketing still isn’t accurate and they listed nine reasons. First two reasons in their list, data quality is poor and the data isn’t integrated.

So what can you do about this? Well there’s a couple of ways you can try and approach this problem and think about in terms of what it’s going to cost your organization. You can either sort of attack that data when it’s born if you will or you can attack it once it’s infiltrated into your system, So here’s a picture from an organization that we’ve worked with that actually highlighted about thirty different places where data was stored in there, in their landscape.

So once that data’s been kind of released into the wild, into these dozens of systems it’s hard to find, it’s hard to maintain, it’s expensive to fix, if you think about going in and trying to fix it thirty different times in thirty different places. So one of the concepts we want to talk about is trying to be proactive and trying to fix things early in its life cycle as possible because when you do that it’s going to be easier to find, easier to control and less expensive to fix and maintain and that way you’re going to get better data as you work downstream in your applications.

So this is a concept we’ve been talking about called proactive MDM and the tagline we like to use is the concept of cleaning the rivers, not the lakes. So there’s three steps to that, first is to get clean, where you want to match and merge and cleanse and survive records as close to the source as possible to try and get a golden record so that you avoid polluting your downstream applications. This can be done by implemented controls over sources or even providing a place to centrally offer the data so that it’s consistent and accurate before you release it into the wild.

Enrichment is about adding value to the data as you go, including reference data and other information that’s not in your source operational system, so this might be hierarchies, it might be segmentations for example that aren’t inherently stored in those systems.

Then the third part is about staying clean and here’s where you want to make sure you’ve got processes in place and people identified to do the stewardship and to do the governance, have tools available to enable those processes and then proactively match and you cleanse your data on a periodic basis so that it doesn’t, so you avoid the decay that naturally is going to happen, for example when people change jobs or they change their residence or so forth.

So how do you get this? Well, there are a number of options that you could use to attack this problem. You could use a data quality tool, you could use a data integration tool to extract the information, you could use a work flow tool in order to portion work out to different people. You could use a governance tool to keep track of who’s doing what. You can actually string together all those different heritage tools and throw a lot of people at it.

But that’s all very expensive, it’s very resource intensive, it’s going to be slow to deploy and it’s going to be hard to manage and you might even want to start with your customer data but you’re also going to want to eventually manage your products, your list of products that those customers have, and the list of suppliers for those products, and a chart of accounts that you’re using across your business to keep track of what’s going on, manage your employees who service those customers and so forth. So now you’re talking about multiple domains, suppliers, products, chart of accounts, employees and so on to try and deliver a 360-degree view of your entire business.

So ideally what we think you want to achieve is one solution to integrate, match and cleanse your customer master data, one solution so you can manage the stewardship and the governance and one tool that you can use to manage every data domain as you start with customer and move on. So that is the objective behind a new offering that we’ve just announced called Magnitude ONE. Magnitude ONE is an MDM offering designed for companies to integrate, harmonize and manage their master data across those popular SaaS or non-premises applications that are in use as we talked about earlier and so Magnitude ONE includes a number of components.

First thing it includes is our Kalido MDM solution, which has been deployed in some of the world’s companies, and Eric, you were talking about your exposure to master data and management back in 2003, I think this product originally came out around 2004. So we’ve been an early pioneer in this space, with this tool. We started out with using it to service the analytical use of the information to make sure that good data was getting into the warehouse and over time our customers have used it more and more on operational use cases and managing multiple domains including customer and product and financial and vendor and employee and so forth. So Kalido MDM is a core part of this solution.

We also deliver connectivity and integration to a wide variety of source systems by a partnership with SCRIBE software, using their SCRIBE online integration platform as a service. That’s a cloud-based integration offering with connections to over forty systems both on premise and SaaS systems that organizations use. So with those two together, with our Kalido MDM solution it also includes and ability to have a workflow-driven environment for master data management and manage it through its whole life cycle. We have a matching engine that’s in there that’s specifically designed for handling customer data and we also provide in addition to the software, some virtual classroom training on the Kalido MDM product and the modeling components.

