Kisha Gulley was once kicked out of a Facebook group for mothers with autistic children after a c...Read More
On Thursday January 6th I hosted Gartner’s 2022 Leadership Vision for Data and Analytics webinar. Here is the link to the replay, in case you are interested.
There were 80 or so questions or comments posted and I was not able to respond to all of them live in the webinar so here are the verbatim questions and an individual response to each on. I hope they are helpful. Enjoy!
The questions are listed in order they were tracked in the QA tool. I try to relate as much published research as I can in the time available to draft a response.
Which trends do you see for 2022 in AI ML technology and tools and tool capabilities?
– In the webinar and Leadership Vision deck for Data and Analytics we called out AI engineering as a big trend. I would take a look at our Top Trends for Data and Analytics 2021 for additional AI, ML and related trends. We will publish a new Top Trends for DA for 2022 in a couple of months. Keep a watch on Gartner.com. You could also have an inquiry with Anthony Mullen who leads our AI research.
Simulation is a very effective way of generating synthetic data
– Yes indeed. Here is a recent note on synthetic data: Maverick* Research: Forget About Your Real Data — Synthetic Data Is the Future of AI.
What are your thoughts on a data mesh vs. data fabric in large heterogeneous environments?
– This is a question for our data management team. Try Mark Beyer . I believe that the two ideas may have originated at different times and maybe in different places. But in general I believe our tam would suggest that a data mesh is like a data fabric but without the self-learning add. Data mesh seems to mean a network of systems and hubs connected to help cope with silos. Perhaps a data mesh is what a pure architecture might design. A data fabric has the same model but it adds a set of capabilities that continuously monitor the whole landscape, and continuously update the intelligent part that infers insight as new data is discovered and/or added. This insight is then made available to other uses such as application integration, governance, analytics and so on. So in summary they are very similar concepts but data fabric seems to be the more rounded of the two. Again, I’d check with Mark Beyer.
Where can I get more information on Data Fabric?
– Here is a very popular data fabric resources: Data Fabric Architecture is Key to Modernizing Data Management and Integration.
How is Data Analytics systems doing on-Premise in comparison to on-Cloud – specifically with respect to data security?
– This is a question for our data security team. Try Jay Heiser who leads our data security research. As to broader data and analytics, literally every capability you might have sourced on premises would need to be re-evaluated for cloud options.
In Latin America, especially in Peru, AI is still at the level of large corporations. I would like to know if there are organizations to promote AI at the level of the Central Government and Small companies. Thanks !
– I am sorry but I am not aware of this myself. But I would search on our web site for AI and mid-size enterprise; or ask Anthony Mullen who leads our AI research in our data and analytics team. I am sure he can point you to the analysts that cover this.
Can you remind where we can find the mentioned blog?
– If you are reading this, you have found it
How D A , AI would help to reduce Net zero emission or decarbonization ?
– This is a very broad question with many angles and I will only be able to help identify some of the angles here. First, how we measure emissions and carbon footprint is about data design and policy. For example I would argue that most organizations that report their carbon footrest are not doing it consistently and nor are they doing it correctly. So much hype compares electric vehicles to traditional engines. Yet there is no inclusion in the conversation about the costs and issues related to the battery and materials used in the most expensive part of the EV. In other words, DA plays a key role in the foundational measuring angle. Second, DA can help inform business leaders on smarter decisions about where to make changes in their business to meet carbon or environmental goals. DA is just a capability – really – to help through how to improve decision making. This means ANY decision. I am sorry that this is not a complete answer. I would encourage you to set up an inquiry with our DA team to explore these and other angles.
Julian son pls confirm, Mark Baya pls confirm names and their topic you are breaking pls more detail Governance Atlas
– Here you go: Julian Son and his note: Composable Analytics Shapes the Future of Analytics Applications.
Mark Beyer and one of his notes: What Is Data Fabric Design?; Guido De Simoni and his note: Tool: The Gartner Data and Analytics Governance Technology Atlas.
Can you provide a link to your blog, or will we get notice of your info posted through email?
– As above, if you are here, you have the blog URL. Several questions asked about the blog URL so I will remove all that follow.
What is your thought on ecosystem play? Do you see this as an emerging trend in Data and Analytics?
– As I said in the webinar, ecosystems is a hot word. It’s “in” again. All kidding aside yes it is important to DA. The discussions we are having at this level concern things like what is a platform? Where will you invest in DA in the future? Will it be driven or controlled by a single cloud vendor of some kind? What happens to best of breed concepts? I blogged on this recently: When is a Platform? What is an Ecosystem? What is a Market?
