Reltio Connect

 View Only

How to Build a MDM Practice from the Ground Up Part I - Show

By Chris Detzel posted 07-11-2022 16:54


Find the Presentation here

In this Community Show, Reltio's, @Chris Iloff , Director of Customer Success, walks through how he has successfully built MDM practices from the ground up. He will share lessons he learned, educate on how to approach creating an overall Practice of Master Data Management (more than just a technology solution), and help show ways you can expedite the practice maturation to increase the velocity of customer triumph and accelerate the growth of your practice.


Chris Detzel (00:05):

Well, hello everyone. My name is Chris Detzel, if you don't know me by now. And welcome to another Reltio community show. Today's special guest, Chris Ilof, he's a Director of Customer Success, will be presenting, specifically on how to build an MDM, Master Data Management practice from the ground up. So, Chris has had several years of experience, previous to Reltio, on building Master Data Management practices. So I thought he'd be a perfect fit to do this particular show.


Chris Detzel (00:41):

So Chris, thank you so much for coming. Really do appreciate it. Definitely excited about that. So the rules of the show, as normal, please keep yourself on mute. All questions should be asked in chat or feel free to take yourself off of mute and ask. I will make sure to ask your questions for sure, if you have any. Community show will be recorded and posted to the Reltio community, as usual. And then, we do have several community shows coming up. I'll let you know what they are. One is today's show, on "How to build Master Data Management practice from the ground up."


Chris Detzel (01:18):

Next week, we have the new modern Reltio user interface, our user experience. We have a CCAB lunch and learn session. So that's an ongoing session to really have an influence in our content strategy. So if you want to be a part of that, let me know. And then July 14th... I did skip the early July timeframe just because of July 4th and the holiday here in the US, but we do have a "Reltio Community Show: UI-Configuration Self Service". And then on the 19th, something very interesting and I'm looking forward to, a little bit different outside of the Reltio product, but it's called "Natural Language Processing - Text Representations in MDM".


Chris Detzel (02:06):

And then, "Moving Data in and out of Reltio: It's a "Cake Walk", and a lot of people are very interested in that. And then another exciting one is "Machine Learning Based Matching". I know we wanted to do that a while back and we finally rescheduled. And then we have a customer coming on with iron on "Reltio Integration Hub with Sajid Sayed at Advarra". So looking forward to all of those. We're packed with tons of really great shows coming up here over the summer. Let's get started Chris, with today's show on how to start with how to build an MDM practice from ground up.


Chris Ilof (02:50):

Thanks Chris. I get to share my Screen here in a minute. And of course, right when I went to pull it up, I had the Okta message telling me to sign in. So perfect. So we avoided a little bit of that. Hang on a second, want to make sure you all can actually see my screen accurately. I know it gets a little wonky. Sorry about that.


Chris Detzel (03:16):

Not show anything yet.


Chris Ilof (03:23):

Yeah, I know, it got a little spooked on me there when I went to the slide show. So we'll just do it old schooly here. It did full screen on me.


Chris Detzel (03:36):

Cool. We see it.


Chris Ilof (03:37):

So everybody can see it. Unfortunately, you'll see the little side banner here, but oh, well. It's the content and it's our conversation that matters today. So you, what we're going to talk about today is really more about methodology. So unfortunately we're not, for some of those that are wanting to be into bits and the bites of things, I know we can talk about that a lot, and it's very exciting to talk about, but there's an equally important part around your master data management piece, of treating it more like a practice and not just yet another IT system. And that's going to be a theme here that we have, not only in this session, but this is really the kick off to a series around this.


Chris Ilof (04:20):

So on the methodology, the practice of MDM. So the convergence of the great technology around, Reltio and other MDM technologies, but not just that systemic view right around technology and what you can do and technology capabilities, but how you apply that and the practice of it. And how you build teams around this and coalesce around building up this practice and how that approach is critical to being successful with your MDM and continuing to add value into your organizations with MDM.


Chris Detzel (04:51):

And Chris, I just should have mentioned this, is we would like this also to be a little bit more conversational. So we know that you have had, or a lot of folks here, have had a lot of experience in master data management in the practice of, so if you agree, disagree, or feel there should be something added to this, then please feel free to speak up, because today, I want to see this as more of conversational, but I also, for future sessions like this, we want to do more conversational around the practice itself. We'll get deeper, like Chris said, but this one is a little bit different than usual and we are pushing for more of that conversation. Thanks Chris.


