The Gartner 2021 Leadership Vision for Data & Analytics Leaders Webinar Q&A

Share on facebook
Share on twitter
Share on linkedin
On January 4th I had the pleasure of hosting a webinar.  It was titled, The Gartner 2021 Leadership Vision for Data & Analytics Leaders.  This was in fact an overview of a piece of research we update every year.  This was for the Chief Data Officer, or head of data and analytics.  It is meant to be a desk-reference for that role for 2021.  Gartner also published the same piece of research for other roles, such as Application and Software Engineering.  What is unique about the D&A Leadership Vision is that it crossed over into business since for many organizations, the CDO reports into the CEO or COO (as examples).  So the material is not designed for IT  - but spans business and technology. The fill report is here: Leadership Vision for 2021: Data and Analytics. The webinar was very popular and I was not able to respond to all the questions during the live recording.  You can of course listen and watch the webinar from this link.  But below I copied all the questions we received  and I post a quick response.  Hopefully this helps, and I hope you enjoy/enjoyed the webinar. Here are the questions as they were tracked through the webinar.  I hope you can find your question. We are creating a new D&A department , what should be our approach towards other departments ? ex : we help you to improve your performances ! ? What you think - it's the main drive for D&A. As with any good consulting response, “it depends.” It really does. As the three perspectives in the deck suggest, your positioning will depend both on how D&A is perceived today and how you believe it should be in the future. In general, though offer a ‘improve performance’ might be a tad too general. It might be more productive to target very specific outcomes and try or pilot something. With a success behind you, sell that experience as the kind of benefit you can help improve. What is your vision for D&A for small and medium enterprises? Do you recommend a consulting approach strategy rather than a CDO strategy? We have specific research for midsize and small enterprises. In summary the message and focus should be the same: outcomes are primary. But the specifics may change based in size, industry and maturity. But as I said during the webinar, we should not get caught up with the CDO thing. You have need or must have a CDO to add business value from D&A. See 3 Questions That Midsize Enterprises Should Ask About Data and Analytics and have an inquiry with Alan Duncan. Which industry, sector moves fast and successful with data-driven? Government, Finance, ... Tough question...mostly as it’s hard to determine which industry due to different uses and needs of D&A. In general, I would offer that information-based industries tend to be more advanced than non-information-based industries. As such banking, finance, insurance and media are good examples of information-based industries compared to manufacturing, retail, and so on. Interestingly public sector might be low/behind but in pockets there is very advanced use cases. How do you think Technology Business Management plays into this strategy? I am sorry I don’t know what that is. Please drop a comment or email me and I will respond. Where does the Data Architect role fits in the Operational Model ? Assuming a data architect helps model and guide and assist D&A then they play a key role. More importantly the work is important; the label of the person helping is not so important. See Tool: Sample Job Description for the Role of Information Architect and have an inquiry with Guido De Simoni. What's your view in situation where the IT function still reports to CFO (Finance Director)? The ICT organization we have has more than one head and is run in some kind of 'divide and rule' style. Does this promote efficiency? Tricky but would you believe not that unique a question? Many organizations operate with a fragmented structure and this remains a longstanding challenge. More visibility; more transparency; more collaboration may help. Certainly, more communication across the silos will help. The secret, in my view, is with the outcomes and how each silo behaves in response to them. Understand the outcomes and you can predict better and understand behavior. "Garbage in - Garbage out". How can we make sure we are capturing valuable data? Ooo good question. Would really like to explore this one in debate. GI-GO of course we all know. But we also know not all data is equal, and not all data is equally valuable. Some data is more a risk than valuable. Additionally, the value of data may change, and our own personal judgement of the the same data and its value may differ. So, given these and other variables, should we try to capture all the value? Can we? Should we? Is that a pathway to succeed? How do we define success? I am sorry but this is a deep question. It has no simple or single answer. How does CDO overlap with Market Research functions? Do Qual data sources for example text, which tend to live in MR cross into the CDO world? I suspect some of our analysts who cover market research would have insight here. In my experience there does not have to be a gap or conflict here. For example, it is possible the CDO is the head of Marketing. What I mean to say is that the executive leader who leads marketing and/or marketing research might take on the role of CDO for the organization. If not, we should also recognize that having a CDO does equate to a centralized or equate to bureaucratic control of data. So, a market research