Transcript
Kevin Petrie: Hi everyone. Thanks for joining. We will wait one minute to. Allow any stragglers to join us. Looking forward to the discussion. Great. So let's get started. Thank you for joining today. My name is Kevin Petrie. I am vice president of research at Eckerson Group. We are an industry research and consulting firm focused exclusively on data and analytics. So we help data and analytics leaders and practitioners make better decisions, build and execute on data and analytics strategies. I run the research team. Our founder, Wayne Eckerson, runs consulting, and he advises various fortune 2000 and global 5000 firms about how to build and implement their data and analytics strategies. So our topic today is ROI for master data management predicting, predicting, delivering and measuring business value. Really pleased to have Stephen Lynn join us to share his perspective and expertise. Stephen is with Semarchy. Stephen, why don't I let you introduce yourself? Steven Lin: Yep. Hey, Kevin, thank you so much. My name is Stephen Lin. I'm the product marketing manager here at Semarchy. So in my past life, I was a technology strategy and transformation consultant at Ernst and Young. So helping clients kind of on the other side, you know, understand their business strategy, you know, predicting and delivering and measuring business value for their initiatives. So kind of, you know, on the implementation side and now bringing it to Semarchy, which is their leader in the master data management and data integration space. And, you know, our claim to fame is we help you guys master your data really fast and help you deliver value. So excited to be on this call, Kevin. Kevin Petrie: Great. Likewise always enjoy working with you Stephen. And this particular topic I think is a is a very timely. We're doing a lot of research on artificial intelligence, on advanced analytics initiatives. And the overwhelming thesis that I have right now is that generative AI, advanced other advanced analytics initiatives related to artificial intelligence create incredible opportunities, but they're colliding pretty quickly with fundamental challenges related to data quality, data governance, and as part of that, master data management. So I expect that we're going to see a real resurgence in interest in master data management to ensure that text, unstructured text, is using appropriate business terminology and to ensure that tabular data is is accurate because organizations want to do more with their data than ever, but they need to get the fundamentals right. So it's going to be an interesting topic and discussion today. We want to start by letting you know the results. When you registered for this event, we asked you what was the main barrier. What is the main barrier to investing more in your organization's data quality? Um, and the the findings here are right up our alley, our underscoring think the value of the topic at hand, which is that 38% of you say that your primary barrier to investing in data quality is that you don't have a clear business case or plan for executing on that. Other primary barriers include a lack of skills and resources to deploy solutions. That's certainly a persistent problem with a lot of the enterprises that we work with on the research and consulting side. And of course, macroeconomic uncertainty, will we really achieve? Will the fed really achieve its soft landing here in the United States in terms of raising interest rates without causing a recession? We don't know. And we have seen that different practitioners and their management chains are tapping that, tapping the brakes in terms of budget. Um, another challenge, of course, is difficulty evaluating vendors and capabilities. Speaker3: Okay. Kevin Petrie: So what the the fundamental challenge that master data management and data quality observability tools seek to address is that enterprises have the proverbial multiple versions of the truth. Most organizations have a fair amount of heritage, process, heritage, technology and systems, some residing on premises, some in new cloud platforms and business happens. So entities change by entities. We're referring to customers, partners, suppliers, locations and so forth. Processes drift. Organizations are under extraordinary pressure in our globalized environment to innovate quickly and then often means acquiring ventures. It often means spinning up new ventures, perhaps on the cloud, perhaps on multiple clouds. And as a result, you can have a real patchwork of of datasets. You can have silos. You can have errors that get propagated throughout an environment, or you could have a certain environment that exists and duplicates or conflicts with records in other parts of the business. So these three challenges changing entities, drifting processes, and human mistakes can create real problems where you've got conflicting data sets, siloed systems, rising costs and risks. And so it's no surprise that if we look at Eckerson group's research partner, Barc, which is based in Europe, but does a global annual survey of about 2000 data leaders, data quality slash master data management are the number one data management trend, according to respondents there for seven years running. And I think that's very interesting because we all rightfully get excited about the possibilities of new technologies augmented analytics, machine learning models, generative AI now and so forth. But the the real the reality is that data quality remains a problem that enterprises need to stay ahead of, because you've got the business that continues to create a moving target, but enterprises need to keep up with that. I moderated a track at the at the CDO event in Boston last month, and great sets of attendees, chief data officers from all over the world. And overwhelmingly, they spend most of their time worrying about solving those fundamental problems in order to enable the new stuff. So those fundamental problems don't go away. They need renewed attention. Master data management, of course, can help. And I won't spend too much time on this, this graphic here. But fundamentally, we're defining master data management as a discipline comprising practices, tools, techniques that seeks a single or at least far fewer sources of truth for the business. It creates