Chris Ilof (05:35):

No problem. So yep. Another picture of me. You don't need to see that, but where am I coming from, this is really having experienced a lot of what you were all in different phases of experiencing in the past as well, or in some of you currently, but in my past. And then, with conversations with all of our great customers of Reltio. And being able to have these conversations as customer success representative with great customers like you all and really vetting out these similar life experiences and the experiences of not just the implementation, but how did MDM come to be within your organization? What was the inception of it? What was the key driver? We'll talk about commonly how that happens. What brings MDM to life and gets us all assigned to these projects that build this awesome practice called MDM.


Chris Ilof (06:26):

And we'll talk a little bit about the implementation and keys to success in pitfalls there. Things that some of you may be involved in implementations today. And if you are please, feel free to ask about that. And I know that many of you on the call today are probably post implementation, so feel free to share some of your stories as well as we move through this around pitfalls you've seen and keys to success that you may have as well. And then we get into really what we call continuous evolution. And this is the big, big chunk of it. So it's the day one right after your go live. And how do you keep things moving forward with MDM, which is a big, big struggle.


Chris Ilof (07:11):

You can struggle with this if you don't avoid some of the pitfalls and it's hard to avoid the pitfalls if you don't know about them. So that's why this series is really coming to fruition here, is so that we can share our experiences together. I can share what I've seen with my own experience and experience of our customers as well. And we can all get to a point where we're super successful in building out these practices. And we don't see this lag between implementation and continuous evolution. We'll talk about that here, going forward as well, but onward. So key here, and you'll see this as a theme that we're going to keep driving home and I already mentioned it's in the title.


Chris Ilof (07:50):

We've mentioned a few times MDM's a practice. But it's a practice of this continuous evolution as we talked about. We see inception comes and some of us are involved in that phase, but I know in my own experience and a lot of others, not really you're handed this project that more of the "We're ready to start, and we need you on this project" type of phase. But there's definitely some strategic meetings and things like that going on. There's a whole process around how these large programs, if you will, spin up an understanding of MDM as in need, and then fund the project, if you will. Then we get the implementation phase where everybody's focused on getting a technology in place, data transformation, all that stuff. And then what happens after that?


Chris Ilof (08:40):

That's the day one stuff like I talked about a moment ago and how you roll through on the continuous evolution, which is the longer running part of really your MDM practice. So just laying it out as this is high level, very high level as we go through this series. So today we'll talk at this high level through these and these bullet points, but as we go through and we have other sessions via community, we will dial in. We'll focus on very specific points of each one of these containers here. Just out of curiosity, has anyone on this call been involved in this inception phase where they were part of the strategic discussion, maybe that brought on that, "Hey, I think we need MDM or a decision there". I know some may have been involved in a rip and replace where you had an existing solution. You're replacing that MDM, but has anybody been involved from what started the first build of an MDM?


Speaker 3 (09:44):

Yeah. Yes, I have been involved, this [inaudible 00:09:47].


Chris Ilof (09:48):

Okay. Awesome. Awesome. Well, let's make sure we get some of your feedback here and great insights as well as we continue this conversation and better understand some of what happens in those spaces and how we declare the dependency of MDM. But in general, IT is part of the business. It may in large enterprises get treated as a second class citizen in some cases. Because it's, "Oh, that's an IT thing, and not necessarily always on the same level as your marketing or finance or other, key, large business functions within an enterprise, but it's absolutely there. And they're part of strategic planning, part of understanding how, what is the work that IT needs to do to achieve the business goals for this year or onward. Long term running goals.


Chris Ilof (10:42):

And where we see those long term running goals, and this need for a long investment, is where things like programs get spun up in programs harness, funding of multiple projects. What we see are, today, a lot of businesses are driving in initiatives around customer experience, operational efficiency and risk mitigation. We see this in the light of programs that are talking about digital transformation where we're moving things into the cloud as enterprises or even Salesforce implementations. Things like that. So this is where the strategic, how do we meet these key business initiatives and objectives? Well, we need to do X, Y, Z. And those conversations that come up to put things like MDM at the forefront as dependencies. Digital transformation, Salesforce, migrations, things like this, that we're seeing in more current times, all get conversations around, "Well, we definitely got to focus on the data and what do we need to do to the data".