function can be supported by a CDO however the organization is structured. This is a good thread to explore since every organization will various elements of D&A work across the organization. What is your view on data sovereignty? when some business has moved data centers offshore This question touches on legal and regulatory controls. These are fluid and change quite quickly. You only must look at how the US and Europe continue to explore changes in where data can reside and be processed. As such any Data and Analytics strategy needs to incorporate data sovereignty as per of its D&A governance program. Finance background - what would be the key technical skill to develop? Coding skills - SQL, Python or application familiarity - ETL & visualization? Sorry not sure if this pertains to the role of CDO or some other role within the context of D&A. Please email me and I will see if I can help and even point you to some research. What resources are available to assist with getting D&A Leadership Buy-in & Support and Deepening Adoption across the organization? We have a lot of research in this area as this is a common area of exploration and query from our clients. Here are a few research pieces I hope you can explore. The titles will help identify the various angles of attack here: How CDOs Can Use Data Storytelling to Engage and Influence Stakeholders The Keys to Achieving a Successful CFO-CDO Business Partnership CDO Success Factors: Culture Hacks to Create a Data-Driven Enterprise What are you seeing as the differences between a Chief Analytics Officer and the Chief Data Officer? In a nutshell a CDO’s responsibility may focus or span across the following (as an example): Strategy Governance Value Management or monetization Risk Management (most likely within context of governance) Product Management Architecture Note: Delivery of data, analytics solutions and the sustainment of technology, data and services is a question. This is where most often conflict with the CIO takes place or at least boundary overlaps are spotted. In any sizable organization each of these elements may have their own leader. As such a head of analytics, BI and data science may emerge. CAO may well be a name for that role. But the use of ‘Chief’ should be deliberate in that it should align to executive authority and P&L. That’s the idea. The real world is more complex. We have seen firms with a CDO and. CAO, and even where the CAO reports into a different organization. These are few, thankfully, as they tend to lead to organizational confusion. In all cases it might be better to focus on the work that needs to be done (see list above) and only worry about the title afterward. See Survey Analysis: Fifth Annual CDO Survey — Growth Must Continue in Order to Achieve Real Impact and have an inquiry with Deb Logan.  The sixth survey is complete and we are working on the results right now. When does a CDO (Data) move to become CDO (Digital)? Do you see these as one role or two in 2021? The trend is the reverse. Historically the digital officer has taken on the role of change agent. Once the digital strategy is in train, more than vision and change agency is needed: fulfillment now needs to be digitalized. As such the long term trend is that the role of digital officer will likely wane and the work itself will morph both into marketing, operations and also data officer as fulfillment becomes part of the digital strategy and the change itself has become accepted and evolved. In many ways the digital officer role seems to be following the trajectory of the previous head of e-business. How can one get hold of the Tool Kits? All the toolkits are mentioned in the appendix slides - strategic and tactical. These can be found on our website using the names of the notes on those two slides. Are you anticipating continued separation of "BI/Analytics" teams from "Data Science" teams or are those roles merging in the years ahead? Our Content Leader in this space, Carlie Idoine, is working on this as we speak. I don’t know if we have a view yet. The data suggests several things: The work of traditional analytics and BI continues towards democratization in the business unit directly, we call this domain analytics in our research, part of domain D&A. The work of data science is more tied to machine learning and so AI and those projects do not focus only on analysis but also automation. Many data science labs are set up as shared services. Where the two report into can also be split; ideally both would still report into a common leader though such as the CDO. Technology vendors are falling over themselves to acquire each other across the analytics space, as well as the data management space, and the D&A governance space. We call all these worlds collide in our research, but it is bigger than just analytics, BI and data science. It is across all D&A. Thanks, Andrew, for the briefing. The question that I've - how do we mind shift the thinking for the organization to go on D&A from the traditional way of managing their assets. Good question. Not one with one, single response though. Becoming data-driven is not simple. But I do believe it is though to be so hard that firms don’t get the idea you can start with small steps. There are in fact many different steps and tactics you can try. Some of those in our research highlight: Storytelling. Do a pilot and see what happens? Try some gamification? This would be part of a Data Literacy program. Fusion teams. Try a little DataOps or ModelOps and see if a small, skunks-like project fan turn some folks on. Decision modeling (one of my favorites). Explore in