Alter Everything

MaddieJ · ‎04-05-2023

How can you reach analytics nirvana? Analytics democratization! Melissa Burroughs and David Sweenor are co-authors of "Democratizing Analytics: Compete and Win by Empowering Your People with Data." In this episode, they'll explain why making advanced analytics more accessible in your organization will empower your teams to solve complex problems more efficiently, and they'll share actionable steps you can take to get started.

Panelists

Melissa Burroughs - @MelissaBu, LinkedIn, Twitter
David Sweenor - @DavidSweenor, LinkedIn, Twitter
Maddie Johannsen - @MaddieJ, LinkedIn

Topics

"Democratizing Analytics: Compete and Win by Empowering Your People with Data" - e-book by Melissa Burroughs and David Sweenor
Alteryx Analytics Maturity Assessment
International Institute for Analytics - Analytics Maturity Model
Alteryx ACE Program
Alteryx Innovators Program
Inspire 2023
Kotter's 8 Steps for Leading Change

Youtube Thumbnail2.png

Transcript

Episode Transcription

MADDIE: 00:02	Welcome to Alter Everything, a podcast about data science and analytics culture. I'm Maddie Johannsen, and today's episode is all about democratizing analytics. Joining me are two of my Alteryx colleagues, Melissa Burroughs, and David Sweenor. They're co-authors of the book, Democratizing Analytics, compete and win by empowering your people with data. Melissa Burroughs is the director of technical portfolio marketing at Alteryx.
MELISSA: 00:31	I'm sure all of us have experienced clicking into a cell and seeing the 15-level nested formula. How can you even explain what this formula is doing? How are you supposed to know if it's still functioning appropriately? What if the data sources has changed or formula's doing something unexpected? Oh, my blood pressure just went up thinking about it.
MADDIE: 00:51	And David Sweenor is the Senior Director of Portfolio Marketing at Alteryx.
DAVID: 00:56	Organizations really need to establish sort of a risk framework that has these different levels of governance, and Alteryx actually has a ton of experience in this. We have some of the largest customers in the world in highly-regulated environments that have solved this challenge. So the net of this is, if I was to make a headline, democratization versus governance, are they at odds? No. They can coexist.
MADDIE: 01:25	In this episode, Melissa and David will define buzzwords and hot topics in the analytics world. And they'll also share tips for how a modern analytics democratization and governance strategy can be the cornerstone of how you advance your organization's analytics maturity. Let's get started. In the analytics world that we occupy, there have been these buzzwordy concepts that have surfaced over the years. So digital transformation, data literacy, analytics, maturity, and what we're here to talk about today, data democratization. So let's start by defining this buzzword. What is data democratization?
MELISSA: 02:10	This is such a good question, Maddie. Thank you for starting us out with it because really it sounds like $40 language. Data democratization, it can be thought of as data to the people. It's the ongoing process of training and empowering professionals to work with and understand data. It's the act of making data accessible to more stakeholders and of educating them on how to use it to create insights. And in particular, data democratization is the act of removing barriers to working with data, especially the barrier of needing technical know-how. And democratizing data can really bring immense benefits to your organization, including things like faster decision-making, higher employee performance, greater efficiency, faster go-to-market for new offerings and new services, and even better customer experiences. And taken together, all of this can create a meaningful and sustainable competitive advantage for the business.
MADDIE: 03:21	Well, Melissa, thank you for that insightful response. And David, I noticed that the name of the book that you and Melissa wrote is Democratization of Analytics. It's not the democratization of data. Is there a difference between the two?
DAVID: 03:37	They're closely related. So as Melissa mentioned that data democratization is really the process of making the data available and accessible with the appropriate permissions to everybody within the organization, community, or society. And so we really want to break down these barriers. And that's something Melissa emphasized, and make sure people that need to access the data and have the right to the access, to the data they can. Now, the difference is analytics democratization really that's the process of making advanced analytics more accessible to a wide range of people. And don't let that term advanced analytics scare you. That's okay. Let's just say analytics. And so the goal of analytics is really democratization is to empower individuals with the ability to use the data and analytics to make informed business decisions and solve complex problems regardless of their technical background or expertise. So that's how I think about analytics democratization.
MELISSA: 04:36	I completely agree with you, David. I think that both are good. Democratizing data and providing people access to the raw materials they need to generate insights is the beginning of the journey. Right? We must start there. But it's typically not enough to actually complete the story. Right? Those tools and techniques that David referred to, especially those more advanced techniques when we talk about statistical analysis and predictive analysis, not everyone has got the training, the technical background, the code savviness, to be able to turn those data into those business-impacting insights that the organization requires. So democratizing data is step one, but democratizing analytics is where you really see that data turn into business value, broadly, across the organization.
MADDIE: 05:27	Got it. Okay. Yeah. That makes sense. And what is the current climate at organizations when it comes to analytics democratization?
MELISSA: 05:38	Another incisive question, Maddie, thank you. David, if it's all right, I'll be happy to take a shot at this first. Does that work?
DAVID: 05:45	Yeah. You take a whack at it, Melissa.
MELISSA: 05:48	It's become a pinata, Maddie. So current state of data and analytics democratization at organizations today, I would estimate that most organizations are at a place where they're broadly taking advantage of data-driven reporting, but perhaps only 5% of the workforce can access and fully leverage the data that they need to do their jobs superlatively. And you might be wondering, "Okay. Great. That's abstract. What does that look like? Is 5% good? Is it bad? What is 5% of the workforce enabled with data and analytics? Is that a bad thing?" And we can take, for example, what a low level, which I would argue 5% is a low level of data and analytics democratization did to the legacy reporting processes of a large European telecom organization. And I'll make this brief. But essentially, to generate some very necessary operational reports, this Telco had to link data from more than 140 different spreadsheet models together and do that sequentially.
