How to create a successful data strategy? John Bottega, EDM Council

Hi everyone. You're listening to the Data Insiders podcast, the show about data and how to use it to shape the business and the world for good. I am your host Thomas Kirk and with today's guest we're here to talk about how to build a business-driven data strategy and what to do with it once you've got it. John Bottega is a senior data management strategist and the president of the Enterprise Data Management Council. John has held various roles in driving the Enterprise Data Management Strategy. He has worked as a chief data officer in organizations such as Citibank, Bank of America and Federal Reserve Bank of New York. John, warm welcome to the Data Insiders podcast. Thank you for having me on this podcast. I really appreciate the opportunity. So John, you're president of the Enterprise Data Management Council. Can you tell us a little bit about what that does? In short EDM Council, we are a non-profit trade association that focuses on data management and best practices in data management. So our constituency are the data professionals, the chief data officers, the head of data governance, the head of data quality, etc. We've seen a tremendous expansion of these roles in businesses over the past decade. The Council has been around since 2005, started initially within the finance industry but has grown to other industries as now data is important to everybody. What we do as a trade association across a number of things, we are advocates of best practice because we believe that facilitating good practice on data management is key to ensuring that there's trust in the data. We provide training and education. We collaborate and what we do is as a again as a trade association we don't want to be the smartest folks in the room, we want to bring the smartest folks into the room. So we work with our membership, we work with experts in the field and we conduct work sessions and webinars and things of that nature to share that knowledge amongst the data professionals in the space. We have enjoyed very good relationships with the regulatory communities as you can imagine. Lots of you know the legislation over the years on on transparency of data and good data practices. So we have great relationships with a number of global regulators because again as a non-profit we don't represent anybody other than we represent good practice. So that's kind of it in a nutshell. I mean we're here to support our membership and I could say knock on wood. It's been very successful as our membership has grown and our opportunity to do work in the industry has grown. I think you're being a bit modest saying that you don't have you don't have the data experts in the in there that you pull them in because I think you're you've got a rather good experience yourself. We have some background in it to be honest. You know my background as a Chief Data Officer for you know almost a decade certainly helps my opportunity to have conversation with other Chief Data Officers but this industry is changing so rapidly that the opportunity to hear from people you know what they're doing and what new innovations they're coming up with and so forth. So that's what we try to do. We try to track all of this new knowledge and bring it together so we can share amongst the professionals. Thanks and you've been working for EDM Council for over eight years now and what kind of changes have you seen in in the demand for the EDM's Council's work over that time? Oh I would say dramatic. So my involvement with the Council started as a member when it was first launched. I was the Chief Data Officer at Citibank and I was a member of the Council. About a year and a half in I was asked to take the role of the Chairman of the Board of Directors which I did that for almost eight years. When I joined the Council eight years ago we were a third of the size. We were still kind of you know collectively industry licking our wounds coming from the financial crisis you know trying to get our arms around all the things that that took place and how data could help facilitate a safer environment etc. But we've also seen the industry shift from being purely defensive you know I have to add this report to the regulator and being offensive in the sense that hey let's use data for you know growing revenue, better service tower clients and customers, new innovations and so forth and that shift has has really driven the industry to increase the number of hires. I mean the number of CBOs that have been hired now dramatically increased and like I said across multiple industries now you see the manufacturing, pharmaceuticals and insurance. The government agencies have increased their play you know you see more government organizations with Chief Data Officers now and Chief Data Analytics Officers. So that that's dramatically changed and I just snuck in the word they're calling the Chief Data and Analytics Officer and that's another indication of the change. You know the CBO I would say when I was CBO was probably more around the operational side of data but as that role has evolved now the accountability of not just not only curating the data but making it available for analytics has also dramatically changed. We're seeing more accountability of the analytics agenda in firms falling to the Chief Data Office. Something to catch up on later of course the cloud coming in and causing all sorts of new issues and new opportunities. Just going back to the you mentioned the different industries starting off in financial sector but now seeing much more industries involved. Are these other industries sort of catching up with the finance sector that had that bit of a head start or how are