dialogue decisions and outcomes rather than focus on data and analytics asked for. This can expose new angles of attack on the business challenge. See Roadmap for Data Literacy and Data-Driven Business Transformation: A Gartner Trend Insight Report and also The Future of Data and Analytics: Reengineering the Decision, 2025. Will you (Gartner) do some research publication/webinars on those three top challenges? measuring value, prioritizing (where to start), and data literacy? Great idea. I think some of our earlier webinars touch on these. But I like the idea of a program here. I will suggest this. Thanks for the tip. See recorded webinars: Emerging Practices for a Data-driven Strategy. Data and Analytics Governance: Whats Broken, and What We Need To Do To Fix It Link Data to Business Outcomes Does Data warehouse as a software tool will play role in future of Data & Analytics strategy? Yes. You cannot get away from a formalized delivery capability focused on regular, scheduled, structured and reasonably governed data. Data lakes don’t offer this nor should they. They have a different sweet spot. Any recommendations on influencing the leadership/executives to adopt the D&A DNA? I find that a good dose of D&A-impacted business results is the only effort that seems to most of the time. A question above led to a few pilots and projects that may help expose the value. Remember, it’s not about how many records were cleaned up or how many dashboards were generated, it’s about how much of an impact on the outcome the worm of D&A has that counts. What are the new trends around the Data solution architecture (centralized vs de-centralized?) Not sure I understand this one. From a D&A organization perspective the future is hybrid: we need to locate the work of D&A in the right place. I guess that means that solution architecture also needs to respect and reflect this? Try this: Tie Your Data and Analytics Initiatives to Stakeholders and Their Business Goals. Hello. Very interesting. Thank you so much. What do you recommend to small companies of 50 or 100 persons? We publish research for small organizations, not just larger organizations. Alan Duncan leads our research here. From where I sit an embarrassing number of modern approaches and best practices are applicable to all. In fact, I might even say that the best best practices are those defined for smaller firms; and these should be universally used. Larger firms should use the same, and add the bits they uniquely need. See 3 Questions That Midsize Enterprises Should Ask About Data and Analytics and have an inquiry with Alan Duncan. This is not how it works though. In our economy the larger firms invent their practices and the market looks to them as leaders, because they are largest. This is not the history of innovation. This is a personal area of interest to me: how ideas and IP diffuse across industry. Large and small firms play a role here. Concerning artificial data generation: Does this mean, that simulation models are going to be even more important? Yes! And not just for synthetic data techniques. Simulation in general is coming back in vogue due to continued interest and innovation in and around the AI space. I am reminded of my days when I was in supply chain and used to work with simulation software tools. For a time, I believed simulation was more useful a capability than optimization, at the time that larger firms were seeking optimization solutions. Do you play SimCity? Imagine running our businesses like that? Would you say that confidentiality of information is casing a problematic barrier? Yes. Well, it is and will remain a challenge. But a new and growing innovation is that some firms are exploring how to bring together the various silos and threads of privacy, security, retention, quality, and so on. So, confidentiality will be included, respected and reconciled with D&A governance. Do you have an example of how an organization improved data literacy in a really practical useful way? Yes, and no. We do have good examples and bad examples. Storytelling is a nice one to use early on to test the approach. But we are seeing increasing data suggesting that broad and bland data literacy programs, for example statistics certifying all employees of a firm, do not actually lead to the desired change. New data suggests that pinpoint or targeted efforts are likely to be more effective. See: Tool: A Living Library of Real-World Data and Analytics Use Cases. Where do you see the data ethics in 2021 challenges? We have written on data ethics quite a bit. It remains a growing are of interest for our clients. Overall, I do see a growing connection to D&A governance to help align the effort, to reduce the tendency to create yet another silo. See: Data Ethics Enables Business Value. Which are the key skills for a CDO? Do you think it could have an IT background ? Yes indeed. However, I as mentioned on the webinar, the skills most often lacking are on the business side. Empathy, communication and engagement skills are critical. More often the technical or IT skills are not lacking as much since they tend to be widely available in the IT organization. In cases where this is not the case the yes, a CDO does need to have some technical awareness; they tend not to be the hard-core IT lead. Do you see AI as being a sort of 'micro-AI' where individual moments are automated through learning-by-example (e.g. improving quality at a factory process) vs. 