The Business users believe IT takes too long to deliver and IT believes the Business users don’t know what they want. What suggestions do you have to bridge the gap that has existed historically and still does in many organizations between IT the Business users?
– Yes I touched on this in the webinar but I didn’t have a slide to allow me to go into this topic in any detail. We did in fact do so last year for the Leadership Vision for DA Leaders 2021. There are many tactics to employ that might help – including data literacy; storytelling; use of our value pyramid/business value model; use of our decision intelligence model. At the end of the day, we would like to shift the conversation away from a request for a report or a dashboard, toward a discussion about the business outcome the leader or team is trying to impact. That is the key. If we can shift that conversation we might work out that there is a better way to help achieve the business goals; and data and analytics becomes part of that conversation. Search for data literacy, value pyramid, business value framework, storytelling and decision intelligence to get some ideas from our web page.
I’ve seen many scenarios where an organization tries to build out a net new DA team and it eventually begins to compete with the more traditional technology delivery teams/capabilities within the organization. What’s your view on the actual delivery of DA initiatives and capabilities? Should you have a central DA team to provide guidance/oversight and leverage your traditional delivery pipelines or should you stand up an entirely new pipeline?
– This is a very topical question. We did a note to help organizations work out the most approbate place to target the work of DA: Where to Organize the Work of Data and Analytics. This research does not tell you where to do the work; it is meant to provide the questions to ask in order to work out where to target the work, spanning reporting/analytics (classic), advanced analytics and data science (lab), data management and infrastructure, and DA governance. It is designed for a CIO, CDO or DA leader.
Why lifespan of CDO is inferior at 2 years ?
– We did some early work a few years ago that look at the career path of a CDO – see from 2016 Build Your Career Path to the Chief Data Officer Role . We found anecdotal data that suggested things such as a) CDO’s with a business, more than a technical, background tend to be more effective or successful, and b) CDOs most often came from a business background, and c) those that were successful had a good chance at becoming CEO or CEO or some other CXO (but not really CIO). This was not statistic and we have not really explored this in any greater detail since. I suspect we should. Since this analysis we have manured the success of CDOs and it is true that in some regions of the world, some industries, the success rate of CDOs are variable. The practices we talk about on the webinar today should help extend that success and assure a more effective lifespan and more importantly, career path.
Could you share a good realistic skill set for a Data officer?
– We do have a job description of a CDO. Here is that and some other resources: Do You Need a Chief Data Officer? and Toolkit: Chief Data Officer Job Description. The last item should be updated in 2022.
How important is the role of the citizen data scientist, and what recommendations do you have for organizations looking to augment their data science teams by developing analysts and domain experts into this role?
– There remains some confusion in the market concerning citizen data scientists and even citizenry in general. For example, you don’t and can’t hire a citizen data scientist or any citizen role. It is not a discrete role that is sought. It is a name we use to denote that a role or practice is adopted in the wild without any formal structure or management. As such it is like finding natural forming resources when you didn’t set out to hire or build it. That being said, it is a very important role.
To drive a successful Data Analytics strategy do you think it is a multidisciplinary activity and if so, what additional roles would you expect to see involved
– A DA leader of CDO or whoever will lead the development of the DA strategy will of necessity work with business executives. They will need to since they have to explore and discover the role DA plays in the organization meeting or realizing its goals. As such a CDO does not need to have a deep technical background, but it can help. What they do need is an ability to talk in business sense about how data and analytics help improve decision making and how data (in all its forms) can be the basis of a new business model. So a wide experience helps, and one that has a business angle too. But there are also many CDOs that are technical in nature that are successful too. So I am only inferring the general idea, not the only path to success. I would look above at the question re career path and lifespan for details on the skills required of a CDO.
Is there a good map that shows the connections between data, advanced analytics, digital, innovation, etc.
– Yes and no. I would have to admit that there are few documents that talk about all the connections across any set of topics. It is really hard to maintain such things since a) the list of topics changes all the time and b) the connection between them also changes all the time. I did make an attempt recently for CIOs and CDOs. Have a look at this and see if this helps: Data, Analytics and AI Form the Foundation of Data-Driven Decision Making.
Can I book you for a team (DA analytics community) session for 30-40min – with exactly this session ?