MELISSA: 07:01	It was, as you can imagine, cumbersome, error-prone, and shockingly time-consuming to run. It often took up to four weeks to execute, and it required multiple runs before the necessary reports were successfully completed. They were working with, I think, 1,500-plus unique data sources and methodologies, and I'm sure you can guess, no one could comprehend or maintain the overall workflow. It was simply too complex. Each area required its own unique set of rules and processes. An absolute nightmare for the analysts involved and for the organization as a whole. Now, if you're interested to learn how this Telco finally ended their bad dream here, I do encourage you to pick up a copy of our new book that David mentioned, Democratizing Analytics, and learn about who you can reap the benefits of faster data-driven insights while maintaining proper data controls and discover best practices and actionable tactics from leading organizations all around the world.
MADDIE: 08:14	That's great. I love that you said nightmare. As soon as you started describing the use case, the first word that popped into my head was nightmare. So yeah. It's a good way to describe that. Yeah. And David, just thinking about analytics maturity, I brought this up earlier as an example of another buzzword that we've been seeing. I'd love for you to introduce us to this concept as well.
DAVID: 08:40	Yeah. That's a great suggestion there, Maddie. I think really democratization is related to analytics maturity. And for those who are not aware of what analytics maturity is, it's a maturity model, and there are different ones that are out there, but generally, they have five stages. So stage one is I'm not really doing anything with data and analytics or is completely random to stage five, which is the highest level of maturity. That'd be like analytics Nirvana. Now, I don't know if you ever get there, but some organizations are quite close to analytics Nirvana. And what this analytics maturity measures, it measures, your effective use and governance of data. It measures technology and how you're using that technology. It measures organizational structures. It looks at people in practices. And what we found, and there's lots of research to back this up, the more analytically mature your organization is, the more competitive you are in the marketplace across just about every financial metric. I think some of the research we've seen there are 32 financial KPIs and then maybe 28, 29 of them all showed positive increase. So in a nutshell, democratization helps you accelerate your analytics maturity. Because if you think about this, you can't have three people in a large organization leading the charge. You have to have everybody in the organization to be able to participate and use data and analytics to solve their business challenges that lead to better business outcomes.
MADDIE: 10:27	So democratizing, if you get to stage three, it's not where you start democratizing. It's more so how you get there to the different stages. It's how you advance.
DAVID: 10:42	Yes, and it's how you advance more quickly. And Melissa, I bet you have a thought on this. I know you've done a lot of work on the analytics, the assessment that we actually have our website.
MELISSA: 10:53	Yeah. That's a great point, David. Thank you. Yeah. I think Maddie, you're right. What you've described is there's not one point in time where you start to think about democratization. Hearkening back to the maturity model David mentioned, we can use as an example of an analytics maturity model developed by the International Institute for Analytics, which is an independent research firm that's been collecting benchmark data on analytics maturity for I believe nearly a decade. So they have a very robust five-stage model that really assesses where a company is on their journey through analytics sophistication. Just as David described stage one, absolute analytic beginner all the way to stage five. I think David called it analytics Nirvana. I believe the IIA calls this the Analytic Competitor, which is an organization that has turned its data and analytics into a distinct competitive differentiator in their market. So that's very powerful. So on this scale, from 1 to 5, the average company scores a paltry 2.2, 2.2 out of 5. On a curve, that is not a passing grade.
DAVID: 12:06	I want to eat at its restaurant that's a 2.2. That's for sure.
MADDIE: 12:09	True.
MELISSA: 12:10	Exactly. And what that means practically is that these organizations are just doing localized analytics. As David said, a large leading company can't just have three people doing it. And really a 2.2 is that small percentage of folks. Maybe they're using reports in specific business units to address common business-as-usual questions, kind of little sad trumpet use of analytics in the org. But as David mentioned, I think this is something really important to know about your own business. And if you're curious to find out where your organization lives on this spectrum, I am happy to share, again as David mentioned, that Alteryx offers a free online self-assessment that you can take that will give you your analytics maturity score. Just go to Alteryx.com/analyticsmaturity, and in less than 15 minutes, you'll have your answer.
MADDIE: 13:12	Awesome. Yeah. I love that resources are available to people, and we'll be sure to link that in the show notes. So okay, let's say that I get a two or a three or a score that I'm not happy with. Like we were saying, data democratization is a really good practice of how you can get to the stages that you want to be at. You want to get to that five-star restaurant quality analytics maturity. So with the data democratization, that can be kind of potentially scary to think about. It could be a big undertaking to think about. And I think something that is useful to make this less scary would be to have a governance strategy. And we've talked about this a little bit on the podcast. We love talking about it on the community. But I'd love to hear from you guys. How can folks implement this simultaneously? And can you share an example?
MELISSA: 14:10	So absolutely, Maddie. I love your point that this sort of takes out some of the fear of change, having a solid governance strategy. And indeed, when you're enabling more of the business to have access to and interact with your valuable data, you also need a strategy for maintaining appropriate control and oversight. So as a quick review, some of the major principles, let's say the big four principles of governance are accountability, regulation, stewardship, and quality. And a lot of this is intuitive stuff. Right? An accountability includes understanding the value and the risks involved with data access to ensure that data is being used properly. And this is best captured and represented by a cross-functional governing body like an analytics council, for example, that you might create within your business. Regulation describes a set of well-defined rules for data usage and access, which balance the flexibility you need with security and privacy concerns. Data stewardship provides the basis on which that regulation of data can be built as each area within the org needs expertise around its own data. And finally, the principle of quality, that's pretty intuitive. Data is useless if it is incomplete or incorrect. So standards related to accuracy and completeness of data, they need to be established. So putting these principles into action is all about figuring out what works for your business, and I'll segue into an example because I have a feeling that David has some strong opinions on that phrase what works for your business.
DAVID: 16:00	No. No.