you seeing that? It's an interesting question because there's always the assumption that the finance industry is is farther ahead in data management and to some degree they are because of the financial crisis right. You know regulators were putting pressure on the banks to say listen you got to get your data on the control and so forth. So there was a I would say more a couple of years ago where you saw a greater difference in the maturity of data management practices within finance versus non-finance. But I'd say in fairness the the rest of the world is caught up. Think of things like manufacturing. The efficiency of manufacturing is largely based on analysis of data coming back into into those processes. You know evidence of this is our membership includes organizations now like you know Coca-Cola and Pepsi right retail organizations. The U.S. Navy became a member of the council. You know that there are large institution with many many employees and they need to manage their data effectively. Insurance you know go across the spectrum of players pharmaceuticals all recognized how important data is to their operation as well as to their clients. So I would say we're starting to see a leveling out of maturity as as other industry you know run into this hey we got to we got to get good this too. Our main topic for today was voted by the data insiders community and that's data strategy. So how would you describe a successful data strategy? One of the the things that we have been huge advocates of in the in this data management best practices is a model called DKEM data capability assessment model and the very first component is data strategy. As a as a former CDO and I was a CDO in both public and private sector it is still so important to communicate the objectives of your data program to your business and management. You know we've become accustomed to the chief financial officer being responsible for all financial activity or the chief risk officer being responsible for risk activity. The CDO is still relatively new so a lot of firms a newly appointed CDO has to now go to management go to the business and say well this is what we're going to do. Now in developing this strategy I always advocate the idea of a two-pronged approach. One is a data strategy and one is a data management strategy. What's the difference? The data strategy really talks about the content. What information do I need to you know improve my business to facilitate my operation. So you know oftentimes here you know the beginning of the year the CEO steps up and I'll say something like you know we're going to be this is the year of the customer. Well if I'm a CDO I just heard that my customer data had to be you know 100% right well it's how can how can I help the company improve our customer service right. So what kind of data do I need is part one of building the strategy and why that's so important nowadays because data isn't just two-dimensional anymore it's it's on structured data it's data coming from IoT devices. One of the best examples one of the government agencies responsible for mortgages built a fantastic system to assess the risks of their mortgages by incorporating climate data and geospatial information so they could look at properties and go well this property is is near a coastline it's near a river that floods it's near you know so they were incorporating all these different data sets into their into their you know operation that in in in my advice has to be facilitated through that data strategy what data do we need to get our job on the data management side how am I going to operate this thing you know who's accountable many firms have tried this and it doesn't work that well when they try to centralize all data management it's a it's a federated approach and because everybody from senior management all the way down to the brick and mortar are touching data in different places so how am I going to operationalize this culture of data management how do I ensure quality how do I ensure I have ownership you know accountability things like that so those are the two key prongs of a strategy is what data do I need and how am I going to you know how am I going to facilitate efficient operation you mentioned there about the the insurers then making more use of stranger types of data and I think we're seeing some cases where there's more of not just business data in the data strategy but then data actually going out there into the business strategy changing driving new business opportunities and changing how they work so how are you seeing that developing I guess the best way to describe it would be alarming experience right you know none of this is a hard line between the data team and the business team it's a collaborative activity right I've seen this over the course of my career and with our members presenting data to business spawns ideas right you know the business may not be aware that I have this data available to me right we've been huge advocates of what what I like to refer to as a sandbox you know working with business giving them huge amounts of information to play with so to speak not productionize but you know innovation comes from business so if they know they have this data available they have ways of you know measuring you know this is a an oversimplified example but I'm sure you've heard of this where aerial shots of a company's parking lot could be indications of you know the flow of customers and and how busy you are and you know change my hours so I can accommodate the you know the customer flow that's a simple but good example of if the business wasn't aware that that data was available they wouldn't know to adjust to accommodate those types of scenarios so yeah