– I hope we can help. If you have an account manager or leadership partner, check with them. Have them figure out if we can set this up for you.
What is the difference in your view between Data Analytics and Data Science?
– Just to be pedantic “data analytics” is not something we write about. We write about data and analytics. This is somewhat pedantic because if we accept that these are different, we have to explore what is data, and what is analytics? At the end of the data, analytics are data – and analytics are derived from data: everything is a form of data. So, data is one discrete thing, and analytics is different (and includes its own data). But a lot of data is not used in analytics – such as operational data. So the term “data analytics” is not precise enough – I can’t tell if you mean all data and all analytics or just analytics and its data. I wrote a blog on this: It is all in the name – what does data and analytics mean?
– With that in the clear, data science (as far as our research is concerned) is part of the analytic and BI topic, and these represent elements of the “analytics” part of DA. Data science and analytics and BI includes, increasingly, the data it needs to do its work. But that is not the “data” in DA – it is implied in the term “analytics”.
How do you think is the optimum way of governance and organization of data products and data projects and program?
This is a big question. I would suggest that one way to explore it is to explore the idea of a standard operating model. We have resources to explore this and it helps organize the people, process, data and technology for every initiative, be it a program, project or product. Every initiative has an element of governance and decision rights (as an example) and so our operating model include governance and change mgt as part of its planning. This is the same for scope, outcomes/metrics, practices, organization/roles, and technology. Check this out: The Foundation of an Effective Data and Analytics Operating Model — Presentation Materials.
Is Decision Making Maturity keeping up with the Technology/Technical Maturity?
– Interesting question. I am not sure we have a maturity model for decision making, but the power of the question makes me think that no, our collective use and understanding for improving decision making is not keeping up with technology maturity. There are many more technologies that can help; but so many organizations do not even think methodically about how we take decisions. Additionally, so much of any such understanding needs to focus on people and psychology, not even technology.
How can we practice this plan in the Department of Education in the Philippines?
– We can help. I’d love to come to the Philippines ( ) but more seriously we do have resources that can help. Here is our contact details: Please contact Sunny Hemrajani, Phone: +65 6771 3770, Fax: +65 6 333 6768, firstname.lastname@example.org
What are the business key focus areas for AI and ML in 2022
– This is similar to the first question above. Here is that response: In the webinar and Leadership Vision deck for Data and Analytics we called out AI engineering as a big trend. I would take a look at our Top Trends for Data and Analytics 2021 for additional AI, ML and related trends. We will publish a new Top Trends for DA for 2022 in a couple of months. Keep a watch on Gartner.com. You could also have an inquiry with Anthony Mullen who leads our AI research.
Link to item 6 on slide 27 is broken, https://www.gartner.com/en/documents/3987664 , for Dashboard to measure business impact, can you provide a current link?
The note is this one: Create a Dashboard to Measure the Business Impact of Data and Analytics
People Competence is such an an importance part of achieving your data vision, is there some information on the critical skillsets required to keep us on the front edge of this journey for certain roles??
– This is another very broad question, given that there are many roles involved in DA. Overall I would offer that a focus on data literacy and job descriptions for the roles you are looking at should help. We have a lot of data literacy material published and we annually update our must-have roles work. Here is the 2021 report: What Are Must-Have Roles for Data and Analytics? .
Can you give a plan, with respect to an employee working in Data and analytics field for adapting the fast change happening in the organization to stand as better contributor in this change within the organization?
– This is quite an open ended question but I would focus on decision intelligence ( Decision Intelligence Is the Near Future of Decision Making ), storytelling ( Impactful Storytelling: Become a Powerful Data Storyteller), data literacy ( A Data and Analytics Leader’s Guide to Data Literacy ), our value pyramid ( Tool: How to Connect Data and Analytics to Business Value), and business value framework ( The Gartner Business Value Model: A Framework for Measuring Business Performance ) and look at each of these areas as tools to put into your toolbox. They can all help you at any time depending not the context. There really is not one plan per se for everyone.
Most of DA concerns and activities are done within EA in the Info/Data architecture domain/phases. How should you manage this overlap with the CDO role? Are enterprise architects now being operationalized under new roles such as CTO and CDO etc?