So Robin, you talked about the model, that’s a really critical part and that’s actually where we start in our solution and we’ll show you that in a moment, how you take that white board that Dez showed and translated that into something that can actually set up your MDM system. Your final point about Magnitude ONE is it’s available on premises or as a cloud service, you can get a subscription license or a perpetual license. The idea is it’s going to going to be easy for you to buy, maintain, implement and maintain.

So what this looks like then is Magnitude ONE at the center here, with the robust capabilities to do everything in the white and the blue boxes. So connect to and access customer data through the SCRIBE connector that I talked about. Then do all the mastering exercises you need to do around matching data, merging, surviving and enriching the data to get it clean. Then authorize and publish accurate and consistent data out to your consuming systems along with an access layer for people to search for data, browse data and even author new records so that your operational and analytical systems can stay clean as time goes on.

We provide a web-based user interface for both the stewards and the admins, which you’ll see in a moment, as well as the business users. Not only can they just browse and access that published master data, they can even play a role in the stewardship process. So imagine your sales representative’s out talking to customers, they learn something new about the customer, they can raise a change request and say hey this customer’s, they’ve changed their title, they’ve changed their email address, they’ve switched companies, maybe this physician has switched an affiliation with this hospital, we want to make sure that we keep track of that sort of stuff, or this insurance broker is now carrying these products, we want to make sure we market these new insurance products to them, for example. So those kinds of things can be raised and serviced right as your customer-facing employees are dealing with those individuals.

Couple of other attributes about our solution. Number one is this business model, remember that white board picture that Dez showed that had the circles and the arrows. That’s basically the business requirements for how the data needs to be, how it’s used in the real world. We start with something called a business information model and we can basically capture those requirements and the attendant business rules and actually deploy that to create the rules and the MDM repository. So it effectively acts as a way to bridge the communications gap that we so often see between business people describing a requirement and IT having to go back and translate that into tables and mappings and so forth.

So we have the business-model-driven approach to make sure that it’s right from when you start. We also include automated processing for that and the embedded work flow and change management so that you can, if you have a change in your model where you add to it, you can rapidly deploy that and do that with a small team because of the automation, it doesn’t require as much coding as maybe you might have expected to do.

I mentioned the model-driven nature that also drives the screens that actually appear. So when you have a description of a customer and you have their attributes there, what you’ll see on screen is the attributes that are defined in the model, so it’s all created for you, you don’t have to create any specific interface screens to map for the data, that’s all driven off of the model.

Another cool feature that we’ve introduced is the concept of Excel integration for data stewards. This means that data stewards can use Excel as a place to edit the records that couldn’t be automatically matched and approved and deployed. Now you might think, well this is just, you’re just dumping data to Excel, right? Well it’s much more than that because the cool thing about this capability is it overcomes the problem of just having renegade data updates by loading up data from Excel.

We actually, when you download that data from Kalido MDM to the Excel interface, it comes with the validation rules. So it’ll tell you which of these cells need to be filled in in order to make it a valid record, it’ll give you a drop-down list of the available values, or the approved values for example so that you basically avoid creating errors when you’re updating the master data records.

Then on embedded workflow engine, make sure the data’s all processed and authorized for publication and it also keeps track of who did what and when and allow you to basically review and audit all of those former master data values so you can see how data’s changed over time.

So the benefit of this, in terms of customer data, is you can get to a place where you can have more personalized and relevant dialogues and interactions with customers. MDM’s becoming more business critical, especially when you think about one-to-one marketing that’s going on there and this is a good example of the cycle that occurs.

So you start with data about your customers, this is the stuff that you’ve mastered, who are they, what products do they own, what can I match in terms of customer information across multiple systems? Then you enrich that with more information about them and how you’ve interacted in the past. What have they responded to? Or how do they want to be contacted? Maybe they want to be contacted by fax so that’s why it’s still on their business card. But that information that then gives you the insight you need in order to interact.