absolutely it's it's a it's a partnership if you will to learn from the data that's available to to drive the business in different ways somewhat like those known unknowns and unknown unknowns from that old quote as they say I know what I know and I know what I don't know right so and this really speaks to a data culture there has to be an understanding that there is this learning curve there is this experience that can be achieved that and benefited from it's not a clean cut here's an answer that said there you go right there's a back and forth and and boy if you throw an artificial intelligence in there that just opens up a whole another dimension of learning and experiences and so forth so the culture of the organization has to be in concert with what we just talked about for it to really to benefit the organization you mentioned the decam so the data management capability assessment model including data strategy so I'll give you the opportunity for a quick sales pitch and so what's the benefits of using the framework and what kind of organizations benefit most well like like anything when you can codify best practices it becomes a really powerful tool when we developed decam um this is going back six seven years ago uh in fact even before that the original idea was what i was chief data officer at the federal reserve and some of my and we were just starting to have more chief data officers you know we got together one day and said you know we don't have anything you don't have a book on this nobody's written anything on it how do we make sure that we're doing this the right way so the the the impetus of this was can we capture these different concepts and codify them into it into a framework that everybody can benefit from so in models like decam you're not going to find any earth shattering things in there but what you are going to find is a very methodical approach to how to build and sustain a successful data management program so starting with strategy then how do I build my organization how do I how do I incorporate in and partner with technology architecture data quality data governance uh analytics all of these are included in this model as a guideline now what's really important in building these models they are not prescriptive and they're not intended to be because we don't want to dictate to somebody how to run your organization instead we want to say to run a good data quality program you need these capabilities and how you facilitate this capability organization how you implement them is up to you but as a as a um an auditable tool I can use that to say have you built your data quality strategy are you using these tools are you do you have checkpoints do you have assurances from an audit perspective that the data is of high quality see that's the difference it's a capability model more than it is just a maturity model identifying the capabilities you need in those different aspects of a good data program so if we get then from here down into some of the the details and and practicalities and keeping in mind that many of our listeners are going to be working in in smaller and non-financial companies so that's somewhat different to the scale of things on the other side of the pond um but first off uh having a data strategy so some quite a few organizations don't seem to have a documented data strategy at least so does everyone need one I mean tell me everything is scalable everything that we do was designed to accommodate the organization its size and its structure and its and its its culture um but I can tell you unequivocally that the concepts addressed in this decam model and similarly in the in the cloud model that we build really is ubiquitous to scale and what do I mean there I'll give you a good example I was chatting with an organization a while ago that was collecting sales data from sports arenas and they're facilitating it through bio recognition you know thumbprint on a dial so I don't have to pull out my credit card every time I can do that kind of thing small company growing um the question that was asked what are you doing with all this great data that you're gathering and they what while we're just facilitating the the transaction side think about how rich that data is and how you can support your customers so even a small company can create a data strategy right that can help grow their business now you know you're not going to have a dedicated person here dead again you may not have a chief data always a chief risk always a a chief governance over that type of thing but the functions need to be there right you want somebody who's going to be accountable for the data the data governance and data quality could be the same person in small organizations so we don't we don't try to um suggest that you need all these individuals we suggest that you need the capabilities and that scales very nicely to companies of all size really and we've been working with mid-tier banking organizations we've been dealing with boutique companies um the the benefits they realized by just stepping back and going I have this great asset in my company information and and this is how I can better use it uh it's a win for everybody in those cases if we actually now think about the actual documents the actual things that someone's tapping away at their keyboard to make what kind of thing do you actually recommending people create so the course of my career I had to I was on both sides of the table uh where I had pen and hand as well as I was giving advice to someone with the pen and hand um my recommendation and and again not just based on my my opinion but seeing how it you know what works and what doesn't work um I was in a meeting some years ago in a relatively