– I remember that I tried to answer this live during the webinar. Of course, many firms don’t have EA functions or architects. So it might help to ignore the idea of a discrete resource or team, and just think about the work that needs to be done. As such, I would say that it is critical the right work is undertaken. That might be facilitated by EA but if there is no EA, that should not stop us. I also said during the webinar that the emerging CDO/office should/does have a very strong dotted line to information architecture for the reasons noted here. However, I would change that slightly here. I would say that the CDO should have a strong dotted line to EA overall. Much of the analytics architecture takes place at the solution architecture level since we should be looking at business process and decision design.
Could you precise to which complementary research you mentioned when you talked about a data governance survey ?
– Here is the one I mentioned during the webinar: The State of Data and Analytics Governance Is Worse Than You Think. This was from 2020. Here is the new survey report: The State of Data and Analytics Governance: IT Leaders Report Mission Accomplished; Business Leaders Disagree.
On the topic of “Composable DA” – are there examples / use cases where a biz has achieved this on a larger scale?
– Our Julian Son is leading our research in the application of composability for analytics apps. see his note: Composable Analytics Shapes the Future of Analytics Applications; I would check with him in an inquiry. You could also ask the Apps and Software Engineering teams as they are doing a lot with composability.
How do you see next generation data management tools (governance, quality, catalogue, integration) incorporating AI/ML evolve in 2022 compare to traditional data tools?
– We see most, if not all, of data management being augmented with ML. Much as the analytics world shifted to augmented analytics, the same is happening in data management. You can find research published on the infusion of ML in data quality, and also data catalogs, data discovery, and data integration. MDM also has picked it up though mostly via the DQ/entity resolution angle but there are some intriguing impacts that are emerging – see this note: The State of Master Data Management . MDM will never likely go away as a concert and business priority (who can survive with untrusted customer or citizen data?) but how MDM will be achieved continues to evolve.
Would you classify Data Governance as a “trend” or “challenge” with respect to this presentation
– Data (and analytics) governance remains a challenge. It is misunderstood and incorrectly positioned by many. I believe that we are going to experience a renaissance however as the number of modern best practices available that can help are still diffusing across the industry from modest beginnings. But governance as a topic is also growing in popularity due to the focus on DA, and digital business. So governance is also a trend from that perspective.
Great presentation, thank you. How can we learn more about some of these topics, for example, composable DA (a term I wasn’t familiar with), and Generative AI?
– Thank you. I would search Gartner.com for the topics of interest and also search for Gartner webinars since they are a source of some useful material too.
What are some strategies/tactics that you would recommend for elevating the value and strategy discussion in large, siloed organizations?
– I will revert to a response earlier. Any number of tactics can help turn the dialog around away from challenges and constraints such as silos to focus on the real needs. Silos are not a problem as such, they are only a problem if and when they get in the way. Figure out what you really, really need first, and then we can find practices and tools to bypass such constraints. I would focus on decision intelligence ( Decision Intelligence Is the Near Future of Decision Making ), storytelling ( Impactful Storytelling: Become a Powerful Data Storyteller), data literacy ( A Data and Analytics Leader’s Guide to Data Literacy ), our value pyramid ( Tool: How to Connect Data and Analytics to Business Value), and business value framework ( The Gartner Business Value Model: A Framework for Measuring Business Performance ) and look at each of these areas as tools to put into your toolbox. They can all help you at any time depending not the context. There really is not one plan per se for everyone.
Looking at the decision making framework , it appears a re-engineering of applications and functional grouping of modeling into capture /interpret/model /act etc. Is that how you have seen using this framework ? If yes , then in many businesses the application landscape contains many third party business applications. How will this framework handle those cases ?
The framework is tool that you can use, in a workshop setting if you will, to explore and unpack with business roles the decision and outcome they are trying to improve. As you use the model you can/will explore different angles and inputs to the decision and outcome. This gives the team time to work out where the opportunity is to improve (re-engineer) the decision making process.
elaborate on synthetic data
– I responded to this topic above with this published research note: Maverick* Research: Forget About Your Real Data — Synthetic Data Is the Future of AI . Here too is a blog ( By 2024, 60% of the data used for the development of AI and analytics projects will be synthetically generated) of mine on the topic. I suspect there is much less Maverick to synthetic data today.
What would be your definition of interoperability and to what extent would standards and semantics play a role here?