consistent, trusted records for key business entities by matching and merging data across systems. So some of the modular systems that I highlight here are customers customer records for for marketing, for sales outreach. Supply chain records related to factories to distribution partners and so forth. Human resources looking at individuals, HR records, finance of course, which will derive from or link to customer relationship management systems and tabulate revenue each quarter and so forth. And then of course, product systems that help with the product planning process. And you can see that across these systems, you've got different glimpses of the proverbial elephant where each each blind man is touching a different part of it. You need to make sure that they are aligned and have accurate views of the business. That requires matching and merging data across systems in order to agree on and foster adoption of consistent definitions of key business terms and consistent attributes and identifiers for those terms and the entities that they describe, you need to reduce duplicates and resolve discrepancies. Now, there are four different architectural approaches for master data management registry consolidation, centralization, and coexistence. And I'll go into those in a little more detail, but suffice it to say that there are a few different options to create what we can call golden records. That that elusive version of the truth, single version of the truth that's going to help you conduct business in a more effective way, accelerating operations, making analytics more accurate and strengthening governance, thereby reducing compliance risk. So the value of master data management, if we were to boil this down in terms of the business value, which is our topic today, we all understand from the prior slide, having a single version of the truth is a fundamental good thing to have. But how do you measure that? How do you win budget for it, especially in today's uncertain macroeconomic environment and build that business case? And so at Eckerson Group, we've really distilled this into three sources of business value. One is that you can streamline process. You can remove friction from transactions from communication among stakeholders, which helps reduce project time for all types of projects and improve team productivity output per worker, our infrastructure. So here we're looking at reducing because you'll have fewer copies of data, the infrastructure on prem services, on prem servers or cloud services, reducing the number of data copies of the infrastructure needed to support them. And of course, there's the value of reducing risk, operational risk and compliance, risk variability and customer satisfaction. So if you can quantify the time you're gaining the system cost you're saving and the risk you're avoiding by adopting master data management or MDM, you can really start to articulate within a certain range what the ROI on investing in MDM is going to be. So if we boil this down, risk time, resources, execution is everything. And I, I personally like this chart because I think it reflects the reality, which is that you got to make things worse before they get better. My 16 year old son had to get knee surgery earlier this year because he heard it playing lacrosse, and he got worse before he gets better. He's on the path towards getting better, but you need to accept sort of a near-term hit. To efficiency, looking at risk time and resources. Master data management as during the implementation process is going to require some resource, some time and some risk. You're going to have potential disruption to projects. You're going to have individuals that need to spend time learning new systems, getting trained on new tools. And you're going to have to have some investment in resources to assist migrations or things like that. Reality. That's the reality of implementation in the near term. Maybe it's six months, maybe it's 12, maybe it's a lot less. Steven will be talking about that. But afterwards you're looking for that plateau of productivity, if you will, to take a turn from Gartner here. So MDM projects need to deliver a long term outcome that justifies the short term cost. Okay. So there are two ways you can we can do that. One is one of the key enablers is to choose the right architectural approach. And the four methods that that we find that we see enterprises adopting are one. Registry. Registry is it's an index. It's a central index that's going to assemble master data, key attributes and terminology associated with various entities that various distributed business units are going to consult and periodically revise their systems to align with. You can view the registry as sort of a loose federation, because the business units are, in this case, often autonomous. They need to run their operations and they can't afford to slow everything down. But they they do want to periodically consult the master data in order to align with it. Then you have consolidation, which is a it's a central hub. It's going to create master data for business units to adopt. It's going to require the business units to go to the master to go to that consolidated hub quite frequently, in order to get that master data and work it into their processes. Now the the more. Stringent approach towards central governance, which especially works for compliance sensitive organization as centralization. In here, you have that central hub that's going to create master data and require business units to really integrate their processes with the central hub. So that's going to require business units to. Rewire some processes and take a hit in terms of efficiency. And that works for really compliance sensitive organizations or business units that have real regulatory hurdles to clear in order to demonstrate compliance to auditors and to and to regulators. There's there's certainly value to it. And it's a question of how much efficiency matters to you versus compliance risk. The final one here is coexistence. Coexistence is interesting because I think it's going to require sort of an ongoing juggling act. Coexistence might be the de facto reality of a lot of organizations where you've got you do have a central hub that's trying to keep up and create and and distribute master