large organization and the chief data I was said I I had my data strategy held up a single sheet of paper that might be a little too thin okay um the flip side of it is don't don't give me a tone of pages and pages and pages which no one will read right you have to implement your artifacts in alignment with your audience and even that a data strategy you know there was a Ted Talks a while ago which one of my favorites called called talk nerdy to me right or don't talk nerdy to me uh we don't want to you know talk too technical with the business and likewise I want to talk to with the tech technical people so so building a strategy artifact doesn't necessarily be one per se but you could then customize that artifact if I'm going to meet with the business what do they care about sales and customer support and so forth so gear my strategy to that audience if I'm talking to the technologist and we're talking about AI and machine learning and data lakes and all that gear my strategy towards that audience and likewise senior management they probably don't want to hear about either of those things they just want to know are you you know responsibly managing our information so a core data strategy can easily springboard into other types of artifacts but the best practice guide or the best advice is make sure you're tailoring it to the audience you're presenting it to and and then you have less worry about how big my my artifact should be it should be did I get my message across to the audience I'm speaking to so then you've got your your great message uh in its various forms the different audiences how do you then make sure that those don't just become bits bits and bytes sitting in SharePoint and actually some a change in the way that you're you're operating so there's a a a number of key stakeholders that I always call out that you should become very friendly with and one of them and I say this kiddingly because I know you're not supposed to do this but take your auditor to lunch what I mean there is audit has to be data management's best friend what audit enables is a little bit of teeth behind your policies right but also not just at the end but invite an audit perspective into your data program and into your data strategy auditors think differently they think in ways that you know you show artifacts of evidence and you show you demonstrate progress and things along those lines so so important that a program be tightly aligned to it now that doesn't mean audit won't audit me right if I'm the chief data officer and and they should but I can use their expertise in ensuring that the lines of business the HR department the you know the finance department and so forth that they follow the policies that are established in my data program and it doesn't just become what I call a coffee table book you know sits there and just people thumb through when they come and visit it's got to be part of your day-to-day activity so audit is a critical part of the other and I've used this in in presentations I'll show a picture of the chief data officer's office and it's a chair that's empty and the point being get out of the office you know go talk to your in essence the data function is almost like an internal consulting group you know go talk to the business go talk to risk go talk to the other parts of your organization because you have to learn what they do and what their needs are and make sure you're providing that asset of information to them so in general you're a consultant and partner to your stakeholders and again having that audit expertise is so important uh so that it as it becomes a implemented BAU and not just as you know our strategy sits in the corner so sitting there in your your own office with the with a door shut is is a mistake what are the kinds of mistakes you're seeing organizations making and and how can they address them oh wow that that's an unfair question since I was one of the first chief data officers and I probably made more mistakes I don't over promise one of the things that especially new chief data officers have a habit of doing is they want to please everybody so they say yes to everything uh I've seen it happen where the chief data officer becomes the lightning rod of every problem in the organization um you could be the leader you know from a you know corporate leadership perspective but you have to be able to bring in the other players to help resolve those issues see data is kind of funny um when you think about data quality for example and people will complain about data quality in many cases it's not the data but it was the business process that created that data right so uh you know a mistake that data manager might make is to take sole responsibility for the quality of the data in the database and not look back to the business processes so documenting business process is just as important as the data itself so I guess in short time it's it's don't try to internalize every problem in the organization say yes to everything bring be humble enough to go out to different groups and and solicit their their contributions and help and then create that environment where everybody's involved in in the game you know it's it takes a village story right I mean you know you're not going to solve old data problems yourself it's it involves everybody in an organization everybody in essence touches data and they should all be part of the solution so flipping that now back to the the positive side for other than mistakes how about a uh a success story perhaps even with decam being involved yeah so I mean I'm happy to say there's been a lot um just off the top of mind one one example we had a member um a financial institution um this is a couple of years back