– Gartner has written extensively on this topic over the years. I tend to think that most folks spend too much time on talking about technology. For example, with EDI we could electronically interoperate across organizations. But how well did those EDI programs go? From a technical view point is there that much different from connecting EDI to connecting APIs? No, not really. Where these efforts break down is in the data that goes into the connection at one end and comes out the other. In too many cases rubbish went in, and rubbish came out. Worse, our definition and understanding of rubbish was different. So, I hear you say, let’s share metadata and make the data self-describing. Sure, that can help for sure. But it is not an answer to what is needed if either party are to understand and share in the business context and semantics for that informations use. As such I think there are different layers of interoperability and the easy stuff – technology – is solved. The upper levels of interoperability, where real business value exists, is much harder to attain. I wrote a note some years ago on the topic and more recently several blogs on the topic: Electronic Health Record (EHR) Interoperability – Why is it so hard? (2015) and What is Wrong with Interoperability (in healthcare)? (2016) though I have followed the topic in retail and CPG for years. See New P2P Solutions Will Redefine the B2B Supply Chain.
Should composability be used to create a pattern for integrating core solutions to 3rd party services or should it be a provider that can provide that?
– Yes. Composability is a pattern that should support anyone the ability to connect packaged (i.e. software) business capabilities from different vendors. Here is a suggested note: Use Gartner’s Reference Model to Deliver Intelligent Composable Business Applications.
When you talk IT are you including CIO or also CISO?
– The topic of security and data security came up (I think) when I mentioned connected governance and the state of governance (see above). In that context, CISCO’s might report to CIOs (Chief Information Officers) or they might be peers. This is analogous to Chief Data Officers, who may also report to CIOs or be peers.
Is composability Gartner created terminology ???
– Well I am not sure Garter invested the word but we certainly focus on the concept significantly in the last couple of years across Applications, Software Engineering, Data and Analytics and AI.
You mentioned Data Fabrics, and this is something we’ve looked at; however, it seems that the space isn’t clearly defined what a data fabric is. Do you feel it’s mature enough to pursue at this point?
– Yes, the concept is mature enough to pursue. As to the technical elements you would need to do more homework. Hope the links above give you some material to work with.
What impact are data fabrics and data virtualization expected to have on the data capture?
– I am not totally sure what you mean by data capture. Data fabric (its a design pattern, see references above) includes all of the data management landscape. As such tools and techniques to access, read and capture data would be needed. A data fabric that can’t read or capture data would not work. Data virtualization is a way to use data without actually copying it; it helps achieve some of the benefit of data integration and at the same time avoid some of the costs. Seems to me these things are all connected and impact each other. I’d set up a call with our team covering data management technologies.
Hi Andrew, You said people is not obvious but important. Nowadays it is loudly spoken that big resignation is very dangerous for companies. In this case, should not be people become obvious? What companies are doing to develop new skillful people in data domain?
– I can’t remember – sorry – the context of my comment. But I would agree with you that people are critical – more so than technology. And yes we should continually work to improve our teams’ capabilities.
How should an organization think about MDM from an investment priority standpoint in its DA journey? Is there a framework for investment priority?
– MDM is a data and analytics governance-related program, focused on master data. It enabled many other initiatives achieve their goals more effectively and often more efficiently. As to its priority, that should be determined in context to the overall DA Strategy. Its a good working assumption that it should be explored early, since it is so foundational. But MDM has been so poorly understood for many years that it remains a challenge for many organizations to get right. See Don’t You Need to Understand Your Business Information Architecture? and Four Steps to Start an Information Architecture Practice
Please enlighten us with the investment sequence in data management, data governance and data analytics…!
– I wish I could. Alas I was referring to a survey we did and it was fascinating. We published a note on the findings: Sequence Your Data and Analytics Investments to Maximize Business Value. This was one of the most interesting projects I was lucky enough to be involved in. We are doing more work on this in 2022.
What sustain the analysis that the delivery of business shifted from “as a service” to “operationalize lean and agile”? Shouldn’t it still be a service?
– I think two ideas we explored during the webinar that might have been conflated. The “as a service” was in context to the point that DA is often perceived differently in organizations. The oldest perception is that DA is an office and they delivery reports and dashboard to order. They operate “as a service”. More mature than this is a more consultative capability where the DA office offers insight and new ideas to those that used to ask for reports and dashboard. Lastly is the leadership role in the business side. The operationalization and lean and agile discussion was more to do with physical delivery models. These are related to the perception topic but independent of the perception, we might explore delivery models. I suppose we could say that a product delivery approach, adopting lean, agile and DevOps practices, can operate as a service but I am not sure that terminology is used this way. It might lead to more confusion – not sure.