data. But business units also might have their own master data, and they might for certain types of entities, for certain types of records, they're going to be the source of truth. So in this case, you've got sources of truth that depending on the domain, depending on the records themselves, will vary. In some cases it's central. In some cases it's the business unit. And the challenge with coexistence is you want to make sure you've got incredible flexibility, which is good. Good for efficiency depending on how you run. But you need to make sure you have some sort of overarching plan for that. So data leaders really need to select their approach based on its impact on risk, time and resources both during and after implementation. And it's going to depend on the degree to which they need business unit autonomy, compliance with regulations, efficiency. And how they kind of handle those those objectives. Now it's critical if you're going to build a business case and you're going to measure against it over time in the medium short to medium term, with implementation or long term after implementation, you got to figure out what are the dials you're going to you're going to really hone in on. And we've identified eight here. One of course is data quality. You need to make sure that you're in a pretty rigorous way monitoring records to ensure there are not errors. There are not accuracy issues. There's not a lack of consistency or a lack of completeness, you know, null values or things like that. So data quality is critical. There's a complementary set of tools. And there that that relates to data quality observability, where you're doing these checks on an ongoing basis, looking for consistency, completeness, you're validating the accuracy of the records and so forth. You're doing word counts and things like that. Um, user adoption is is is really tricky piece here because we all know and we get excited about technology itself. The reality, of course, which we see on the consulting side at Eckerson Group all the time, is that you need users to actually adopt the tools and the people and process. Part of the equation is oftentimes a lot harder than the technology, so it's critical to measure user adoption. You might have an MDM plan that you're putting into place that requires business subject matter experts to approve or even create golden records for a new customer, something like that. You need to make really, you need to make sure that you have the people who do that. The business domain experts who do that are motivated, incentivized, and approved and also have the time to actually perform those activities. Execution time is critical. Execution time related to an implementation project, but also execution time for various operational tasks and analytical tasks. Whatever the project is that the data is feeding into, you want to make sure that you're not slowing things down during implementation and that you're actually, if anything, speeding it up after implementation. Productivity is important. Closely related here. Output per worker hour. How much work is your team getting done and how does that change over time? Project cost derives from some of these, but also relates to compute costs. In particular. How much are you processing data now versus before? Ideally, you've got fewer copies of data. You have less need to reconcile errors than you did before because you're starting from a more accurate place. But project cost is always important. Customer satisfaction in many ways. The bottom line nobody likes calling a an airline or a customer support center for any type of vendor and repeating themselves. And yet we all have to do it, and it drives us nuts. So there's no faster way to alienate a customer, in my view, than to have bad master data or data quality issues. Process impact relates to a lot of these, and it's really how much are you impairing slowing execution time, productivity and so forth. Existing processes and then compliance of course, means that in regulated industries such as finance or health care, you want to make sure that you're not introducing compliance risk and that you can demonstrate in a satisfactory way compliance with internal auditors and external auditors. Now, the trick with compliance is that when we're looking at customer data, there's also there are regulations that cut across verticals such as the California Consumer Privacy Act in California, which effectively covers the United States, and then GDPR in Europe. So you need to track these KPIs during and after implementation of your project. Okay. So I'll conclude my section by by start offering an analogy. The United Nations, which I think is an incredibly ambitious and just justified in Denver, endeavor has 173 members, something like that. And the native languages for these countries are in the hundreds, but they've managed to winnow down their official business to just six. They've made iterations along the way. Think they added Arabic in 1973? Um, and they continue to balance how well they represent those six in terms of their external communications. So it's not perfect, but they figured out a way to have this highly. Multilingual community. Uh, keep them on the same page and think seeks. We're talking about geopolitical transactions and communication with the UN with we're talking about the language of business is creating value by ensuring a consistent language of business. It's not perfect, it's not pretty. But boy, it can have a good result in terms of ongoing efficiency. So is going to create value and deliver ROI by streamlining process infrastructure and risk. If you choose your architectural project approach, excuse me based on the short and long term impact, and then very closely measure your KPIs, reporting back to business owner and to executive sponsor and huddling with different domain experts such as cloud ops engineer, data engineer. The data stewards can be doing a lot of this. You can keep people on the same page, and you can make sure that you're actually tracking against an achievable goal in a good way. There's no perfect formula for MDM. Each enterprise needs to craft its own path to achieve ROI, but it really is achievable, and it's never been more important given the ambitions we have with operational and analytical analytical use