when you know the scrutiny of the regulators was still pretty you know pretty strong um one of the things that we have always said the use of decam becomes a great tool in a regulatory audit and why is that the case because what decam does in providing those capabilities it creates a roadmap for an organization and it enables you to then communicate with your auditor or with your regulator these are the things I'm doing these are the things I'm not this is our plan to or mediate the things I'm not and here's my strategy that is so different than being on the on my heels defensive when an auditor comes in how come you're not doing this oh we didn't do that so it turns it around it creates a common language between the uh you know the institution and the auditor might be now in this particular example we got a call from this this member and they said we just had our regulatory audit and they it went fantastic uh they were hugely happy that we were using a standard model they completely adhered to our strategy and approach using the model and you know no pens were taken out of pockets to write MRAs or audit issues now that didn't mean that they're not going to come back in six months to make sure that we you know follow up on the things we were doing but it it really created a foundation for that organization to talk with with confidence to their auditors and their regulators and it helped them you know avoid any kind of negative uh perspective of what they were doing so that was one example another is again a financial institution uh in the uk have have introduced decamp at the board of directors level now at that level they basically set the organization even to senior management we're actually going to consider your compensation based on how well you adhere to these data best practice because data is driving our organization and over the course of a number of years now uh they've communicated to us tremendous success in implementing these programs uh reducing errors in you know in where bad data whether it would be you know in the operational side of clearance and settlement or you know sending out information to the wrong customer whatever it might be you know you can talk about all the things that go bad when there's data but they were able to minimize those those issues and also helped to facilitate that offensive side more insights more and analytics uh so you know there's a cause and effect here the decam gives you the foundations to make these things happen right and and we've seen absolute correlation between when you start to you know it's like anything else if I eat good food I I'm healthy right so if you if you practice good data management capability your program just continues to improve and bring benefit to the organization other than using decam what would be your top tips for uh someone starting out to create their data strategy or who are in the early days of creating it nothing other than decam is what I would say uh no I I think you know as the role of the chief data officer continues to evolve um you know people are bringing different skill sets into that role as well uh so you know I've talked to a number of people in the health care and um industry uh you know combine those elements of expertise with the elements of data so in health care where you know people are very skilled in uh patient care uh follow up analysis of you know hospital visits and things of that nature bring those two together bring your your your um industry expertise in with the data expertise and that is what that's how people see value in in data management uh yeah that's that's what we we we tell people to uh you know it's not one-dimensional right it's it's a combination of your your your um knowledge of your field mixed with data uh we're talking to an organization in Canada with with um that supports uh auditors and CPAs right the CPA now has a unique role of not only bringing their expertise in accounting but bringing their knowledge of data so you know how can they help to improve their organization's use of information even as they sit in an accounting role as as the CPA so you can see how bringing these two things together is is so critical uh to developing an insightful strategy and building a good programming um I promised earlier that we'd we'd loop back around to the cloud so cloud of course opening many possibilities you can do much much more with so much ease a whole new system up and running in a couple of clicks of the mouse but of course with great power comes great responsibility so what uh opportunities and threats do you see the cloud platforms creating that should be addressed in data strategies so uh let me stop by saying my favorite line from the spider-man series because that's where that came from right with great power comes great responsibility I'm a big big believer in that um cloud has introduced from a opportunity side I should just say tremendous opportunity and accessibility to information um you know even from a capacity perspective right you know we have more flexibility and in how we store information but over the course of the past 20 or 30 years on-site databases have become disparate and sometimes they're consistent and so forth and the last thing we want to do is lift and shift take all those you know mistakes we made over the years and drop them into cloud so we have a unique opportunity to develop an environment in these cloud implementations where we do data management right so you know minimize the duplicative data avoid you know lack of a better term you know data swaps versus data lakes you know that that that type of thing so more and more organizations doing it like I said there's almost a herd mentality everybody's got to