MELISSA: 16:03	But I'll share a quick example, Maddie, since you remarked on it. One that I'm aware of is when UBS, a bank in Europe, was working to make data and analytics more available while also maintaining governance over that data. So this exact case you've described, they created a repeatable, easy-to-use, and easy-to-maintain system to provide analytics insights across the organization. They used a self-service method for data access and preparation via, of course, the Alteryx analytics automation platform, and they classified their workflows into different types based on things like what the workflows did, the data that was being used, who was being served, and they subjected each type of workflow to uniform requirements. And so UBS empowered their workflow-building analysts to classify their workflows as well as govern them to meet the standards that were required by the organization. And all the workflows are regularly run through this governance review process to ensure compliance. And if you'd like to learn more about the details of UBS's approach to governance, there is a bunch more information available in our new book, Democratizing Analytics.
MADDIE: 17:30	That's such a great example. Putting things in a very digestible list is always helpful to understand the concept. So thank you for that, Melissa.
MELISSA: 17:39	My pleasure. I think people think it's scarier than it is. It actually can be quite straightforward. And I do want to invite David to add his POV on governance because I know you feel strongly about it, mister.
DAVID: 17:54	Yeah. It's just really a top of mind for many of our customers and our prospects, and Melissa and I were just at the Gartner Data and Analytics Conference, and there was a central theme throughout it. And I think they had a statistic that said, "Through 2025, 80% of organizations seeking to scale digital business will fail because they do not take a modern approach to data and analytics governance." And I think we see, a lot of companies when they think about governance, they approach it like it's a one size fits all. We're closing our financial books. It has to be absolutely validated, vetted, locked down, put in Fort Knox, and they lock the whole organization down. What happens when you do that is that stifles innovation. So I think they need to have a governance framework. It's not a one-size-fits-all. And so think of it as a spectrum. On one side, if I'm level one and doing analytics just for myself, that probably doesn't need to be governed at all. If I'm submitting financial statements to SEC and want to hear it of Sarbanes-Oxley or the FDA, if I'm in the pharma industry, that needs to be highly governed. But in between, there are a whole spectrum of things. And so I think organizations really need to establish sort of a risk framework that has these different levels of governance, and Alteryx actually has a ton of experience in this. We have some of the largest customers in the world and highly regulated environments that have solved this challenge. So the net of this is, if I was to make a headline, democratization versus governance, are they at odds? No. They can coexist. So you don't really need to boil the ocean with this governance topic. Melissa, I have a question for you. Why don't organizations govern spreadsheets?
MELISSA: 20:02	That is a great question. If you think about the principles of governance that could be applied, right, having someone be accountable, have rules in place, having experts, data stewards, and ensuring quality, the enforcement level required spreadsheets are this ad hoc tool in the mind. I think of much of our audience and our analysts, the overhead required to monitor and keep on top of and create rules for all of the unique and unexpected use cases for spreadsheets, it feels like you've gone straight through the looking glass. David, would you sign up to govern spreadsheets in an organization? Would you want that role?
DAVID: 20:47	No. I can barely use a spreadsheet. They're always wrong. I've never got into spreadsheets because they're not auditable. There's been lots of examples where spreadsheets cause multi-million-dollar errors and companies. So I would never sign up for that. And I think that's why that example of the telecom company earlier you mentioned, they had a lot of spreadsheets, and now they don't. Spreadsheets are the wild west. So I would encourage every organization that's out there, if you want to be scared, take a look at the spreadsheet usage within your organization and how they're being used for critical business processes.
MELISSA: 21:21	Yeah. Absolutely. The spreadsheet is the scratch pad of the analytically-minded thinker, and you should not be attempting to productionize work that took place on a scratch pad. It's just not an auditable environment, David, as you commented. And auditable can mean a number of things like change, tracking, and history, and watching data lineage, but also just incomprehensible. I'm sure all of us have experienced clicking into a cell and seeing the 15-level nested formula. Have you ever seen that, David? You click on it and all the cells light up in bright colors all over the spreadsheet as if that's [crosstalk]?
DAVID: 22:01	As I mentioned, and it's always wrong. I get an error because I'm not good at spreadsheet formulas.
MELISSA: 22:05	You get the ref error.
DAVID: 22:07	Yeah, question mark ref.
MELISSA: 22:10	Oh, and it's just, how can you even explain what this formula is doing? How are you supposed to know if it's still functioning appropriately? What if the data sources has changed or the format of the data has changed, and the formula is doing something unexpected? Oh, my blood pressure just went up thinking about it.
DAVID: 22:28	I could talk all day about this, but I think the message to our audience really is, there's a better way. There's a better way than spreadsheets.
MADDIE: 22:35	Yeah. No, I agree. And I think switching gears a little bit if I'm a practitioner, and my leader has just laid out a new analytics democratization strategy. If I'm thinking about how I can use this to potentially advance my career, something that you said, David, about needing a modern governance strategy in order to encourage that innovation. I think that there's something in there in terms of encouraging your employees to still be creative, still be thoughtful about their workflows and what they're doing with analytics, and again, a potentially an opportunity to advance their career. So let's talk a little bit more about that. David, I'd love to hear your thoughts.
DAVID: 23:20	Sure. So maybe to first start out, there's probably five different areas where people start with data and analytics for business outcomes. They're going to use data analytics and AI to either improve decision-making, better understand customers and markets, deliver more intelligent products, services, maybe to improve a turtle process, or to monetize data. So let's assume we're going to do one of these things. And all the research you read, "Hey, we've got these intergalactic projects that are out there." And they're multiyear projects. We see some of these, of course, but a lot of this starts with the individual analysts. They are doing something with data and analytics. That affects their personal productivity. That's where it starts. So I would say, if you're an analyst, you can help champion this initiative with your peers, and some guiding principles that I would have is really focused on the problem you want to [small?]. It could be a big one, but it might be small, but I typically just follow Kotter's eight-step model. You have to have a sense of urgency. You need to form a guiding coalition. You need to make sure that the organization understands the strategic vision.