of data today. All right. So now I want to bring in Stephen and we can talk through I think some some big topics here. So, Stephen, I think you're there. Maybe we'll just do a quick audio check. Steven Lin: Yes. I'm here. Can you hear me? Kevin Petrie: Excellent. Yes. Perfect. I'm also going to invite Q&A from from the audience here that that often makes these webinars even more valuable. People can learn from their peers. So please chime in. Um, so, Steven, I've kind of offered my view of how to make it work and and how to achieve success. What are your thoughts on the primary success factors for MDM? Based on your work and some archey's work with different types of customers? Speaker3: Yeah. Steven Lin: You know, I would say, Kevin, I think we talked about this a lot and I still think it's understated, but the people and processes piece is still absolutely critical, right. You know selfishly as like MDM vendor I want to say that, hey as soon as you, you know purchase our solution or any MDM solution immediately all these problems go away and you don't have to worry about anything else, right? Um, but that's just realistically, you know, not the truth. Um, you know. Speaker3: And clients. Kevin Petrie: And vendors. Steven Lin: You know. Yeah. And as you know, and it really is about kind of, you know, talking to the people and building it for them and saying, you know, what is your actual problem? Is it, you know, your customer's support, right? Like, you don't know, you know, the customer that's calling in and you know, their history of challenges and what you can do to help them now. Or is it, you know, your data stewards are spending just so much time exporting everything from Excel and then just, you know, doing like these ad hoc reports and it's taking, you know, two hours per day when they could be doing something else, like if they had better trust in their data or better quality and then put it in like chat or some machine learning model, right, to just do that all for them. Um, so I think that's really important is like, you know, a lot of times we just say, okay, great master data management. Awesome. Let's just go do it and then we'll, you know, have everyone else, you know, they understand it. So they're going to adopt it and no questions asked. Right. And we find that that's, you know, rarely if ever the case. Um, and you know, it's not just me saying this. You know, there's a lot of research think new vantage partners, they were conducting some research, uh, you know, from some of the bigger, you know, maybe more qualified qualifiable personas, right? Like from executives from Chase, Sanofi, even the white House. But, you know, 80% of the challenges are people, process and culture. And that, you know, that leaves 20% to technology. So it can do a little bit of the lift and help, you know, set a precedent, foundation in a way to communicate and collaborate. But there's no way it can lift all 80% of the challenges for you there. So to keep it simple, it's still people and processes are the primary success factors, and you just had to talk with them and empathize and understand what do they actually need. Kevin Petrie: Yeah, that makes sense. I agree completely, and as I said, it's refreshing to hear vendors talk about the people and process elements because they're so important. Another factor is when it comes to evaluating tools, figuring out not just the ease of use, but also how much training it's going to require and how much of a of a shift or jolt it is for customers to stop using one tool and start using another. And the more they can consolidate tools, the better. Because we've seen survey data that shows that folks are continuing to add to the number of tools that their teams use, which creates. A lot of overhead, trying to figure out what to put to work where and trying to make sure that it all works together. Steven Lin: Yes, definitely. Yeah. And, you know, in kind of my experience doing a lot of, you know, the financial modeling and business case for these projects, you know, most time, I think ten, 15, 20% should be allocated to, you know, this kind of larger bucket of like change management training, you know, those kind of things. Um, and, you know, if the larger the project, the bigger that number becomes. So I think it's definitely, um, you know, it's not a small number or factor to ignore. Speaker3: Yeah. Kevin Petrie: So, um, so extending this, that that question of success factors, what can go wrong, the biggest pitfalls in risks that you've seen with, um, with some of Simakis customers and can offer perspective as well. But I'll let you go first. Speaker3: Yeah. Steven Lin: You know, I think probably one of the biggest things, you know, this might be a kind of lead on to the question following, but I do think, you know, having dreams and visions is amazing, right? Like, you know, having a vision for what you can achieve. You know, all for that. I think sometimes the biggest pitfall is, you know, there is kind of this notion that. There needs to be this massive thing and everyone needs to get on board. And there's only kind of one way to do it, right, you know, this way or or no or no way at all. Um, and, you know, the larger the project gets, you know, the more intricate and intricate and complex and interrelated each kind of factor contributes to it. So, you know, I think those that's one of the things then, you know, not being able to kind of then, you know. From that recommendation. How do we narrow down and identify like what is what one should go, you know, then what should I start solving first? Right? Or you know, and how do I find the budget and the business case to, you know, actually prove that, hey, this is worth it for the whole organization and everyone's gonna, you know, benefit at some point and kind of like you're saying, you know, this time, this timeline of, you know, short term kind of pain, right? For the long term gain, you know, is that worth it? And how long is that going to be? I think there's a lot of these like unknowns oftentimes that, you know. We just don't know. And we're kind of hesitant because it's so big, it's hard to think about everything connected instead of, you know, let's pare down and, you know, choose one case that's really, really helpful for us and let's go