So what other preferences? Some of it may be coming from social sources for example. Then you can decide from that what’s the next best interaction for those customers, what offers should I make? That’s going to generate some sort of interaction, they’re going to download something, they’re going to purchase something.

That’s going of course to create more data that you want to feed into this virtuous cycle of marketing interactions. As a result then, you’re going to be able to find and close new customers faster, increase upsell, deliver better customer service, eliminate errors, eliminate duplicate shipments, shipping for marketing materials for example, and eventually we get to reduce sales and marketing costs.

So one example of a customer of ours who did this, the U.K.’s post office was using Kalido MDM to deliver better customer data so that they could deliver the right products and carry on their customer dialogues in the right channel which ultimately led to higher sales volumes and increased margins for them.

So that’s just my introductory comments, I’d like to now to turn it over to Diana, to take you through and show you exactly how we do some of this.

Diana Collins: Thank you John, so hopefully we’ll be able to bring some of this to life for you all. So what you should be seeing on your screen right now is an example of a Kalido business information model. So part of the solution, what we’re going to be showing you today is an integration of data from salesforce.com. Here we’ve got out our salesforce.com model on the lower left. That is obviously a web-based application, the software is service kind of application. We’re going to be integrating it with data from our on-premise implementation of Oracle, a business suite.

So our goal is to take our contacts and account information from salesforce.com, integrate it with our accounts receivable accounts and contact information into a single harmonized account and contact structure that we will then load into Microsoft Dynamics CRM. So our scenario here is we’re migrating from having used salesforce.com in the past to using Dynamics CRM. We want to make sure that we have a fully integrated, harmonized list of customers, 360-degree view based on our new Dynamics CRM environment.

So to build this we’ve moved the data from salesforce.com and EBS into Kalido MDM, we’ve actually run the harmonization process. So in the interest of time we’ve sort of done the cooking and we’re going to enjoy the meal. So let’s switch over now to our MDM environment and just show you some of the things that we can do in the added features that an MDM solution adds to a simple connectivity integration of these platforms.

But one of the things that would happen of course, it’s you’d lose your history. You’d end up with your data in Microsoft Dynamics, but would you know where anything came from? That’s what the MDM, one of the things the MDM solution can provide us, it gets us a history.

So if we take a look at our list of harmonized accounts and we’ll pick one of those. Let’s say we picked Albert’s Stores here. This gives us some information about where this Albert’s Stores record came from. We can see that it’s an integration of two records, one it came from a salesforces.com account called Albert and Gerard and one came from EBS billing account called Albert’s Stores and they were integrated together and harmonized into this single parent account called Albert’s Stores.

We also see its original ID, we can see this day it has already been migrated to Microsoft Dynamics because here we have the CMR ID from Microsoft Dynamics. I can see the time when the data was last updated. In addition to this we provide another view that not only allows you to look at the data, but also with our graph view you can look at the associations that the data participates in.

So here we have that same record, our Albert’s Stores with its associations to its accounts receivable account, its salesforce.com account and the contacts. If we select one of these contacts we can see that that contact was actually a salesforce.com contact. Likewise our Adam Albert account was an EBS contact, so of this movement, I think on the screen it’s happening automatically, a couple of them I’m doing just keep things easy to read. But as we keep going we can take a look at the contact information and see that it came from our salesforce.com account. That’ll actually build up a view that shows us all of the relationships that our data participates in.

In addition, seeing the ways in which we classify our saleforce.com data and that there are other accounts out there that are too numerous to list. Well those things that are too numerous to list, we can still get to them. We can just scroll down the page here and get to the list of all those extra accounts that were too numerous to list in the graphic view. Of course we could start in graph view for any of these as well. So that’s one way of dealing with things. We can see the data, we can manipulate the data, we also want to be able to remediate and to fix data. So a couple of ways to look at that.