run to cloud because it's the thing we want to do and the risk is it's it's not a overly simple environment you have to know what you're doing you have to build the right infrastructures around it and the right garden structures around it so that was the impetus really of it was Morgan Stanley and Google approached us last year and said you know we're working with different organizations and we find ourselves repeating the same best practice we'd like to reach out to you EDM council and can you help build a model similar to what you did with DKAM to give us those best practice capabilities in cloud that time went from two or three companies as I mentioned earlier so over 100 companies in a matter of a couple of months and 300 individuals engineers business people data people and so forth and think about this this was a volunteer army we tallied up the hours 45 000 hours went into buildings and to sustain that type of contribution over the course of 18 months is almost unheard of right I mean people have day jobs but everybody involved said you know what this is important for my organization and it's important for the industry to make sure that we again to use the word used before codify these capabilities into a framework that everybody can benefit for and by the way create an environment where we can certify I can now have confidence that either it's my product has certified against these capabilities or my implementation is certified against these capabilities an unbelievable effort all these individuals we built this model it's a 160 page document with advice recommendations and it covers all the things you would be concerned about in cloud ownership protection cross-border issues you know cloud could be you know jurisdictional locations become a big factor so all this was included in this model and we're really excited about it because we're seeing tremendous uptick we had four or five 600 downloads of the model in the first two weeks and what's really exciting is we see a way that the industry can can really leverage this to build really good and sound infrastructures and look we know there's risks out there right and in cyber attacks and things of that nature so we feel that this is an excellent service to industry in an environment that's just going to continue to grow in this cloud and of course all that great framework is is available downloadable for free from from your website and it is an open license I'm not an open license but a free license to the entire industry just come on to the EDM website cloud and you can download it we are partnering with a lot of different organizations consultancies who are now offering to their clients you know a facilitated assessment and if you think about other types of certifications missed for security or COVID for operations this is falling into that category SOC2 these are all you know ways that we can ensure that we're doing things the right way and and we're seeing the interest in in CDMC getting to that level where you know if you get that badge and you now sit in front of your auditor you sit in front of the regulator or you sit in front of the public and say you know we have a certified environment where your personal information we're ensuring we're doing the absolute best in securing that data you can see how that benefits everybody on both sides of you know the the marketing table if you will the consumers the producers and the and the regulators yes a great way indeed to to not make the same mistakes when we're when we're running off into the cloud as you said and we'll include links to in the episode description to to the frameworks and and your website before we sign off our next guest is going to be Jonas Blumquist who's the head of analytics and data at Scandinavian Airlines and we'll be speaking with him about their experience of migrating to the cloud what kind of questions would you like to ask from Jonas I got a kind of a kick out of the fact that you're meeting with an airline because so even though I look extremely young Tom I've been around a while and I go back to the days of computer time sharing which in essence was cloud back then it was you know I would rent space on a computer at a large organization and I would do my work off of that so almost kind of the precursor to cloud one of the biggest benefactors of computer time sharing was the airline industry as they moved their their operations their scheduling you know their their passenger manifests and all that to this this this capability called computer time sharing so it's funny it seems like we've gone complete around in circle where the the early innovators of that type of capability airlines you're going to be talking about cloud my opinion they invented this stuff back you know 30 40 years ago so I would say to your guest remind them of the fact that they were the leaders in this space and and we welcome them into the cloud environment now so that would be a hopefully a great conversation you have with with that individual planning your data strategy get out of your office go talk to the business management technical team auditors and the other parts of your organization but don't leave it there you need to put your findings together gear your strategy to the audience who's listening you don't have to reinvent the wheel when it comes to data management models when you bring in best practices and combine them with your organization's individual needs you're able to create a powerful tool for change and growth not just a coffee table book no one really reads did you know that Tieto every is also a proud member of EDM council so if you want to know more about how to assess your capabilities learn more on these frameworks reach out to me