forward, you know, implement it, learn it, see, you know, and then go on and do the next thing. Right. So I think the biggest pitfall is, you know. Dreaming too big without the action plan or the business case right to back up like you know why this would be worth it, and how long it would actually take to make that vision come true. And then, you know, continue on. The kind of first train of thought is, you know, this change management component that's involved with it. You know, a lot of people, you know, you and me included, we've been through, you know, our own kind of anecdotal experiences of, hey, we're changing to this new project management tool or this new, you know, we're going from Slack to Microsoft Teams, right? People are hesitant. We're humans. Right? Like change is inherently uncomfortable for us. And we need some kind of certainty and continuity. So just introducing change without, you know, this kind of comfort that there are things that are going to remain the same. And, you know, there are still familiar items. Um, it's super important to actually make sure that it's successful. And, you know, users actually adopt and use it within their day to day. Speaker3: Yeah, yeah. Kevin Petrie: And so think. You're right. It's it's hard not to talk about biggest pitfalls without getting into the question of a big bang or incremental change. And what we find is that, yeah, it makes perfect sense. And I think intuitively, people understand that they want to identify one, one point of demonstrable pain and fix that, isolate that and fix it, and then move to and thereby demonstrate success to the business. Get more executive buy in, more political support, more budget to tackle a next the next challenge and sort of build up this string of wins and hopefully quick succession. And so that makes perfect sense. The challenge, of course, is that you alluded to this. There are a lot of interdependencies with master data management, especially the the cliché about becoming half pregnant is very true because you might think you've isolated one set of problems. But in all likelihood, if you're a big organization, you've got some customers, some partners, some factories, some parts that have global implications. And so it's hard not to create this cascading set of changes, no matter how modular or modest your initial project is. Speaker3: Yes. Steven Lin: Completely agree and think, you know, to maybe, um, you know, since we kind of answer some of the that their question maybe adding some perspective on like what we see, you know, the first starting point for a lot of our customers and their approaches, I think that could kind of be helpful to supplement, you know, because, you know, a lot of these conversations are related. It's like one of those things that it's easy to understand, but it's not or it's simple to understand, but it's not necessarily easy to do once it's kind of, you know, on your plate to say, okay, now I have to figure this out. Um, so, you know, think in terms of like, you know, the big bang and incremental change. You know, we pretty much always recommend doing kind of this agile, iterative approach, kind of like start small scale fast. Um, and in terms of, you know, what that looks like, right? Um, you know, it doesn't necessarily have to be like a single process or like, you know, appeasing to like a single person or decision maker. Um, what we usually see is, you know, a lot of our clients start from a couple different domains, and that could be either, you know, um, think about 60 to 70% are kind of customer, um, B2B, B2C because, you know, that's where a lot of your money comes from. That's usually your biggest source of, you know, master data. So about 60 or 70% start from there. And that's, you know, extremely valid and, uh, you know, provides immense value starting from there. Other times it could be geographical, you know, if you manage a lot of locations or if you just want to start with United, you know, the US, and then you want to expand those kind of learnings into other countries. You know, that's another approach. And then sometimes even just departmental, right. You say, hey, finance. You know, they might just have a analytics and reporting problem because they're not getting, you know, sales data from Salesforce and they're not getting, you know, supply chain and partner data from SAP. You know, those can also be approaches there on you know where to start. And then in terms of um, kind of okay, how should I model or architect this solution? So then, um, you know, in the future, I don't regret having this massive solution that I can't customize. I can't kind of adapt, you know, when chat comes in and, you know, I don't know if I can legally use all this data that I have to, you know, based on, like, my, you know, enterprises compliance and policies. Um, you know, a lot of times we see our clients just starting with, with just like, you know, something as simple as, like, um, you know, the registry or sometimes just like a reference data management play. Right? So, you know, essentially just organizing, like, what does a customer look like or what are all my products, what are all my sales and, you know, hierarchies that it has. Right. And you'd find, you know, it's it's like, oh, wait, why have we not done this? You know, we have all these like screenshots of org charts and where these territories and products have. So, you know, there's kind of this false conception that we already have this, but you'd find that a lot of people might not know that. Oh, I didn't know we had this, you know, hierarchy or this division or, you know, these product lines. So even something as simple as that is, you know, often it's really fast to start and stand up, but provides immense value without, you know, a lot of pain. And it's a great starting point. And especially in the compliance space, you know, we have, you know, financial services clients and, you know, healthcare organizations that need a lot of external data that they frankly just haven't, you know. Use to kind of tie standards and consistency to. So that could be like ISO standards, you know, that could be, um, you know, medical terminology and stuff like that, that, you know, you. Ultimately need to adhere to. So I think those are kind of two good ways to think