DAVID: 24:27	You have to have help. So you need to have a list of volunteer army. This isn't Kotter's words, but you need to create a Tiger team to remove barriers, essentially. And you need to generate these short-term wins. Gone are the days are we going to have multiyear super long projects where we're waiting and waiting and waiting for business benefit and business outcomes. We have to generate these short-term wins. We want to sustain this acceleration and really make sure that this change is sticking within the organization. Don't be afraid to take that plunging. Maybe you don't know how to write code. That doesn't matter. The software does this for you these days. It's super easy to use. Maybe you don't have the appropriate level of analytics or data literacy. That's okay. There's plenty of free training and how to understand and interpret data out there. We have a wealth of information on our community, but there are certainly other sources as well. So I would say many organizations, many people, they just need a little nudge. They need a little activation energy to get going. And once they're going, they will accomplish amazing things for your organization. So my key message is just to get started, don't wait. You can do it.
MELISSA: 25:45	Yeah. Exactly. I 100% agree with David here. Absolutely become a champion and work to upskill yourself. I would add to that also, especially if you're looking after your own career, work to become an analytics teacher or a trainer as soon as you can. When you learn something, share it with others. It's surprising. That Delta doesn't have to be so great. You can be just a few days or a few weeks ahead of your peers, and you can still provide value by training and sharing. And you become an integral part of the solution when you multiplex your knowledge this way, and leadership tends to notice. You'll begin building an analytics community around yourself as you build your own reputation as a trusted and emerging expert in this high-growth field. In fact, if you are an Alteryx user, I encourage you to look into our Alteryx Innovators Program which is where you can earn rewards for participating in advocacy activities and gain exposure to a worldwide audience. Plus, if you're a true die-hard Alteryx aficionado and expert, consider becoming an Alteryx ace. This is a great start to aim for. Alteryx aces are analytics experts and community advocates who innovate the global analytics marketplace and community. So there's a lot you can do to develop yourself while supporting this change initiative of analytics, democratization for your business.
MADDIE: 27:32	Awesome. That's great. Yeah. Thanks for plugging both of those programs. I think that there's a lot of opportunity. I think people sometimes think about LinkedIn or social media or those kinds of platforms to get their name out there, and they're awesome. But I love that we also have programs as well in-house that people can participate in. And that's a really good segue to my next question. What are some best practice recommendations that you have for leaders?
MELISSA: 27:59	This is an important one. So we visited the practitioner side of the equation and then if we're in leadership, what do we do? I have a couple suggestions to offer. When you're setting out to democratize analytics in your organization, I believe it really pays to keep people front and center in your mind. We really can't just impose things on folks. There is resistance to change. And there are three groups of people I'm thinking about here, in particular, analysts, of course, other business leaders, and your senior leadership team. So first, think about those folks with their hands on the keyboards who are turning data into insights every day. You need to make analysts the foundation of your change process. And they can help you. They can provide you important educational content, information, training, right, if we've got those champion advocate analysts and reference materials that your organization needs to develop a more analytically-minded culture.
MELISSA: 29:13	These are your internal thought leaders, and they can make change stick. Also, do think about your peers. Right? Work quickly as you can to create a cross-functional group of leaders to maintain oversight of analytics, resources, and activities across the business. This kind of analytics council, as I mentioned before, that's going to help enable important data governance activities, as well as identify opportunities and potential issues as your program really gets off the ground. And of course, we cannot forget about our senior executive leadership team. Ensure that the top of the organization and really the company as a whole, understands the benefit of investing time and resources in analytics initiatives. If you have close relationships with senior executives or you are one, yourself, work to start modeling new analytically-driven behaviors such as using data and analytics as a part of your leadership process to be the change that you want to see in the business.
MADDIE: 30:35	Got it. Yeah. No. Those are great best practices. And David, question for you. Let's say that I am a leader. I maybe took the analytics maturity assessment. I didn't score what I wanted to, and I'm thinking to myself, "The people who did get a five, what are their teams doing that my team is not doing?" I'd love for you to share your thoughts there.
DAVID: 31:00	Thank you for that question, Maddie. First of all, out of the hundreds or thousands of people that have taken this maturity assessment, I have yet to see a five. So maybe they're juicing the system if they have a five. I'm not sure. But when we look at sort of what high-performance teams really do differently, and actually, this is experientially and some research to back this up. I think number one is really align your strategy to business goals. We don't want to do analytics for the sake of doing analytics. We have to understand how is the business going to change its behaviors because of this analytic output. You just say, "Hey, the trend is going up. That's great. And you don't change your business. Why bother?" So make sure it has meaning. Number two, invest in talent and training. So Melissa has covered this. Your analyst, they have the knowledge of the business. And if they don't invest in training, then invest in your talent to make sure they can implement that change you want. Number three, collaboration. You can't do this in isolation. The whole company needs to be behind this. There's cross-functional teams that are needed along every step of the journey. Number four, play strong data and analytics governance processes, you don't want to get in a hot soup. We talked about that earlier. Number five, make sure whatever you do you can establish a way so it's repeatable, so repeatable methodologies. Don't keep reinventing the wheel so you can apply this to your next project. And the last one, number six, is just ensure adoption of this. Make sure this change is sticking. Maddie, you mentioned start with a maturity. Melissa and I actually put a checklist in the back of this book, eight steps to getting started with analytics democratization. Step one is to take the maturity assessment. So thank you for reading all the way to the appendix in our book. Everybody go get a copy and pick it up and follow that checklist. I think it's got some real practical steps.
MADDIE: 32:50	Thanks for listening. Links to the analytics maturity assessment and Melissa and David's book, Democratizing Analytics are available on our show notes at community.Alteryx.com/podcast. Catch you next time. [music]