Chris Ilof (11:45):

And this is where, well we need MDM. We need a customer 360. We need a holistic view. We need a source of truth. Those things are understood. And this is where the funding comes into play that usually brings MBM into the space as something that needs to be executed on and built. So the dependency gets declared in these strategic sessions, the projects get funded and then things get scheduled out and MBM becomes one of these foundational type of, or enablement type of projects that needs to happen really before the target space thing that's getting a lot more of the attention happens. But that's the initial set of investment and funding. Investment and funding is key to understand, because it's not only why you did the project in the first place, but it's going to be what you need to keep that continuous evolution going forward.


Chris Detzel (12:35):

Hey Chris?


Chris Ilof (12:36):



Chris Detzel (12:37):

Sandro mentions that one, he started with both inception and migration and he says IT drove the MDM inception at Empire. So that caused them some grief because many of the business use cases were not known and thus not factored into the design. So business involvement was "Well, we'll optimize your success". So in a number of those initial business use cases will help instead of trying to anticipate how the business may use MDM. So I think that's a very good point.


Chris Ilof (13:08):

Yeah. It's a fantastic point. And we'll touch on that a little bit, Sandro, and more around the periphery of that, but alignment with business and all of the data stakeholders is absolutely key. And as we dive into the implementation factor and really you want to do it upfront up front as best you can here, but alignment to those stakeholders is 100% key so that you understand where you're adding value to the business and you can attach to that directly.


Chris Detzel (13:35):

Somebody told me once, and it's from our executive council. He goes, "The key is to not talk about master data management sometimes to the business, because they don't know what that is". What business problems is it going to solve at the end of the day?


Chris Ilof (13:50):

Absolutely. So I know in my experience, I wasn't part of the inception type of decision but was brought in more in this space, in the implementation space. So once it was decided that we were going to do MDM, it was, "Hey, we're going to do this thing, the project scheduled, let's go build it". So one of the key lessons learned I had was, "I'm an IT guy, development background all the way through". Done a lot of different things in the SDLC for many years of my life, until I got this project, back in I think it was like 2008, 2009.


Chris Ilof (14:33):

And I focused on it like an IT system. So when I say build a practice, not an IT system, this is a lesson that I learned as we do learn lessons in life, sometimes the hard way. You do it wrong the first time, and you spend a lot of energy trying to correct it and then double back and take a view and say, "So, how could I have avoided some of these pitfalls"? So this is a lot of my own personal experiences as well as validating again with a lot of you all as customers and understanding some of the experiences like Sandro just shared, which are great. Implementation I think oftentimes is where a lot of us first get our taste of MDM, and get involved. And we're trying to understand this as we're executing this implementation. A lot of times there's partners obviously helping it. Either consultant firms or software consultants like professional services with Reltio and things like that, are involved to help you along that.


Chris Ilof (15:30):

But oftentimes unless you're doing a complete rip and replace, this is something that is a little bit foreign to most people involved in the IT implementation. But obviously the basics here. Scope gets defined, the project gets executed and go live is when you have whatever your initial domain is. If it's customer, if it's provider, if it's member, if it's whatever initial domain that is responsible for enabling the core key program, that's the overall funding mechanism, is usually the scope and focus of your initial implementation. And some of the things we see with this usually is that there's a focus on building it as an IT system. And as Sandro said exactly, there's not a lot of business inclusion. So this is where we talk about the data stakeholders. And these stakeholder partnerships are absolutely 100% key.


Chris Ilof (16:25):

You have to be tied to the folks in the business app space that own the data. And we can talk about some of the back and forth between owning the data and not really owning the data on another time or perhaps over some cocktails, but having that partnership way up front as you're kicking off the project so that everybody's really aligned right on what the practice roadmap is, you can build that together. That you're building a practice execution framework because there's accountability involved in many constituents that are just not IT constituents. It may not be part of the close knit MDM team. It's important to, as you're building this, as you're doing an implementation of technology, that you're actually trying to understand what are the skillset I need to support this thing when it goes live so you can manage it and keep the lights on as well as more accurately steward the data.


Chris Ilof (17:20):

And have partnerships with those stakeholders. So that it's clear who's going to be doing the data quality remediation. Who's going to be the experts in and other pieces of stewardship from anything from potential match work, and again, to other data quality issues that maybe need to pipe back to source, or that hit downstream systems and consumers. It's really important to know your data, of course. So being familiar with it, having done some data analysis upfront so that you understand and have some predictive understanding of where the gaps are in the data, that you understand what the outcome expectations might be, so that you're better aligned to, it will save you time iterating through the outcomes of your MDM and save you time as you're developing and constructing match rule sets and the technology aspects of it like that.