How do we capture the data driven insights and decisions in an organization when it is spread out in different departments and very hard on management level to see if we get the insights but even harder to see if we act on them. Any advise on this dilemma?
– This is a question with many angles. I can offer a new angle that is gaining traction. And this angle takes me back to my roots. We recently wrote a note that explores the idea that decisions in one department might impact another, which is obvious. But in terms of business process design, in terms of decision design, many organizations do not think this way. As such these ideas might help alleviate the challenge your question poses. See Align DA With Value Streams to Optimize Decision Making and Business Value Creation and Toolkit: Use Value Stream Mapping to Optimize Your Product Information Supply Chain.
Do you have any recommendations for a business model that helps determine value for DA
– Not sure I fully grasp the question. We have various tools, best practices and techniques to help explore the ROI for a range of DA investments. We also have same for how to possibly value data assets themselves, what some might call monetization. There are too many to list really. On our website we have resources on “business value” of DA and also how to measure same. Start here maybe: Achieving the Business Value of Data and Analytics.
Q1. Do you think Data Management should be a part of IT Organization or part of Business Organization or somewhere in between?
– This might look like a simple question but it will really depend on what you mean by data management. This might mean the strategy and tactics of data management or it might mean the work to delivery and serve data (to analytic projects) or it might mean the sustainment of core data stores. The better question is where should this work take place? We addresses this question above with a note on Where to Organize the Work of Data and Analytics. This research does not tell you where to do the work; it is meant to provide the questions to ask in order to work out where to target the work, spanning reporting/analytics (classic), advanced analytics and data science (lab), data management and infrastructure, and DA governance. It is designed for a CIO, CDO or DA leader.
Q2. Many a times Teams working on data are looked at like doing a backend work. How does one explain Business that Data is the Oxygen behind analytics.
– I would try to leverage any number of techniques (listed above) to help. I don’t know that one idea works everywhere so try them all :). Try decision intelligence ( Decision Intelligence Is the Near Future of Decision Making ), storytelling ( Impactful Storytelling: Become a Powerful Data Storyteller), data literacy ( A Data and Analytics Leader’s Guide to Data Literacy ), our value pyramid ( Tool: How to Connect Data and Analytics to Business Value), and business value framework ( The Gartner Business Value Model: A Framework for Measuring Business Performance) and look at each of these areas as tools to put into your toolbox. They can all help you at any time depending not the context. There really is not one plan per se for everyone.
I would think a key importance would be how you generate and what data points are used to analyze especially if it is manually retrieved. Are their best practices on how to pin point the data points and most effective data collecting practices?
– This could be two questions and both are quite open ended. If your business decision can be well defined and the data needed can be well modeled (a known/known) then a standard analytics and BI approach should work well. If the decision is vague and ill formed, and more to do with an exploration or innovation, I would try a more advanced analytics and data science approach, perhaps using time-based approach. These are more unknown/unknown kinds of problems. As to how to collect or store the data – not sure if this means the data integration tactics (varied) or data storage options (varied). I would have an inquiry with our data management team.
In this decision making framework, it appears to infer Humans / Machines, I’d like to interpret it is Humans + Machines in order to be more effective.
– Yes, good point. It should me humans and machines. It might be human or machines in some situations.
There have been cases when we have huge amount of data that we have gathered from different sources with lot of efforts. But it will not be useful until or unless we make necessary impact. As you correctly said that reports and insights are not enough its more about impact we can make. How to accomplish that in real time? Any strategy or tool that can be super useful..if you can suggest
– Not sure there is one tool here. There might even be a range of options in terms of technique and tool. Real-time analytics is a hot topic on its right. Maybe this helps: 5 Essential Practices for Real-Time Analytics.
Would you agree that Decision capabilities should be included as a key component of hyperautomation?
– Yes, I would agree. Any program focused on methodically improve or exploiting automation should focus on business process design and that implies decision design. As such, decision intelligence should help organization achieve higher degrees of automation.
Data rich and Insights Poor, where should you start with a large organization that is data rich but want to ask questions and insights. Is the first step to understand what data sets you have. If not what is the first step
– I would argue that you should NOT start by cataloging your data. All you will do is spend a lot of money and collect a lot of data. Who knows or cares what the value is of that data? Why not start there first? Other than starting with a DA strategy I would (again) use any of these to try to engage with business leaders to get their buy in to the effort: Try decision intelligence ( Decision Intelligence Is the Near Future of Decision Making ), storytelling ( Impactful Storytelling: Become a Powerful Data Storyteller), data literacy ( A Data and Analytics Leader’s Guide to Data Literacy ), our value pyramid ( Tool: How to Connect Data and Analytics to Business Value), and business value framework ( The Gartner Business Value Model: A Framework for Measuring Business Performance) and look at each of these areas as tools to put into your toolbox. They can all help you at any time depending not the context. There really is not one plan per se for everyone.