MADDIE: 00:02 Welcome to Alter Everything, a podcast about data science and analytics culture. I'm Maddie Johannsen, and today's episode is all about democratizing analytics. Joining me are two of my Alteryx colleagues, Melissa Burroughs, and David Sweenor. They're co-authors of the book, Democratizing Analytics, compete and win by empowering your people with data. Melissa Burroughs is the director of technical portfolio marketing at Alteryx. MELISSA: 00:31 I'm sure all of us have experienced clicking into a cell and seeing the 15-level nested formula. How can you even explain what this formula is doing? How are you supposed to know if it's still functioning appropriately? What if the data sources has changed or formula's doing something unexpected? Oh, my blood pressure just went up thinking about it. MADDIE: 00:51 And David Sweenor is the Senior Director of Portfolio Marketing at Alteryx. DAVID: 00:56 Organizations really need to establish sort of a risk framework that has these different levels of governance, and Alteryx actually has a ton of experience in this. We have some of the largest customers in the world in highly-regulated environments that have solved this challenge. So the net of this is, if I was to make a headline, democratization versus governance, are they at odds? No. They can coexist. MADDIE: 01:25 In this episode, Melissa and David will define buzzwords and hot topics in the analytics world. And they'll also share tips for how a modern analytics democratization and governance strategy can be the cornerstone of how you advance your organization's analytics maturity. Let's get started. In the analytics world that we occupy, there have been these buzzwordy concepts that have surfaced over the years. So digital transformation, data literacy, analytics, maturity, and what we're here to talk about today, data democratization. So let's start by defining this buzzword. What is data democratization? MELISSA: 02:10 This is such a good question, Maddie. Thank you for starting us out with it because really it sounds like $40 language. Data democratization, it can be thought of as data to the people. It's the ongoing process of training and empowering professionals to work with and understand data. It's the act of making data accessible to more stakeholders and of educating them on how to use it to create insights. And in particular, data democratization is the act of removing barriers to working with data, especially the barrier of needing technical know-how. And democratizing data can really bring immense benefits to your organization, including things like faster decision-making, higher employee performance, greater efficiency, faster go-to-market for new offerings and new services, and even better customer experiences. And taken together, all of this can create a meaningful and sustainable competitive advantage for the business. MADDIE: 03:21 Well, Melissa, thank you for that insightful response. And David, I noticed that the name of the book that you and Melissa wrote is Democratization of Analytics. It's not the democratization of data. Is there a difference between the two? DAVID: 03:37 They're closely related. So as Melissa mentioned that data democratization is really the process of making the data available and accessible with the appropriate permissions to everybody within the organization, community, or society. And so we really want to break down these barriers. And that's something Melissa emphasized, and make sure people that need to access the data and have the right to the access, to the data they can. Now, the difference is analytics democratization really that's the process of making advanced analytics more accessible to a wide range of people. And don't let that term advanced analytics scare you. That's okay. Let's just say analytics. And so the goal of analytics is really democratization is to empower individuals with the ability to use the data and analytics to make informed business decisions and solve complex problems regardless of their technical background or expertise. So that's how I think about analytics democratization. MELISSA: 04:36 I completely agree with you, David. I think that both are good. Democratizing data and providing people access to the raw materials they need to generate insights is the beginning of the journey. Right? We must start there. But it's typically not enough to actually complete the story. Right? Those tools and techniques that David referred to, especially those more advanced techniques when we talk about statistical analysis and predictive analysis, not everyone has got the training, the technical background, the code savviness, to be able to turn those data into those business-impacting insights that the organization requires. So democratizing data is step one, but democratizing analytics is where you really see that data turn into business value, broadly, across the organization. MADDIE: 05:27 Got it. Okay. Yeah. That makes sense. And what is the current climate at organizations when it comes to analytics democratization? MELISSA: 05:38 Another incisive question, Maddie, thank you. David, if it's all right, I'll be happy to take a shot at this first. Does that work? DAVID: 05:45 Yeah. You take a whack at it, Melissa. MELISSA: 05:48 It's become a pinata, Maddie. So current state of data and analytics democratization at organizations today, I would estimate that most organizations are at a place where they're broadly taking advantage of data-driven reporting, but perhaps only 5% of the workforce can access and fully leverage the data that they need to do their jobs superlatively. And you might be wondering, "Okay. Great. That's abstract. What does that look like? Is 5% good? Is it bad? What is 5% of the workforce enabled with data and analytics? Is that a bad thing?" And we can take, for example, what a low level, which I would argue 5% is a low level of data and analytics democratization did to the legacy reporting processes of a large European telecom organization. And I'll make this brief. But essentially, to generate some very necessary operational reports, this Telco had to link data from more than 140 different spreadsheet models together and do that sequentially. MELISSA: 07:01 It was, as you can imagine, cumbersome, error-prone, and shockingly time-consuming to run. It often took up to four weeks to execute, and it required multiple runs before the necessary reports were successfully completed. They were working with, I think, 1,500-plus unique data sources and methodologies, and I'm sure you can guess, no one could comprehend or maintain the overall workflow. It was simply too complex. Each area required its own unique set of rules and processes. An absolute nightmare for the analysts involved and for the organization as a whole. Now, if you're interested to learn how this Telco finally ended their bad dream here, I do encourage you to pick up a copy of our new book that David mentioned, Democratizing Analytics, and learn about who you can reap the benefits of faster data-driven insights while maintaining proper data controls and discover best practices and actionable tactics from leading organizations all around the world. MADDIE: 08:14 That's great. I love that you said nightmare. As soon as you started describing the use case, the first word that popped into my head was nightmare. So yeah. It's a good way to describe that. Yeah. And David, just thinking about analytics maturity, I brought this up earlier