So one of the things we could do is we can go over, take a look at the hierarchy, I’ve saved our account hierarchy as one of my favorites, so I can save various categories of information as accounts as well as hierarchical paths that I could use in my hierarchy browser. So here I can drill down through my hierarchy, I can see all of the various contacts that I have with each account.

But one of the other things that this environment provides is the option to find all of the orphans. These are contacts that came in through our harmonized system that did not have parents in their sources, so these are orphans that have been left behind. So we’ve brought these over, we’ve identified them, we know that these are orphans, well how do we fix that? Well we just click this switch to edit mode, which opens up another view of the hierarchy and we can now start classifying these folks. So maybe Bill Murray worked for AC Network so we can take him over and add him to the list and we see him highlighted by pointing out to us this is a change. I can move Sandy and assigned her maybe to AG Edwards and Company.

As these changes are being made, they’re being recorded down here, I can undo them if I realize I’ve made a mistake. I can gang multiples of them together and move them through the system as a unit by giving them a name and then they were processed as a single unit of work through my system. So this is one way and obviously if I’m being proactive, I might want to go in here and look at this and if see if there were orphans and address that problem. What if I didn’t? What if I wasn’t being proactive? Well, again our system includes a workflow, which I mentioned earlier, a workflow solution that allows us to deal with this more directly.

To do that I’m going to log off as system administrator, I’m going to now log on as a data steward, alright? So this would be the individual who’s responsible for managing invalid data. You’ll see as soon as I log on, I get taken to my inbox, where guess what? There’s our 11 orphaned records because the relationship, the association between the contacts and their accounts is mandatory. All of the harmonized accounts that did not have the appropriate connections to an account, are invalid. They move through the workflow and as we can see in the diagram of the workflow, here’s where we are now remediating records. They would then flow to an approval process, approved by the sales manager, approved by accounting, and finally authorized for publication on the next batch update of our dynamics.

Of course this could also be set up to run in real time which as soon as it’s published, as soon as it’s authorized for publication it would just immediately flow out the Dynamics so it’s up to you how you want to configure that last step of the interface. So hopefully this has given us – given you all a brief idea, an overview, of just some of the ways in which our MDM tool can help enrich and enhance our environment. There are many, many other ways that we can enhance your use of your customer’s information, and really get to that point where you’ve got a truly harmonized 360-degree view of a customer with all of the information in one place available to the users. Not only through this provider UI but as I mentioned we also provide a consumer interface, a sort of web portal where if a user knows that there’s been a change in account, he can raise up a change request and address that, and wrap that change request directly to the data steward to make any changes to this record that they see needs to be made. So at this point I think I’ll turn it back over to Eric and we’ll go into the Q and A.

Eric Kavanagh: Sure thing. So we’ve got a couple of questions from the audience here. I’ll throw one out but maybe first Dez or Robin, do you have any questions? Let me start with you Dez.

Dez Blanchfield: One of the things that I come across every single time I go through this journey with an organization is this whole challenge of version control. Could you just touch on the approach towards version control around data or certain – you know, imagine a scenario where three different parts of the organizations are dealing with me as a customer, and then they’re making various updates and changes through now a new tool. How do we address the issue of just version controlling the data that’s coming through business and who’s curating, and controlling, and approving that?

Diana Collins: That’s an excellent question. So one of the things that is built in and baked into our solution is audit trailing and history. So I’ll see if I can find a record with history. Let me see if our Albert’s Stores record that we were using has history, as soon as I click on History Mode what this does for me – I have – this one has no changes in history. I want it as is it would show us any interim changes that were made here, and the date and time in which they were made. In addition, I can go to Full History Details and if I turned on audit trailing, I would not only see those changes and when they were made but the audit trail will then tell me who made those changes, what user made those changes as well.