'macro-AI' where companies try to redesign their entire experiences around AI? Interesting question and terminology. Certainly, organizations seeking to get more value from AI seek processes that can be automated. Not all processes are lined up like that. We are therefore seeing a return of business process re-design skills to help reorganize how work is executed, making it easier to drive mass or hyper automation. So, I think I am affirming your implication though it will be hard to clearly articulate where a bottom-up and top-down approach converge. Where do you see that the Data and Analytics ownership belongs in an already digital business ?- such as SaaS or other tech companies. Not sure I get the question, sorry. D&A is about helping business leaders make better, smarter decisions. All businesses must take decisions, irrespective of where they operate or how they operate. It might be that digital businesses, what we might call data-based firms such as media, use data and analytics more easily than non-data-based firms. But I am not sure if this is what you mean. See The Future of Data and Analytics: Reengineering the Decision, 2025. You mentioned a few times that most enterprises are not good at data governance. What would you say the main reasons are for that? As I tried to explain on the call the way we all view D&A is changing. It is much less just about defensive controls and more about enabling capabilities. As such they could well be the first ‘gap’. Second, too many folks confuse means with ends. Governing data is not the end. What we get from the work should enable the ends. I think that is the second ‘gap’. See recorded webinar: Effective Data and Analytics Governance - Finally! Do you have any advice specific to health care organizations? (Their data systems must support two functions, official "legal" health record governed by HIPAA considerations, AND advanced analytics on a large, complex, and "messy" data set. For clinical decision support in a learning system, these two views of the same data must be reconciled.) We have analysts and others who focus on healthcare who can help. One of our team, Lydia Clougherty Jones, wrote a note concerning how to comply with, as an example, GDPR but still keep the data to power analytics. So, it might be that this and other research we have can help if applied in specific situations. How does the role of Chief Digital Officer intersect with Chief Data officer? Please see previous answer concerning evolution of the digital and data role. Regarding my question re: fit of CDO within organization, you answered it on slide 19. Thank you! What is your view on Data Monetization? As noted in the call this is critical, and a common responsibility for several CDOs. Of course, we need to clear here that for some folk’s monetization means selling data for real money; for others it means value accounting. The former is easier to agree and define and it also focuses on markets and identifying needs for information products. The latter is harder to define, at least broadly and even publicly, and it focuses on all data inside a firm. Both are very important for different reasons. See: Use Infonomics to Quantify Data Monetization Risks and Establish a Data Security Budget. How does this D&A model fit within a project management model because I believe that development and execution of initiatives should be managed using the relevant PM methodology? Totally agree. In our modern data and analytics strategy and operating model, a PM methodology plays a key enabling role in delivering solutions. Out operating model respects programs, projects, and product management methodologies. Do you draw a distinction between a data-driven vision and a data-enabled vision, and if so, what is that distinction? I don’t. I didn’t mean to imply this. It might have been a slip of the tongue. In the world of data and analytics strategy, there are differences in approaches when data is seen as a driver, an enabler or a utility. So, if you were to define a different named strategy using this framework, I think you could argue there is a time when the terms do mean different things. I find most folks say, ‘data driven’ and they leave the details to later/someone else. How does size and maturity of organization impact fit (reporting structure) of the CDO within an organization? We have researched this extensively. We tend to find too many folks seek clarity on who to hire before they have a clear idea for what work is needed. As such the answer to this one is, ‘it depends’. The organization, work, its location and leadership, will all depend on what is needed, in lieu of not knowing, a single person could initiate the change. And that single person does not have to be the CDO. But it would be more effective :). Not sure I answered your question. We live in a global environment in the use of data with laws & regulations to comply on data security & privacy, it's complex, what are the best practices to manage this in a proactive & predictive way to minimize risk? Saul Judah is our main person focusing on D&A risk management. What are the key KRMs on Data & Analytics that needs to be reported to board? I am not totally sure I know what you mean with KRM but I believe you mean Key Risk Management indicators. As noted above Saul is your man here. I believe the focus would align to governance policies that are in focus: quality, privacy, security, retention, access and ethics, etc. As with offensive policies, too many firms mistake hygiene metrics such as the number of records cleaned up versus the impact on outcomes as the measure of success, with risk I would wary of the same mistake. We currently do not have a CDO, how critical is this? Hopefully my answer on the webinar covered this. But no, it