about it is like which one of those can be the quickest wins? You know, you can do it, you know, in a few weeks, right? If you really kind of put your mind to it and then say, okay, we you know, this is really helpful. It gives us a great starting point. What's next? So hopefully that kind of, you know, expands a little bit of value on that third question there. Speaker3: Okay. Kevin Petrie: Um, so that's great. And so the I think the next logical question is to say, okay, we understand that the impact on on time, on risk, on resources is going to help you measure the ROI and the business value that's going to drive the return on investment. Um, I believe that you've created an ROI calculator, but perhaps you could give us a glimpse of what an ROI calculation looks like so that folks can understand and start to put a better picture in their head about how to have the budget conversation on MDM. Steven Lin: Of course. Yeah. So, you know, we've kind of created two versions of it. So one is kind of, you know, a simple, more simplified version. And both of. Speaker3: Those are. Kevin Petrie: Should I go ahead and open this. Speaker3: Up. Yeah. Steven Lin: Of course. So you know, for our listeners, you know, if you come chat with us, it's free and simple for you to kind of access this. But it is, you know, backed by research, you know, by IDC. So they surveyed about 1000 different leaders and practitioners, you know, on what their impact on improving data quality and data trust had on their organization. So, you know, massive increases, you know, in customer satisfaction, time to market, you know, and even instances like profit, cost, revenue. So, you know, those are all kind of ways that we can kind of set some kind of precedent on, okay, if improve data quality, you know, whether that means like reducing time to access the trusted data, time for my data stewards to actually, you know, resolve errors and, you know, remediate data quality issues. Um, it really helps me kind of understand, okay, what are the actual, you know, bottom line benefits as well as kind of the softer benefits? Um, for me as an organization that I can enjoy. And one thing to note is, you know, from, you know, from my experience building multiple of these and going through it with a lot of clients is, you know, don't get so tied up to, you know, a single number, right? That's why, you know, we provide like a plus or -20% range. So you know that, okay, if I invest this much in the, in kind of like this case that I have directionally I should expect, you know, this much ROI and in return over three years. And this is how much I need to, you know, this, my operational cost per year that I need to use to actually maintain these benefits and actually maintain the solution. So take these as like directionally accurate and, you know, don't believe in 15 year forecasts. I know a lot of times that's asked. But you know sometimes we have difficulty understanding what's going to be happening today. Right. So 15 years is such a long time frame for things and variables to come in to change. But. You know, essentially it's, you know, research backed and it's built on, you know, the costs that you have. Main thing is like the software cost, deployment, implementation, training, like we said. And then, you know, their ongoing maintenance and support staff. And then sometimes you would have to do incremental hosting if you're, you know, not consolidating any of your existing hosting spend and need to kind of spin up a new service for this. Those are all considerations to be in there. And then, you know, in terms of the benefits, we kind of have a whole, you know, kind of repository of like, you know, different benefits that you can measure, tie to your KPIs. Or we kind of take a simpler approach and say, you know, on average, you know, there's about a 6 to 10% increase in your financial and operational metrics if you improve your data quality. So we kind of take that as, you know, a stepping stone to tie to like your revenue and your expenses and use that as like a holistic measure and capture the benefits value they can expect over three years. Obviously, with the scaling indicator of, okay, you're not going to realize all these benefits your first year. They're going to be, you know, each year you're going to get more and more benefits after you learn and get used to and accustomed to, you know, the actual solution. And then there's also, you know, variables that you can turn on and off and adjust. So, you know, if you have a different hourly rate for internal versus like your external, if you want to add more risk and variability, you can, you know, control those. So it's really a comprehensive and dynamic calculator to help you give a good directionally accurate view of okay, what do I need to spend and how much should I expect and benefits over the three years? Kevin Petrie: So maybe you could drill in a little bit on the expected benefits. Um, I've offered the high level dimensions of time risk. So risk, there's there's the risk that things go wrong. So you've got a best case and a worst case. You can improve that. Worst case scenario you've reduced risk, which is good. Um, and then there's also obviously the resource costs compute costs and things like that. Um, where do those levers appear in your in your model. Steven Lin: Yeah. So you know. So for the simple one, right. We just take kind of a, um. Holistic measure of like okay, financial financial metrics. We know from these this research that, you know, on average there's a 5 to 10% increase. So we apply that to kind of, you know, the revenue for the for your department that you want to implement the domain to. And then there's a, you know, 5 to 10 roughly increase in operational metrics metrics improving operational metrics. So applying that to the operational expenses. So this is for the simple one to kind of get you a first step conversation and get you some guidance. We do have a more sophisticated one that will have these toggles that say, okay, there's kind of these eight different dimensions on do you want to do you want to focus more on improving your customer satisfaction? You know, your time to market, your reduction in compliance risk, your, you know, employee turnover, your employee productivity, those kind of