Latest / Giant Robots Smashing Into Other Giant Robots / 475: Designing Data Governance From the Ground Up with Lauren Maffeo
Transcript
- VICTORIA: Hey there. It's your host Victoria. And I'm here today with Dawn Delatte and Jordyn Bonds from our Ignite team. We are thrilled to announce the summer 2023 session of our new incubator program. If you have a business idea that involves a web or mobile app, we encourage you to apply for our 8-week program. We'll help you validate the market opportunity, experiment with messaging and product ideas, and move forward with confidence towards an MVP. Learn more and apply at tbot.io/incubator. Dawn and Jordyn, thank you for joining and sharing the news with me today. JORDYN: Thanks for having us. DAWN: Yeah, glad to be here. VICTORIA: So, tell me a little bit more about the incubator program. This will be your second session, right? JORDYN: Indeed. We are just now wrapping up the first session. We had a really great 8 weeks, and we're excited to do it again. VICTORIA: Wonderful. And I think we're going to have the person from your program on a Giant Robots episode soon. JORDYN: Wonderful. VICTORIA: Maybe you can give us a little preview. What were some of your main takeaways from this first round? JORDYN: You know, as ever with early-stage work, it's about identifying your best early adopter market and user persona, and then learning as much as you possibly can about them to inform a roadmap to a product. VICTORIA: What made you decide to start this incubator program this year with thoughtbot? DAWN: We had been doing work with early-stage products and founders, as well as some innovation leads or research and development leads in existing organizations. We had been applying a lot of these processes, like the customer discovery process, Product Design Sprint process to validate new product ideas. And we've been doing that for a really long time. And we've also been noodling on this idea of exploring how we might offer value even sooner to clients that are maybe pre-software product idea. Like many of the initiatives at thoughtbot, it was a little bit experimental for us. We decided to sort of dig into better understanding that market, and seeing how the expertise that we had could be applied in the earlier stage. It's also been a great opportunity for our team to learn and grow. We had Jordyn join our team as Director of Product Strategy. Their experience with having worked at startups and being an early-stage startup founder has been so wonderful for our team to engage with and learn from. And we've been able to offer that value to clients as well. VICTORIA: I love that. So it's for people who have identified a problem, and they think they can come up with a software solution. But they're not quite at the point of being ready to actually build something yet. Is that right? DAWN: Yeah. We've always championed the idea of doing your due diligence around validating the right thing to build. And so that's been a part of the process at thoughtbot for a really long time. But it's always been sort of in the context of building your MVP. So this is going slightly earlier with that idea and saying, what's the next right step for this business? It's really about understanding if there is a market and product opportunity, and then moving into exploring what that opportunity looks like. And then validating that and doing that through user research, and talking to customers, and applying early product and business strategy thinking to the process. VICTORIA: Great. So that probably sets you up for really building the right thing, keeping your overall investment costs lower because you're not wasting time building the wrong thing. And setting you up for that due diligence when you go to investors to say, here's how well I vetted out my idea. Here's the rigor that I applied to building the MVP. JORDYN: Exactly. It's not just about convincing external stakeholders, so that's a key part. You know, maybe it's investors, maybe it's new team members you're looking to hire after the program. It could be anyone. But it's also about convincing yourself. Really, walking down the path of pursuing a startup is not a small undertaking. And we just want to make sure folks are starting with their best foot forward. You know, like Dawn said, let's build the right thing. Let's figure out what that thing is, and then we can think about how to build it right. That's a little quote from a book I really enjoy, by the way. I cannot take credit for that. [laughs] There's this really great book about early-stage validation called The Right It by Alberto Savoia. He was an engineer at Google, started a couple of startups himself, failed in some ways, failed to validate a market opportunity before marching off into building something. And the pain of that caused him to write this book about how to quickly and cheaply validate some market opportunity, market assumptions you might have when you're first starting out. The way he frames that is let's figure out if it's the right it before we build it right. And I just love that book, and I love that framing. You know, if you don't have a market for what you're building, or if they don't understand that they have the pain point you're solving for, it doesn't matter what you build. You got to do that first. And that's really what the focus of this incubator program is. It's that phase of work. Is there a there there? Is there something worth the hard, arduous path of building some software? Is there something there worth walking that path for before you start walking it? VICTORIA: Right. I love that. Well, thank you both so much for coming on and sharing a little bit more about the program. I'm super excited to see what comes out of the first round, and then who gets selected for the second round. So I'm happy to help promote. Any other final takeaways for our listeners today? DAWN: If this sounds intriguing to you, maybe you're at the stage where you're thinking about this process, I definitely encourage people to follow along. We're trying to share as much as we can about this process and this journey for us and our founders. So you can follow along on our blog, on LinkedIn. We're doing a LinkedIn live weekly with the founder in the program. We'll continue to do that with the next founders. And we're really trying to build a community and extend the community, you know, that thoughtbot has built with early-stage founders, so please join us. We'd love to have you. VICTORIA: Wonderful. That's amazing. Thank you both so much. INTRO MUSIC: VICTORIA: This is the Giant Robots Smashing Into Other Giant Robots Podcast, where we explore the design, development, and business of great products. I'm your host, Victoria Guido. And with me today is Lauren Maffeo, Author of Designing Data Governance from the ground up. Lauren, thank you for joining us. LAUREN: Thanks so much for having me, Victoria. I'm excited to be here. VICTORIA: Wonderful. I'm excited to dive right into this topic. But