Are we cycling back to “Decision support systems”? Have we always been there, distracted by the tech? Thoughts??
– Interesting. I don’t think we are going “back” to decision support systems. I think we have had them for many years. I used to be in Supply Chain and I used packaged applications for demand planning. They were very sophisticated; some used neural networks. This was in the 1990s! These were decision support systems. What I think is new is that newer technologies will help us all model our actual decision making process and that will shine light on something we have not looked at properly, if ever. The entire analytics, BI and data science markets (they are several) do not do a good job of discovering, modeling, defining, and explaining decisions – yet.
What do you mean by Data Fabric?
– Please see several answers above.
What is a recommended structure in IT to support the DA strategy…
– I am not sure I can answer this question here. What I can do is point you to a note that can help you start the dialog to work out what you need to answer that question. I would start here: Where to Organize the Work of Data and Analytics. Once your organization has a target state, you can then explore the IT capabilities and look for gaps and organizing principles.
Does Gartner have recommendations on KPIs that support the journey
– The most important measures of success are those that report achievement of the business impact sought. Everything else is secondary at best. This is why we focus so much on practical tips on how to relate data to outcome: Look at our value pyramid ( Tool: How to Connect Data and Analytics to Business Value), and business value framework ( The Gartner Business Value Model: A Framework for Measuring Business Performance).
A lot of rhetoric around Chief Data Officer…Is the role of Chief Analytic Officer gaining traction? The champion of the Analytic Demand which drives the build of Data Products to enable the Analytic Products / Capabilities
– The data we see is that CDO is gaining traction well and the label CAO is not. I would suggest that we not get caught up in the name. After all, if we were to do all this over again we would chose CIO. But that name is already taken. All that said, CDOs do lead the analytics efforts. CDOs do not focus on data alone or even data mostly. There is a lot of research in this area. Here ( CDO Agenda 2021: Influence and Impact of Successful CDOs in the Sixth Annual CDO Survey ) is our last survey of CDOs and equivalents (this includes CAOs) and our newest survey will publish soon and that data will feature prominently in our upcoming Data and Analytics Conference season.
How do you see DA impacting the Learning and Enablement space? Can you share couple of use-cases?
– I am not sure the angle here – I was not really talking about educational industry. But in terms of how to help improve the overall data literacy of the organization we have lots of research on that topic. A Data and Analytics Leader’s Guide to Data Literacy.
Can you describe data source exploitation, which is what you mentioned as an alternative to data cataloging?
– I have seen so many firms believe they can’t start DA governance without first cataloging all their data. This is flat wrong. There is a use case that does warrant starting with a catalog – that is closely related to data privacy risk. But for most other use cases you can start with an outcome based approach that should help align to any budding digital/DA strategy that would be equally outcome focused. This research helps explain and express how data explicitly drives outcomes and can help demonstrated to teams how data links to outcome: our value pyramid ( Tool: How to Connect Data and Analytics to Business Value.
Why do we feel that the focus isnt more balanced between internal and external priorities when it comes to data analytics.? It looks like there is a drastic drop between the top 5 and the rest. The rest seems more internal impacting to me. Internal would inherently impact the external with less cost to a company.
– Sorry not sure the context of the question. I don’t seem to remember me commenting on internal and external priorities. Would this mean internal and external to the organization? Do the DA team? Not sure. The priorities are whatever are needed to help the business meet their objectives as far as I can tell.
Are you seeing organizations applying Lean principles and Product Management mindsets being applied to the DA programs / Value Streams?
– Yes, we are seeing organizations in DA adopt DevOps practices, lean, agile and other iterative and quality-related efforts. Value streams was mentioned above too. Feel free to search on our web page or notes on these topics.
How does Global financial Research initiative in Data Analytics Sector including AI and ML performing technology strategy enhance get a great success in the Data Driven Augmentation operation of simulation?
– Sorry not sure what this question means. It is the last one in the list – but it might have been the first one entered into the system. So sorry for that. Please ping me so we can explore. Thanks.