toggles over a three year period on you know what you expect the benefits to be still based on the same research and be able to kind of be more meticulous on, okay, now we're in the kind of detailed analysis and evaluation phase. This is what I am going to, you know, benchmark my KPIs and my success metrics for implementing this on. So this is kind of a the simple calculator kind of what I'm showing you is to get the conversation going to say, okay, this checks out at a very high level. I want to kind of, you know, go deeper into this more sophisticated one and, you know, control these nozzles and say, you know, I might increase hiring for the next three years. So that impact, you know, how much productivity gains I might have, but also potentially cost, right, because we have to train more people on there. So those are kind of the two views depending on kind of where you are in your journey. But I would recommend everyone starting with this simple calculator just to get a directionally accurate measure and then going to sophisticated one, to kind of be more meticulous and detailed in what you actually need for evaluation. Speaker3: Okay. Okay. Fantastic. Kevin Petrie: Um. So so so good stuff. The the the final question here, and I'd love to hear your thoughts or what steps data leaders should take tomorrow. Um, we all know that eating our vegetables is good for us. We all know that master data management is good for us. Um, but now that we've started to talk in a much more granular, specific and numerical way about the value of MDM Stephen, what do you recommend that people dialing in today, the data leaders dialing in today, what steps they should take tomorrow or Monday morning to get started. Steven Lin: Yeah. Uh, my my recommendation is always do free things right. Free but impactful things. And I think the first thing you know, if you haven't already and you know, it's, you know, we're all kind of guilty of this sometimes is just talk to the people that you think might be having the experiences or, you know, actually opening up the, you know, conversation for that. Right? So maybe it's like, hey, knock knock finance, you know, um, how long are you spending on reporting? Like, is this actually a challenge or, you know, or your CRM team, like, hey, is this, you know, like, is this a problem? What can we do to help? Uh, does this seem like something that would actually help you? Because we think it's beneficial, but we want to make sure that, you know, one, it's actually going to help you and it's not going to, you know, completely, you know, derail what you're doing and make you have to spend, you know, six months in this kind of pain period with, uh, you know, unclear promise of what's going to happen after that. Right? So, you know, the first thing, the first free and most impactful thing is just start talking to, like, you know, each of these kind of core departments or understanding, you know, which which is the largest challenge or biggest kind of, you know, um, area of data opportunity. Right. And just talk to these people. Um, it's free. It might take a while, but that's, I think the first step they should take. Um, you know, once you kind of done that, then, you know, there's a lot of kind of free tools that we've developed, but there's also a lot of free research, you know, Kevin, that you guys have, but, you know, leverage those things and insights from your peers and, you know, analyst groups like Eckerson. Right. Because there's a lot of really good anecdotal and, you know, research backed data that will show you, okay, what are kind of good directions and approaches to take. And, you know. Everything needs to be taken with a grain of salt, but it kind of gives you a good understanding. Okay, how do I start? And then how do I go back to the same people and say, hey, this is what we're thinking based on, you know, industry experts and leaders on what's worked, what hasn't worked. And, you know, go back and propose, okay. What if we did it like this? Would this help you? You know, you know, alleviate some of the kind of concerns you might have and actually help you kind of, you know, get to where you want to be? Um, yeah. You know, and then and obviously, you know, a little bit of a plug, but we, you know, the calculator tool is free to come talk to us. We have this kind of, um. What we call the rapid delivery blueprint, where, you know, with a decade of experience, you know, hundreds of successful implementations, we're usually able to kind of confine this pain period in under 12 weeks. So, you know, if you're, you know, if that's something that's interested in, like how do you actually build this out? You know, we have that available. And, you know, you know, you guys have Kevin, you probably have like evaluation guides and criteria on how to evaluate vendors and you know, what capabilities stuff they need. So, you know, access all of that. As you can get your people involved early and together and, you know, create a kind of core committee group that says, yeah, we're going to be the champions of, you know, this process and we're going to involve these people early on. So then, you know, everyone has their kind of say in what this future looks like. And that's all free. Right. Like, you know, yeah, you know, you can do this tomorrow, but it might take a couple weeks to a month to, you know, get it right. And then, you know, when you're starting to do, you know, more real conversations, valuations then, you know. Speaker3: Yeah. Steven Lin: That's a that's a completely different step. But you know, the first couple steps tomorrow and for the next month or so I would really say it's just to talk to your people and read up on all these resources that are free and available and use them to your advantage right there. So I don't know, Kevin, there's anything you would add? Kevin Petrie: No, I think that's a great set. I think that a common question we get, and I might just put it to you here, is that folks do look hard at architectural approaches. And. I see less interest in centralization these days. There's and perhaps it's because the data mesh is taking hold of a lot of enterprise planning processes that people are trying to figure out how to handle decentralized environments. Um, but you know, which of these are most common