as an example of another buzzword that we've been seeing. I'd love for you to introduce us to this concept as well. DAVID: 08:40 Yeah. That's a great suggestion there, Maddie. I think really democratization is related to analytics maturity. And for those who are not aware of what analytics maturity is, it's a maturity model, and there are different ones that are out there, but generally, they have five stages. So stage one is I'm not really doing anything with data and analytics or is completely random to stage five, which is the highest level of maturity. That'd be like analytics Nirvana. Now, I don't know if you ever get there, but some organizations are quite close to analytics Nirvana. And what this analytics maturity measures, it measures, your effective use and governance of data. It measures technology and how you're using that technology. It measures organizational structures. It looks at people in practices. And what we found, and there's lots of research to back this up, the more analytically mature your organization is, the more competitive you are in the marketplace across just about every financial metric. I think some of the research we've seen there are 32 financial KPIs and then maybe 28, 29 of them all showed positive increase. So in a nutshell, democratization helps you accelerate your analytics maturity. Because if you think about this, you can't have three people in a large organization leading the charge. You have to have everybody in the organization to be able to participate and use data and analytics to solve their business challenges that lead to better business outcomes. MADDIE: 10:27 So democratizing, if you get to stage three, it's not where you start democratizing. It's more so how you get there to the different stages. It's how you advance. DAVID: 10:42 Yes, and it's how you advance more quickly. And Melissa, I bet you have a thought on this. I know you've done a lot of work on the analytics, the assessment that we actually have our website. MELISSA: 10:53 Yeah. That's a great point, David. Thank you. Yeah. I think Maddie, you're right. What you've described is there's not one point in time where you start to think about democratization. Hearkening back to the maturity model David mentioned, we can use as an example of an analytics maturity model developed by the International Institute for Analytics, which is an independent research firm that's been collecting benchmark data on analytics maturity for I believe nearly a decade. So they have a very robust five-stage model that really assesses where a company is on their journey through analytics sophistication. Just as David described stage one, absolute analytic beginner all the way to stage five. I think David called it analytics Nirvana. I believe the IIA calls this the Analytic Competitor, which is an organization that has turned its data and analytics into a distinct competitive differentiator in their market. So that's very powerful. So on this scale, from 1 to 5, the average company scores a paltry 2.2, 2.2 out of 5. On a curve, that is not a passing grade. DAVID: 12:06 I want to eat at its restaurant that's a 2.2. That's for sure. MADDIE: 12:09 True. MELISSA: 12:10 Exactly. And what that means practically is that these organizations are just doing localized analytics. As David said, a large leading company can't just have three people doing it. And really a 2.2 is that small percentage of folks. Maybe they're using reports in specific business units to address common business-as-usual questions, kind of little sad trumpet use of analytics in the org. But as David mentioned, I think this is something really important to know about your own business. And if you're curious to find out where your organization lives on this spectrum, I am happy to share, again as David mentioned, that Alteryx offers a free online self-assessment that you can take that will give you your analytics maturity score. Just go to Alteryx.com/analyticsmaturity, and in less than 15 minutes, you'll have your answer. MADDIE: 13:12 Awesome. Yeah. I love that resources are available to people, and we'll be sure to link that in the show notes. So okay, let's say that I get a two or a three or a score that I'm not happy with. Like we were saying, data democratization is a really good practice of how you can get to the stages that you want to be at. You want to get to that five-star restaurant quality analytics maturity. So with the data democratization, that can be kind of potentially scary to think about. It could be a big undertaking to think about. And I think something that is useful to make this less scary would be to have a governance strategy. And we've talked about this a little bit on the podcast. We love talking about it on the community. But I'd love to hear from you guys. How can folks implement this simultaneously? And can you share an example? MELISSA: 14:10 So absolutely, Maddie. I love your point that this sort of takes out some of the fear of change, having a solid governance strategy. And indeed, when you're enabling more of the business to have access to and interact with your valuable data, you also need a strategy for maintaining appropriate control and oversight. So as a quick review, some of the major principles, let's say the big four principles of governance are accountability, regulation, stewardship, and quality. And a lot of this is intuitive stuff. Right? An accountability includes understanding the value and the risks involved with data access to ensure that data is being used properly. And this is best captured and represented by a cross-functional governing body like an analytics council, for example, that you might create within your business. Regulation describes a set of well-defined rules for data usage and access, which balance the flexibility you need with security and privacy concerns. Data stewardship provides the basis on which that regulation of data can be built as each area within the org needs expertise around its own data. And finally, the principle of quality, that's pretty intuitive. Data is useless if it is incomplete or incorrect. So standards related to accuracy and completeness of data, they need to be established. So putting these principles into action is all about figuring out what works for your business, and I'll segue into an example because I have a feeling that David has some strong opinions on that phrase what works for your business. DAVID: 16:00 No. No. MELISSA: 16:03 But I'll share a quick example, Maddie, since you remarked on it. One that I'm aware of is when UBS, a bank in Europe, was working to make data and analytics more available while also maintaining governance over that data. So this exact case you've described, they created a repeatable, easy-to-use, and easy-to-maintain system to provide analytics insights across the organization. They used a self-service method for data access and preparation via, of course, the Alteryx analytics automation platform, and they classified their workflows into different types based on things like what the workflows did, the data that was being used, who was being served, and they subjected each type of workflow to uniform requirements. And so UBS empowered their workflow-building analysts to classify their workflows as well as govern them to meet the standards that were required by the organization. And all the workflows are regularly run through this governance review process to ensure compliance. And if you'd like to