is not critical that you do not have a CDO. You can start without one, but it will be easier if a CDO is established eventually. I believe I quoted myself on the call: “Every business leader is a CDO in waiting.” What sort of education/familiarity should management pursue to be able to effectively engage with the D&A team, especially when they themselves do not have technical background? Is the manager who doesn't have such a background a dinosaur? This overlaps with similar questions. D&A is not a technology thought it does use technology. The most important education/familiarity (as you ask it) in my view is to understand or be intrigued with how humans take decisions; a desire to learn how we think and behave; and a working knowledge for how a business outcome, process and data interact. Any technology skill is a plus, but does not outweigh the former skills. I don’t mean to ‘put down’ technology skills; I simply want to call out what I thunk is the more valuable of those on offer. Also see The Future of Data and Analytics: Reengineering the Decision, 2025. Data literacy required for non-IT folks in different departments wearing different hats may be different beyond the common denominator. How does a company identify ,plan and deliver the training programs catered to these different roles ? Great question. I believe this and my answers fit at the intersection of those previously asked/answered: Bland or kitchen sink approaches are easy to understand, even sell, but they don’t seem to be delivering sustained benefits Targeted approaches, oriented around specific opportunities, seems to be more reliable As such your answer lays in the specificity of the tattered approach, what ever it is. Thanks for the overview Andrew. On end user clients calls, are you hearing a greater focus on use cases and greater need for prescriptive analytics, ex marketing analytics, sales analytics, healthcare, etc. where performance and data quality is imperative? Thanks. Yes, prescriptive and predictive analytics remain very popular with clients. See When and How to Combine Predictive and Prescriptive Techniques to Solve Business Problems. Data Governance has not only D&A and software angle but has a lot of process angle. I see that it is hard to put that in practice. Do you agree? Which tools would you recommend for getting governance in place? Yes, there is a people, process, data and technology angle to D&A Governance. Tools there are a plenty. Since much of the work is siloed, there are entire markets focused on, for example, data privacy tools, data security tools, data quality tools and more. We cannot of course forget metadata management tools, of which there are many different. A few years ago, we developed what has proven to be a very simple and helpful framework to relate all the parts. The work of D&A falls into three buckets: Policy setting Policy enforcement Policy execution Almost all technology helps with the execution part. Just think of all the rules and policies secured in your business applications, data warehouses, and data security apps. The enforcement was often left to IT who monitored the systems. This is short sighted, more importantly is business monitoring of the same thing that highlight when business actions, work or decisions are held hostage to data problems. Lastly the policy setting is a real challenge. This falls under authority and accountability, which should be business leadership. Long term we see the organizational silos converging for mutual and converged policy setting and more efficient and business driven enforcement (we call this stewardship and called out new solutions for this already). Ultimately the silos technology markets will be forced to converge. That will take the longest, but we might see some enterprising vendors with unified, business oriented UX platforms emerge first. See recorded webinar: Effective Data and Analytics Governance - Finally! It is not that Data Governance was not there earlier, but it is gaining importance only now. Do you think this is only because of Regulatory Compliance and Heavy Penalties or is it more than that? Yes, I think that is part of it. But the growing recognition that mission success is more reliant on data outside the traditional boundaries and control of the organization is also part of it. To some, say intelligence services, that was always the case. But for them, big data evolved into all data and all formats. With Covid, what was chronic has become acute. Hence the growth in demand for CDOs and the focus away from data and analytics toward decision making. Governance really needs a culture shift in the organization, how to you approach improving this? Education is one way. Yes indeed, education can help. But I would off that if we take a modern approach to governance, and focus on outcomes, the changes you refer to should naturally start to take place. If I may say this: When selling the work of D&A governance, we need to change: ‘It’ goes on every day in every company. It’s called citizen stewardship and takes place when anyone simply overrides the formal process and/or fixes data to get their job done At the end of the day, modern D&A governance leads to less work, not more You already have the knowledge you need inside your business. You just may not know the questions to ask to get that knowledge public Don’t emphasize data, data standards, or governance meetings. Focus first on outcomes and where those outcomes and data-besieged challenges are discussed. In other words, in business meetings. See recorded webinar: Effective Data and Analytics Governance - Finally! What