Our approach to versioning is more time based rather than by setting arbitrary labels. You can pick a point in time and see your data as it was at that point in time and migrate the data as it was at that point in time. And we track of course the history not only of the data content but also of the data model. So as your data model may evolve, we add new classifications, we track that as well and you could always roll back and see things as they were at any given point in time.

Dez Blanchfield: Data models are raising challenge on there, I mean you got a significant pedigree with dealing with some substantial articles. Can you give us a couple of examples of some of the data models that are already in place and some you dealt with running this, you know, the key sectors like manufacturing, and retail, and logistics, and financial services. You got banking and loss management and so forth, so is the approach done with a previous model that can quickly spin up a project that people can start to know where the gaps are, or do they have to build and train that model themselves?

Diana Collins: We’ve taken both approaches over the years. We’ve tried coming up with models and found that the more complete a model is really means the more changes you end up having to make, to have more customizations that you can make to it for the customer. So we have really taken the approach of fragments of models, certain basic common elements that we find that really permeate in entire industries.

We have, for example, in financial services we have models for in a capital markets for securities and derivatives, etc. We have models for insurance, for property and liability insurance, for reinsurance, and both of which manage risk in different ways. We have manufacturing models for product bills of materials, bills of landing. We have other portions of the model for a supply chain or any other tracker, intermediate warehouses, distribution models, aging of inventory, things like that. For a lot of our customers, you know, we’ve got customers in almost every vertical you can think of but for many of them we have been able to develop certain core components that we assemble for our customer into a finished model.

John Evans: Yeah. Let me just add to that, Diana. You know, the model that we showed a minute ago with the sort of orange background is really a conceptual model so it has, you know, vowels, and there’s no underscores, I mean it’s that a human being can understand. It’s not an IT concept per se, it’s something a business person can understand. We have these conceptual models, we can import an existing model that you might have and we factor it in order to get it this way but with – as Diana talked about, when we have a model fragment or an example model that we’ve used that before we show to the customer, usually within, you know, a little while of kind of looking at it and putting it up on a screen and sort of pointing and gesturing, they can usually refactor that model to get it to be pretty represented of what they’re trying to accomplish.

So it does accelerate the time in order to capture those requirements so that you can get on with it but the other thing I didn’t show here is, you know, there’s this diagram but there’s also a tab called operations where you basically press a button and it generates all of the objects that you need in the MDM repository along with the rule that you’ve been – you set up for, you know, what’s optional, what’s mandatory, what’s the cardinality, all of that stuff that you want to do but there’s a button there that says Deploy, then it would just generate the model that you’ve got created on the front end. So we do have fragments, we do have experience in a wide variety of industries and our consultants are able to enable customers to get started very rapidly.

Diana Collins: Well, the other thing I was going to find out–

Dez Blanchfield: So I the other quick one before I hand it over to Robin – yeah, sorry, go.

Diana Collins: I would just quick note that we generally run these modeling sessions as sort of a jam session because we’re not so much interested in the details of all of the attributes, we can fill that in later when we get to it. What we’re really interested in is getting the business view of how the data hangs together and how they understand it to be useful, and that’s how we want to build the solution.

Dez Blanchfield: No, that all makes great sense. One last quick one then I’ll hand it over to Robin. So the thing that I immediately imagine would happen in our manager’s conversation to organizations that I deal with is that – they have view, they already have, you know, governance, frameworks, and tools in place – what’s the experience like when you go into an organization where let’s just say the managing team decided they’re going to go down this route, become customer centric and clean up their customer data, or get a single due, and yet IT and other parts of the business may have already felt they run multiple programs of work to get to a good place on that?

Diana Collins: Oh well, that is an interesting question. Yeah, I offer that MDM implementations generally will fail unless there is that sort of high-level support. I think that these projects have to be driven from a fairly high level in an organization because there is a cultural change that needs to be accepted. I think Robin spoke to this earlier that, you know, it is not something that you just do as a project and it depends to be the way it is often approached in IT organization. It is an ongoing program, it is something that requires commitment, and willingness to change if you will to implement and then when you have that I think we found that implementations go very well.