learn more about the details of UBS's approach to governance, there is a bunch more information available in our new book, Democratizing Analytics. MADDIE: 17:30 That's such a great example. Putting things in a very digestible list is always helpful to understand the concept. So thank you for that, Melissa. MELISSA: 17:39 My pleasure. I think people think it's scarier than it is. It actually can be quite straightforward. And I do want to invite David to add his POV on governance because I know you feel strongly about it, mister. DAVID: 17:54 Yeah. It's just really a top of mind for many of our customers and our prospects, and Melissa and I were just at the Gartner Data and Analytics Conference, and there was a central theme throughout it. And I think they had a statistic that said, "Through 2025, 80% of organizations seeking to scale digital business will fail because they do not take a modern approach to data and analytics governance." And I think we see, a lot of companies when they think about governance, they approach it like it's a one size fits all. We're closing our financial books. It has to be absolutely validated, vetted, locked down, put in Fort Knox, and they lock the whole organization down. What happens when you do that is that stifles innovation. So I think they need to have a governance framework. It's not a one-size-fits-all. And so think of it as a spectrum. On one side, if I'm level one and doing analytics just for myself, that probably doesn't need to be governed at all. If I'm submitting financial statements to SEC and want to hear it of Sarbanes-Oxley or the FDA, if I'm in the pharma industry, that needs to be highly governed. But in between, there are a whole spectrum of things. And so I think organizations really need to establish sort of a risk framework that has these different levels of governance, and Alteryx actually has a ton of experience in this. We have some of the largest customers in the world and highly regulated environments that have solved this challenge. So the net of this is, if I was to make a headline, democratization versus governance, are they at odds? No. They can coexist. So you don't really need to boil the ocean with this governance topic. Melissa, I have a question for you. Why don't organizations govern spreadsheets? MELISSA: 20:02 That is a great question. If you think about the principles of governance that could be applied, right, having someone be accountable, have rules in place, having experts, data stewards, and ensuring quality, the enforcement level required spreadsheets are this ad hoc tool in the mind. I think of much of our audience and our analysts, the overhead required to monitor and keep on top of and create rules for all of the unique and unexpected use cases for spreadsheets, it feels like you've gone straight through the looking glass. David, would you sign up to govern spreadsheets in an organization? Would you want that role? DAVID: 20:47 No. I can barely use a spreadsheet. They're always wrong. I've never got into spreadsheets because they're not auditable. There's been lots of examples where spreadsheets cause multi-million-dollar errors and companies. So I would never sign up for that. And I think that's why that example of the telecom company earlier you mentioned, they had a lot of spreadsheets, and now they don't. Spreadsheets are the wild west. So I would encourage every organization that's out there, if you want to be scared, take a look at the spreadsheet usage within your organization and how they're being used for critical business processes. MELISSA: 21:21 Yeah. Absolutely. The spreadsheet is the scratch pad of the analytically-minded thinker, and you should not be attempting to productionize work that took place on a scratch pad. It's just not an auditable environment, David, as you commented. And auditable can mean a number of things like change, tracking, and history, and watching data lineage, but also just incomprehensible. I'm sure all of us have experienced clicking into a cell and seeing the 15-level nested formula. Have you ever seen that, David? You click on it and all the cells light up in bright colors all over the spreadsheet as if that's [crosstalk]? DAVID: 22:01 As I mentioned, and it's always wrong. I get an error because I'm not good at spreadsheet formulas. MELISSA: 22:05 You get the ref error. DAVID: 22:07 Yeah, question mark ref. MELISSA: 22:10 Oh, and it's just, how can you even explain what this formula is doing? How are you supposed to know if it's still functioning appropriately? What if the data sources has changed or the format of the data has changed, and the formula is doing something unexpected? Oh, my blood pressure just went up thinking about it. DAVID: 22:28 I could talk all day about this, but I think the message to our audience really is, there's a better way. There's a better way than spreadsheets. MADDIE: 22:35 Yeah. No, I agree. And I think switching gears a little bit if I'm a practitioner, and my leader has just laid out a new analytics democratization strategy. If I'm thinking about how I can use this to potentially advance my career, something that you said, David, about needing a modern governance strategy in order to encourage that innovation. I think that there's something in there in terms of encouraging your employees to still be creative, still be thoughtful about their workflows and what they're doing with analytics, and again, a potentially an opportunity to advance their career. So let's talk a little bit more about that. David, I'd love to hear your thoughts. DAVID: 23:20 Sure. So maybe to first start out, there's probably five different areas where people start with data and analytics for business outcomes. They're going to use data analytics and AI to either improve decision-making, better understand customers and markets, deliver more intelligent products, services, maybe to improve a turtle process, or to monetize data. So let's assume we're going to do one of these things. And all the research you read, "Hey, we've got these intergalactic projects that are out there." And they're multiyear projects. We see some of these, of course, but a lot of this starts with the individual analysts. They are doing something with data and analytics. That affects their personal productivity. That's where it starts. So I would say, if you're an analyst, you can help champion this initiative with your peers, and some guiding principles that I would have is really focused on the problem you want to [small?]