advice do you have for college students & young professionals who aspire to pursue a career in D& A and how can they keep up with the rapidly changing trends? I think there are so many opportunities here. A technical track, so to speak, would focus on the methods and techniques and tools evolving. Many university opportunities here can help. I personally would look for vocational opportunities since the business knowledge would be the easier of the two sides to develop. But that would need business opportunities. I suppose I would try a balanced approach. Try to keep up with many fronts ‘just enough’, with a leaning toward business over technology, but I am sure that anyone that goes deep and specializes can be successful also. The CDO office seems to be still evolving. In your experience, organizationally where should the CDO org sit? (I have seen Data teams sitting under IT, Finance, CEOs and so on.) We see both too. Our most recent sixth annual CDO survey (interestingly the longest running CDO survey, I think) reinforced this split. In the last couple of years, we have seen a closely even split between reporting into CIOIT and business/CEO/COO. However, of the two, the one reporting into business seems to report better business impact. That may well be tied to how IT is classically managed: as a cost center. When reporting into business D&A may well be seen (correctly, in my view) as a P&L/investment resource. You mention in one of your earlier slides (bar chart) that AI will be a huge trend. Specifically, what industries will be most impacted? We have written on this widely.  One related item that come sto mind is this: Toolkit: Choose and Rank AI Use Cases With This Workshop.  This is a fast moving space so I would suggest a follow up inquiry with one of our AI team members. Define generative ai Here you go: “Generative AI is a variety of ML methods that learn a representation of artifacts from the data, and use it to generate brand-new, completely original, realistic artifacts that preserve a likeness to the training data, but do not repeat it. Generative AI can produce novel content (images, video, music, speech, text — even in combination), improve or alter existing content and create new data elements.” As defined in: Hype Cycle for Artificial Intelligence, 2020. Our operational teams are dissuaded from embracing D&A pilot programs into their operations because the ROI is so small. This is particularly acute when current operations are successful. "Why should we adopt this new stuff, when we're going just fine the way we are." How do we overcome this resistance? Very interesting question. Are the current practices not using D&A at all? Or is it that they are using old legacy D&A? If this is seen to generate value, new approaches can be problematic. Unless you can generate fear from a potential loss of leadership in the future, or if you cannot appeal to the opportunity cost of not making the change, you are probably sunk. Simulation techniques may help. Show the value created with method A compared to method B. If leadership wants even more success, maybe they will change? It all comes down to the motivations of your currently successful leaders.... I am also of the same opinion, as you said in slide 7, about the over-hype around AI. But I would prefer to come to a conclusion with some data. Is there any data that can show the success of AI in enterprises? We have published some case studies. But overall, AI is not something you can acquire and assume success. Case studies show that AI can be useful if applied correctly to very specific problems or opportunities. And as you see in the press every day, you are not guaranteed success if you apply a model to a similar problem, even if it was trained on what appears as related data. For more accurate decision making, forecast and predictions are needed. How this topic (planning, budgeting, forecasting) can feed into a data & analytics strategy and platform? I believe I mentioned on the call an old idea, corporate performance management, that has resurfaced as financial planning and analysis. The former was something finance and IT tried to implement; the latter is today led out of finance. These should be part of D&A. What finance department does not want to leverage prescriptive and predictive techniques to augment their analysis for leadership? Perhaps if FP&A is doing well tHe CFO could take on the CDO role to expand the focus across the business? Does a CDO need to report to the information owner of the data under management/governance in order to be effective? (I think that in large complex organizations, the answer to that determines how many CDOs you need, and at what levels.) Forgive me but I am not sure why a CDO should report to a data owner. It might be we need to agree the scope of the role, and what does data ownership mean. For example, the business owns its own data. There is a leader that owns the funding and cash: CFO. There is a leader that owns people assets: Chief of HR. There is a leader that owns technology: CIO. Why not an owner of data? I noted in a previous response: there can only be on Chief. So, what is the boundary? I’d say it was the balance sheet or P&L boundary for which they are responsible. Thus, every business leader is a CDO in waiting (I said this above too). But if that leads to conflict and loss of synergy, assign one CDO to operate on behalf of all executive leaders. Does this help? In your study did you assess to what extent CEOs are more inclined to acquire a startup to respond to their lack of knowledge on the D&A, or to apply for a Saas instead