Chris Ilof (18:17):

And it's important that, again, these stakeholders understand those pieces of it too. It really is because otherwise it's just this invisible thing that's happening behind the wall. And we all know that. Those of us who've experienced these implementations, those of you or group who are past implementation and really working on the continuous evolution, you all know that when it's a black box, there's a trust issue there sometimes with a lot of business partners. So making sure that they're in fold early on the implementation side, that you're developing those partnerships and that you're really building out all of the constituents of a practice across functional constituency for this practice, are absolutely key. Iterative development is definitely a key, especially as we talk about the continuous evolution, but when we see this and you'll see a graph in an upcoming slide, but we see through implementations oftentimes is because of that block box step of mentality of the magic going on in the algorithm in the background.


Chris Ilof (19:22):

And because you're waiting really for this end state that's anywhere from three, six, nine, sometimes even 12 months later and large implementations for this entity, this master customer, this master provider, this master whatever this business entity is, is the end goal. And it takes a lot of time to get there. And it's really difficult to reinsert people after that, have them trust it, have them understand what's happening and why that entity is right. Why it's correct. Why it's better. So making sure that you can have some iterative milestones and doesn't necessarily have to be agile, but deliver things in an iterative motion, not just a big bang at the very end, "Here's your new entity. Trust us, it works". Making sure that you're tied to those data stakeholders and that you're delivering value at critical points and critical milestones along the way.


Chris Ilof (20:26):

Critical to being able to do that, I think is a modern MDM approach where we're all looking at the cloud now. So there's definitely implications of how you can do that iterative development. How you can achieve value faster to keep the attention of those critical data stakeholders by showing value that your MDM practice is providing along the way as you implement. Whereas some of the legacy MDM still really require a lot of this heavy lifting of data transformation and ETL in and out of their solutions, the building of infrastructure. All of these things that take a lot of time and money. So modern MDMs are something we want to understand there. And we'll dive into to some of that in future sessions as well. So based on your own experiences, any pitfalls that haven't been talked about here for folks that have gone through an implementation and seeing what it's like a little bit after the implementation? Had some time to look back on it, and anybody here have some input as to keys to success maybe, that we didn't talk about here or pitfalls that you saw?


Kurt (21:35):

I have a comment. Hey Chris, this is Kurt. How you doing?


Chris Ilof (21:41):

Hey Kurt. How's going?


Kurt (21:42):

Good. So just coming off of an implementation, I think one of the really interesting things for me was to, and this is exactly on the point of having know your data highlighted on your slide. So not only do have to know your data, like you said, but the business users have to see the data that is being licensed, for example. In many cases, you're getting data from a third party, let's say it's IQVIA or some other vendor of third party data. You need to put that data in front of the users, show them what it contains and confirm with the users that the data has the information they're actually looking for. Because oftentimes they make assumptions, "Oh well, we're going to license this data. It's going to have pharmacy", if this are life sciences, for example, which is the implementation I'm coming off of now.


Kurt (22:36):

Well, of course it's going to have compounding pharmacies. Of course, it's going to have wholesalers and buying organizations. Of course, it's going to have all this. And it turns out that it may, may not, but even if it does, it may not be in a form that's directly accessible from the data. You may have to create some complex and interesting queries or ways of teasing out what it is they're actually looking for. So I think it's really important to put that data in front of the users and make sure it contains what they're looking for. And they're going to be satisfied with the information that they licensed.


Chris Ilof (23:15):

Yeah, absolutely great point. And Kurt, I would even extend that to say to, what Bill's talking about in the chat there, as well as in my experiences, heck I even had business application owners that didn't know the data coming out of their application. And this is like a custom internal application. And we see that too, where you end up with data quality issues or, "Well, that's not my data. I didn't know it looked like that", type of thing. They're so used to their interface and maybe a set of reports that they don't see the raw materials, if you will. So I think that's a great comments from both of you. Think it's suspended.


Chris Ilof (23:58):

Yeah, you're right. It may require further enrichment and transformation before you're able to actually, there's going to be some processing to make the data consumable to your MDM solution and get the outcomes expected. So it's important to have that visibility and alignment through all those data stakeholders without knowing your data.