among successful customers for Sammaki? Steven Lin: Yeah, definitely coexistence, I would say. And you know, Kevin, back me up if this is wrong. But, you know, usually there's like some flavor of registry starting and then it moves to coexistence because, you know, we can't completely replace the saps of the world. Right? But or you know, still Salesforce to the world. You know, we're not CRM system. We're not a ERP system. But it's so ingrained and entrenched with like everyday operations that it's easier to kind of work in lockstep, you know, as kind of this over, you know, this overarching kind of semantic layer that says, okay, you know, there's some product information, customer information in SAP, but, you know, it's better for that to kind of propagate into the, you know, the MDM solution, whether, you know, the MDM is can create new products and new customers and propagate it back. It's kind of putting all these systems in sync. And that what we see is typically um, it's the least resistance one because, you know, if you tell me that I have to, you know, I spent 20 years in SAP and you're like, you know, you're moving off of T coats and now you're going to this, you know, web web solution to do all your stuff. You're like, no, I'm not going to do that. Right. Um, so, you know, it helps kind of like maintain this continuity of like, okay. Yeah, you can still use but now the data is going to be accurate. It's going to be trusted. You don't have to be like, oh man, it's Monday. You know, it's export Excel two hours of cleaning and do my job right. It's kind of keeping that in lockstep. And more importantly think in like customer situations where you know like Salesforce, right. When you need to keep these records almost up to date in real time. Right. Like what is the latest product they bought? What is the latest kind of, uh, ticket that they've, you know, submitted or, you know, latest ad that they clicked. Those need to kind of be in lockstep. But if you tell any kind of marketer, right, or salesperson say, hey, you know, get off of Salesforce and now go into this solution, it's it's unlikely and probably unrealistic, right? So coexistence is how we see it work. The best is, you know, kind of like you're saying, right? Reducing the kind of risk, um, risk factor and time to kind of adopt and, you know, learn new things and just say, hey, this just is just going to work for you. You're not going to have to be like, I don't trust this data in Salesforce or SAP. It's just going to kind of work. You might need to do a little bit of changes. You know, you might get a new field or, you know, some of this might not be there, but it's not going to be a complete 360, right? You know, 90% of it is still going to be the same, and you're just going to have to do the human thing, you know, just change a little bit. Um, but the fundamental kind of day to day operations is not going to change. So, you know, so I would say like registry kind of start off with to kind of understand, okay, what do I have? Like, you know, is this kind of the right kind of main pillar that I want and then kind of moving into this coexistence model to then actually, you know, author or edit products and then syncing those and propagating this kind of golden record back into the systems and vice versa, without introducing, you know, additional disruptions and risks and time to dot. Kevin Petrie: Yeah. I think that acknowledging the power of inertia and and doing no harm over the long term to existing processes is a paramount importance for enterprises that have been around more than a few years, and most enterprises have been around longer than a few years. Speaker3: So yeah. Kevin Petrie: Saying about registry coexistence. Speaker3: Yeah. And, you know. Kevin Petrie: The data mesh more viable. Steven Lin: Exactly. And can kind of completely forgot to mention this. But you know, these legacy systems, right? A lot of them just simply can't work on centralized, you know, on some of these architectural, uh, these kind of centralization approaches because they can't do, you know, real time streaming or whatever. Right. But they have this kind of, you know, decades. Of historical data that you still need right in context. So, you know, that's kind of why also coexistence has seen a lot of like. Uh, adoption and interest because, you know. Most companies just can't get rid of, like some of these kind of core legacy systems that do hold, like, you know, all their customer data, not because they don't want to move to Salesforce, but they might have to for compliance reasons. Right? You know, for example, like I think a lot of banking and healthcare clients, they have to have at least a certain part of these on there for a certain period of time. And if those can't coexist, right, with all these new technologies and this way to kind of feed, you know, this kind of decades of historical data and context to the analytics or into the other systems, then it's it's not useless, but it's not going to be as useful, right, as trying to get everything together and losing that valuable wealth of information. Speaker3: Yeah. Good stuff. Kevin Petrie: Um. Well, good. We welcome questions from the audience as well. I don't see any at this point, but we certainly do invite that. Um, and if not, I think we could probably wrap up here. This is, uh, Steven is always a pleasure to engage with you on this stuff, and I'll be very interested to hear, let's say in a year or 18 months, I'll have to have a conversation and see whether and how the I enthusiasm of today is driving demand for master data management. I view it as inevitable and overdue. The renewed focus on MDM. Um, but we'll just have to see if folks are able to hopefully take some of our pointers, articulate an ROI and go achieve it. Steven Lin: Yeah, Kevin, if you if you're right, you know, 18 months from now on, that prediction, I might ask you about your, you know, sports betting or lottery picks. Uh, so keep that in mind. Speaker3: There you go. Kevin Petrie: Crystal balls are dangerous. You start to have a lot of friends. Steven Lin: Cool. All right. Thank you, everyone, for joining. And, yeah, you know, reach out to me or Kevin if you have any questions on this stuff. Kevin Petrie: Sounds good. Thanks again. Speaker3: Thanks. Bye.