of building by their own an internal solution and team? I am not sure of the angle here, but we do research the methods and degree to which organizations acquire D&A competencies and capability. Those CDOs that report success tend to be more business oriented over technology; and to a lesser degree have inherent knowledge of the firm itself versus an outsider. This is most likely due to the previous comment I made about empathy, communications and engagement practices. SaaS will always play a part somewhere in the D&A environment. Putting Data and Analytics together implicitly assumes that data concerns (only) analytics, while data concerns every single aspect of a business. What is your view on this? if hypothetically we were not be allowed to run analytics on our data, would this mean that we don't need data anymore? Obviously no, so why data is only linked to analytics? Great point. For Gartner and in our research, D&A does mean all uses of data as well as the subset related to analytics, so for things like integration and governance, we span operational and Analytics use cases. It would be wholly wrong to assume or equate analytics, BI and data science as equivalent to D&A. Some vendors, and some folks with an outdated mindset, do this. They refer to their D&A platform and ‘all it is’ is the Analytics, BI and data science platform. Data management and D&A governance also have their own platform supporting all uses cases, including analytics, BI and data science. Look at this note from 2019 where we tried to call out the constituent parts: How to Make Data and Analytics Central to Your Digital Transformation Initiative. Can you expand on the idea of a data literacy program as an opportunity for D&A perceived as consultative? Data literacy is a set of techniques to help organizations get more value from data and analytics. We have written on this quite a bit. Any of these approaches can be used in a consultative manner. That manner by which we communicate, sell and set up the work differs from the service or leadership mindset, but the technique itself may not. See: How to Design an Effective Data and Analytics Training Program to Improve Data Literacy. It seems to me that health care organizations of medium to high complexity must have a "hybrid" approach to D&A, with some centralized services and some decentralized capabilities. In my experience, it seems the best way to manage the necessary standardization is to centrally create and manage an integrated data platform, but then allow intermediate management organizations to customize their own dashboards and tools on that platform. Would you agree? Yes. Nice example of a practice framework. All I would offer/add here is this: Seek local or edge work for speed, innovation and deep domain needs Seek central work for robust, methodical, efficient, and widely reused needs Seek shared work for a balanced approach And ask where for each of the following discretely: Traditional Analytics and BI Advanced analytics and data science D&A governance Data management See Where to Organize the Work of Data and Analytics. How do we verify ROI on data and analytics efforts? Often this department helps organization to get things quicker, is time a factor in the context of ROI? The real end is the impact on business outcomes. Here is an example used form D&A governance: Try to estimate or simulate the performance with ‘bad data’ and then good data. Include in the former assumptions and costs for coping with the data work The difference in impact or profit is the impact More generally use Value Stream Mapping as a method to explore various before and setter situations. For things like data science labs there is no ROI. We would suggest using a different approach to sell and create value, such as pilots and time-boxing. Do you see new organizational structures evolving blending business and IT? currently we have a business side and an IT side and myself as the Senior Engineer blend the 2, barely. Yes, we do. We have been writing about fusion teams (cross disciplinary) and XOps (methods applying DevOps to D&A such as DataOps, ModelOps etc.) focused on team make-up and structure respectively. They seem to be very productive. See Demystifying XOps: DataOps, MLOps, ModelOps, AIOps and Platform Ops for AI. If at number one Analytics the preferred, data would be needed in one repo [preferably], otherwise you don't obtain live data updates. E.g. Data Lakes in Azure - as SaaS. V if data is stored in multiple db's on prem - I have seen spiders webs forming and live data/problems being highlighted is impossible. So why is Cloud computing not rated with Analytics, they go hand in hand? Sorry not sure I get this question. Cloud and cloud computing are just a location and elastic compute capability. Data does reside in many different clouds, and data, and separately analytic workloads are also processed in many different clouds. Cloud can help. It is not a panacea. In fact, it is also a greater pain for D&A governance... How have those Board of Director survey numbers changed over time? Analytics and AI have been top for last two years, I seem to remember. Can we share this recording once it is published? I believe the URL will be public so yes. Thanks no wonder my job is so difficult with people like you around Sorry I am not sure that is a compliment or a criticism that I over complicated the topic. Sorry if I did. It is a huge topic, so I hope I was able to touch on enough aspects to make the webinar interesting.  

Source: Gartner Blog Network On:

Read On

This post was originally published on this site

Leave a Comment