Where we have to struggle in some implementation has been where there has been either not the high-level management support or where the IT organization has been resistant to change but we’ve been fairly successful in both cases in winning them over. I think once we showed them how simple it is to get up and operating, and how it really takes the responsibility for data content off of their shoulders, and really IT shouldn’t be responsible for that. Business knows what makes up good data, IT shouldn’t need to know that. IT should be responsible for the things that they do well – organizing data, keeping it safe, keeping it secure, and how – and usually they come around and see it that way.

Eric Kavanagh: And we’ve got a few questions from the audience, let me throw these out here. We’re going a little bit over time but I think I’ll get all the questions that we can or at least try. I’ll throw this one over to you, maybe John or Diana, either way. An attendee asks, “Do you have functionality to develop to re-parent from bad records to golden records? The transactions like for example sales orders right in the operational systems?” Not sure I know exactly what he means here, but hopefully you can answer to that.

Diana Collins: Well we can certainly re-parent records. That’s a very standard part of this office solution but within the operational systems are not directly. We could do it the MDM environment and then push that data back from the MDM environment once it’s been published from the MDM environment, push it back to the operational system but it would not be directly caught in the – we’d not be correcting it directly in the operational system from the MDM environment.

Eric Kavanagh: Got you. Okay and here’s another question, “Can the tool be used to see data lineage?”

Diana Collins: Oh absolutely, yeah. Again this is not a great model for that kind of illustration, but absolutely. Where you’ve got a history to your data, where data has come from multiple places, we can tag it with its source and carry that information forward up to the published data.

John Evans: Thanks to that. There’s an element of that here in the model, in there Diana, I mean you got the SFDC Contacts and the EBS contacts and that actually came through in a graph field as well. It kind of hangs around the data.

Diana Collins: Yeah. I mean obviously in a real lineage environment, you’d have a more robust solution and implementation and just a basic one was done here.

Eric Kavanagh: Okay, good. Just a couple more questions and then we’ll wrap up. One of the attendees is saying, “How do you support the definition of household? Do you have a way to enrich the customer master data with social networks?”

Diana Collins: That’s on our road map, enrichment with social network from social network data is on our road map. It’s not on the product at the moment but in terms of householding, that’s part of our matching and merging capabilities. In the process of matching, the great many knobs and levers that you can control for weights of particular portions of the data but what it ultimately allows us to do is to gather together all of the individual contact records that may be part of the same household. Then it understands the difference between companies and people. In companies you generally look at the beginning, the sort of the significance of the words in a name; in a company, start from the front and work towards the end. But when you’re doing householding, you really want to start at the end and work towards the front with people’s names. It understands that and is able to do a pretty good job of gathering together contacts that belong to a single household.

Eric Kavanagh: And one final question, what about restaurant customers? We have a good knowledgeable audience member on here asking if you have any restaurant customers?

Diana Collins: Actually no. That will be new vertical for us. We’d be really interested in pursuing that. We have customers that supply restaurants but we don’t have any restaurants that are customers.

Eric Kavanagh: Okay, no worries at all. Well folks, we’ve burned through an hour and five minutes here, so a very big thank you to our presenters today. We will archive this webcast so all these archives are available for later viewing. Big thanks to our presenters today. Big thanks to of course Dez and Robin for their insights, and to Magnitude Software. This is good stuff. MDM is here to stay, folks, there’s no doubt about that. It’s really important to get that central view that’s going to be more important as time goes by. I have to think as our customers decide that they don’t want to be mistreated, they want to get the best treatment possible and that’s the way it’s going to be.

So with that folks, we’re going to bid you farewell. Thank you once again. We’ll talk to you tomorrow on another webcast tomorrow, yes. Hot Technology is the hottest show around these days, we’ll talk to you hopefully tomorrow at four o’clock eastern. Till then, take care, folks. Buh-bye.