Sandro (24:16):

Hey Chris, it's Sandra. One of the things that I'm not sure how common this is going to be, but because it was very lack of experience in master data management at Empire Life, they actually ended up getting a third party to do the development work and then it was handed off. Which caused a lot of problems as well. So that our developers, for example, didn't really have a good understanding of what was delivered. And it took a long time to resolve issues that we would run into down the line. So I guess what I'm trying to say is that make sure you have your IT folks involved right from the beginning and whether they're going to be marrying or working with a third party, that they know everything that's being done right from the beginning, otherwise it becomes again problematic to fix things after the fact when the third party vendors left.


Sandro (25:08):

But having said that, making fixes after the factor, changing the data model or whatever happens as a result of things like that, I have to tell you that Informatica, which is what we use, was very difficult to make those changes after the fact. It took a long time to run through that life cycle. I think the way you guys have designed Reltio, the configuration side of it is a lot more straightforward. So I think you have the luxury to been able to fix things a lot quicker and be able to test it in this environment. Then you would've in a traditional MDM on premise environment. So I mean kudos to you folks for that ability. But again, the more you can do right up front, obviously the better it is and not having to make changes down the line because didn't take into account all the business rules or you didn't have folks have understood how it all worked.


Chris Ilof (26:00):

Yeah, absolutely. Thanks for that. Really good stuff. So next is a slide up. Can I show you where, what we've seen, what I've seen even, like I said, in my own experience and then even being able to map some of the data that we see from Reltio after implementations, and we take a view of this is all a perspective of a view from when a customer buys Reltio to when a customer is done with their go live, to when a customer is ready to expand on their implementation. So the purple line there is what we've seen when it gets treated like an IT solution creation. When you don't avoid the pitfalls that were on the previous. And to everything that we're all aligning on and saying here, when people don't understand the data, when you don't have those partnerships going in and when there isn't really a clear understanding of the business value even, as part of it that keeps people on that journey and through it. Everything that we all just talked about.


Chris Ilof (27:13):

This is how you see it. It looks a lot like, if anybody's ever seen the hype cycle curve, it's very similar to that when we mapped it out. Now this isn't a true representation of the data itself. I'll be honest, it's a sketch up of what that curve really looks like. But what we see is in the purple, when you have that technology focus and you're not building this constituency to build out the full practice, is really this stall. And this lag where it takes a while to get things back up and running. And it's exactly some of the things that you talked about, Sandro, it's a lot of if you haven't focused on building the practice, then you're focusing on it here after you've implemented.


Chris Ilof (27:55):

And so you're really going through that forming, storming, norming of building out a practice and what the heck does it mean to own and operate this thing. So at the same time that you're trying to keep lights on from a DevOps perspective, you're really trying to build out and understand what the heck this practice even needs to be and who owns what, and what are our core responsibilities here around the data. And as you're doing that, the perception of value that your key business constituents and data stakeholders have, just wanes. It just tanks. Just like the hype cycle. Everybody's super excited. They get it. They understand the value here. They understand it through implementation. And then inevitably usually, there's things that happen in implementation, especially when we're talking about disparate data from here to there and everywhere, and the coordination of all these teams and alignment of that things take longer. The understanding of value wanes and disappears as you're focused on very finite technical issues to get this thing launched. This technical solution launched.


Chris Ilof (28:57):

And you lose track of that need to tie everybody together in this practice and the view of what it could look like going forward. What is the long term MDM roadmap for our organization? Because you're only focused on this go live, this thing. This one entity that we're going to create and not focused on really that roadmap of this consistent understanding of valued driven type of roadmap as well. With that, what comes from that? Well, you struggle to fund. So we talked about the investment that's key up front and you're doing a digital transformation. So there's all this excitement, all this money, "Hey, we're going to the cloud and MDM's going to be a big part of taking us there", and you've got this chunk of money. And then it does the implementation and you spent that, and now you want to do all the other cool stuff that you could build onto that entity that you just made and you go to talk to people and they're like, "Ah, I don't have money for that".


Chris Ilof (29:55):

And you don't have money for that, because you've got a DevOps type budget that allows you to do keeping the lights on. And not any of the expansion things that really bring enumerable and expansive value to your organization. When we do keep a focus, we keep this team align, this cross functional team to the practice of MDM and not only what it does and how it applies to technology to build business value for organizations. When we keep that line of sight onward, and we keep that team together and a focus around that, this is where we're able to avoid that trough and we're all focused on this future venture that doesn't stop at go alive. It really just starts. So you continue that enthusiasm. . You continue increasing the perception of value that MDM can provide for the overall organization, because you're attaching it to that business value and a future business value as well.