. It could be a big one, but it might be small, but I typically just follow Kotter's eight-step model. You have to have a sense of urgency. You need to form a guiding coalition. You need to make sure that the organization understands the strategic vision. DAVID: 24:27 You have to have help. So you need to have a list of volunteer army. This isn't Kotter's words, but you need to create a Tiger team to remove barriers, essentially. And you need to generate these short-term wins. Gone are the days are we going to have multiyear super long projects where we're waiting and waiting and waiting for business benefit and business outcomes. We have to generate these short-term wins. We want to sustain this acceleration and really make sure that this change is sticking within the organization. Don't be afraid to take that plunging. Maybe you don't know how to write code. That doesn't matter. The software does this for you these days. It's super easy to use. Maybe you don't have the appropriate level of analytics or data literacy. That's okay. There's plenty of free training and how to understand and interpret data out there. We have a wealth of information on our community, but there are certainly other sources as well. So I would say many organizations, many people, they just need a little nudge. They need a little activation energy to get going. And once they're going, they will accomplish amazing things for your organization. So my key message is just to get started, don't wait. You can do it. MELISSA: 25:45 Yeah. Exactly. I 100% agree with David here. Absolutely become a champion and work to upskill yourself. I would add to that also, especially if you're looking after your own career, work to become an analytics teacher or a trainer as soon as you can. When you learn something, share it with others. It's surprising. That Delta doesn't have to be so great. You can be just a few days or a few weeks ahead of your peers, and you can still provide value by training and sharing. And you become an integral part of the solution when you multiplex your knowledge this way, and leadership tends to notice. You'll begin building an analytics community around yourself as you build your own reputation as a trusted and emerging expert in this high-growth field. In fact, if you are an Alteryx user, I encourage you to look into our Alteryx Innovators Program which is where you can earn rewards for participating in advocacy activities and gain exposure to a worldwide audience. Plus, if you're a true die-hard Alteryx aficionado and expert, consider becoming an Alteryx ace. This is a great start to aim for. Alteryx aces are analytics experts and community advocates who innovate the global analytics marketplace and community. So there's a lot you can do to develop yourself while supporting this change initiative of analytics, democratization for your business. MADDIE: 27:32 Awesome. That's great. Yeah. Thanks for plugging both of those programs. I think that there's a lot of opportunity. I think people sometimes think about LinkedIn or social media or those kinds of platforms to get their name out there, and they're awesome. But I love that we also have programs as well in-house that people can participate in. And that's a really good segue to my next question. What are some best practice recommendations that you have for leaders? MELISSA: 27:59 This is an important one. So we visited the practitioner side of the equation and then if we're in leadership, what do we do? I have a couple suggestions to offer. When you're setting out to democratize analytics in your organization, I believe it really pays to keep people front and center in your mind. We really can't just impose things on folks. There is resistance to change. And there are three groups of people I'm thinking about here, in particular, analysts, of course, other business leaders, and your senior leadership team. So first, think about those folks with their hands on the keyboards who are turning data into insights every day. You need to make analysts the foundation of your change process. And they can help you. They can provide you important educational content, information, training, right, if we've got those champion advocate analysts and reference materials that your organization needs to develop a more analytically-minded culture. MELISSA: 29:13 These are your internal thought leaders, and they can make change stick. Also, do think about your peers. Right? Work quickly as you can to create a cross-functional group of leaders to maintain oversight of analytics, resources, and activities across the business. This kind of analytics council, as I mentioned before, that's going to help enable important data governance activities, as well as identify opportunities and potential issues as your program really gets off the ground. And of course, we cannot forget about our senior executive leadership team. Ensure that the top of the organization and really the company as a whole, understands the benefit of investing time and resources in analytics initiatives. If you have close relationships with senior executives or you are one, yourself, work to start modeling new analytically-driven behaviors such as using data and analytics as a part of your leadership process to be the change that you want to see in the business. MADDIE: 30:35 Got it. Yeah. No. Those are great best practices. And David, question for you. Let's say that I am a leader. I maybe took the analytics maturity assessment. I didn't score what I wanted to, and I'm thinking to myself, "The people who did get a five, what are their teams doing that my team is not doing?" I'd love for you to share your thoughts there. DAVID: 31:00 Thank you for that question, Maddie. First of all, out of the hundreds or thousands of people that have taken this maturity assessment, I have yet to see a five. So maybe they're juicing the system if they have a five. I'm not sure. But when we look at sort of what high-performance teams really do differently, and actually, this is experientially and some research to back this up. I think number one is really align your strategy to business goals. We don't want to do analytics for the sake of doing analytics. We have to understand how is the business going to change its behaviors because of this analytic output. You just say, "Hey, the trend is going up. That's great. And you don't change your business. Why bother?" So make sure it has meaning. Number two, invest in talent and training. So Melissa has covered this. Your analyst, they have the knowledge of the business. And if they don't invest in training, then invest in your talent to make sure they can implement that change you want. Number three, collaboration. You can't do this in isolation. The whole company needs to be behind this. There's cross-functional teams that are needed along every step of the journey. Number four, play strong data and analytics governance processes, you don't want to get in a hot soup. We talked about that earlier. Number five, make sure whatever you do you can establish a way so it's repeatable, so repeatable methodologies. Don't keep reinventing the wheel so you can apply this to your next project. And the last one, number six, is just ensure adoption of this. Make sure this change is sticking. Maddie, you mentioned start with a maturity. Melissa and I actually put a checklist in the back of this book, eight steps to getting started with analytics democratization. Step one is to take the maturity assessment. So thank you for reading all the way to the appendix in our book. Everybody go get a copy and pick it up and follow that checklist. I think it's got some real practical steps. MADDIE: 32:50 Thanks for listening. Links to the analytics maturity assessment and Melissa and David's book, Democratizing Analytics are available on our show notes at community.Alteryx.com/podcast. Catch you next time. [music]

This episode was produced by Maddie Johannsen (@MaddieJ), Mike Cusic (@mikecusic), and Matt Rotundo (@AlteryxMatt). Special thanks to @andyuttley for the theme music track, and @mikecusic for our album artwork.

Alter Everything

Episode Guide

131: How democratization can boost your analytics maturity

Panelists

Topics

Transcript

This episode was produced by Maddie Johannsen (@MaddieJ), Mike Cusic (@mikecusic), and Matt Rotundo (@AlteryxMatt). Special thanks to @andyuttley for the theme music track, and @mikecusic for our album artwork.