first, maybe just tell me what led you to start writing this book? LAUREN: I was first inspired to write this book by my clients, actually. I was working as a service designer at Steampunk, which is a human-centered design firm serving the federal government. I still do work for Steampunk. And a few years ago, I was working with a client who had a very large database containing millions of unique data points going back several centuries. And I realized throughout the course of my discovery process, which is a big part of human-centered design work, that most of their processes for managing the data in this database were purely manual. There was no DevSecOps integrated into their workflows. These workflows often included several people and took up to a week to complete. And this was an organization that had many data points, as mentioned, in its purview. They also had a large team to manage the data in various ways. But they still really struggled with an overall lack of processes. And really, more importantly, they lacked quality standards for data, which they could then automate throughout their production processes. I realized that even when organizations exist to have data in their purview and to share it with their users, that doesn't necessarily mean that they actually have governance principles that they abide by. And so that led me to really consider, more broadly, the bigger challenges that we see with technology like AI, machine learning, large language models. We know now that there is a big risk of bias within these technologies themselves due to the data. And when I dug deeper, first as a research analyst at Gartner and then as a service designer at Steampunk, I realized that the big challenge that makes this a reality is lack of governance. It's not having the quality standards for deciding how data is fit for use. It's not categorizing your data according to the top domains in your organization that produce data. It's lack of clear ownership regarding who owns which data sets and who is able to make decisions about data. It's not having things like a data destruction policy, which shows people how long you hold on to data for. So that knowledge and seeing firsthand how many organizations struggle with that lack of governance that's what inspired me to write the book itself. And I wanted to write it from the lens of a service designer. I have my own bias towards that, given that I am a practicing service designer. But I do believe that data governance when approached through a design thinking lens, can yield stronger results than if it is that top-down IT approach that many organizations use today unsuccessfully. VICTORIA: So let me play that back a little bit. So, in your experience, organizations that struggle to make the most out of their data have an issue with defining the authority and who has that authority to make decisions, and you refer to that as governance. So that when it comes down to it, if you're building things and you want to say, is this ethical? Is this right? Is this secure? Is it private enough? Someone needs to be responsible [laughs] for answering that. And I love that you're bringing this human-centered design approach into it. LAUREN: Yeah, that's exactly right. And I would say that ownership is a big part of data governance. It is one of the most crucial parts. I have a chapter in my book on data stewards, what they are, the roles they play, and how to select them and get them on board with your data governance vision. The main thing I want to emphasize about data stewardship is that it is not just the technical members of your team. Data scientists, data architects, and engineers can all be exceptional data stewards, especially because they work with the data day in and day out. The challenge I see is that these people typically are not very close to the data, and so they don't have that context for what different data points mean. They might not know offhand what the definitions per data piece are. They might not know the format that the data originates in. That's information that people in non-technical roles tend to possess. And so, data stewardship and governance is not about turning your sales director into a data engineer or having them build ETL pipelines. But it is about having the people who know that data best be in positions where they're able to make decisions about it, to define it, to decide which pieces of metadata are attached to each piece of data. And then those standards are what get automated throughout the DevSecOps process to make better life cycles that produce better-quality data faster, at speed with fewer resources. VICTORIA: So, when we talk about authority, what we really mean is, like, who has enough context to make smart decisions? LAUREN: Who has enough context and also enough expertise? I think a big mistake that we as an industry have made with data management is that we have given the responsibility for all data in an organization to one team, sometimes one person. So, typically, what we've done in the past is we've seen all data in an organization managed by IT. They, as a department, make top-down decisions about who has access to which data, what data definitions exist, where the data catalog lives, if it exists in an organization at all. And that creates a lot of blockers for people if you always have to go through one team or person to get permission to use data. And then, on top of that, the IT team doesn't have the context that your subject matter experts do about the data in their respective divisions. And so it really is about expanding the idea of who owns data and who is in a position of authority to make decisions about it by collaborating across silos. This is very challenging work to do. But I would actually say that for smaller organizations, they might lack the resources in, time, and money, and people to do data governance at scale. But what they can do is start embedding data governance as a core principle into the fabric of their organizations. And ultimately, I think that will power them for success in a way that larger organizations were not able to because there is a lot of technical debt out there when it comes to bad data. And one way to avoid that in the future or to at least mitigate it is to establish data governance standards early on. VICTORIA: Talk me through what your approach would be if you were working with an organization who wants to build-in this into the fabric of how they work. What would be your first steps in engaging with them and identifying where they have needs in part of that discovery process? LAUREN: In human-centered design, the discovery process occurs very early in a project. This is where you are working hand in hand with your client to figure out what their core needs are and how you can help them solve those core needs. And this is important to do because it's not always obvious what those needs are. You might get a contract to work on something very specific, whether it's designing the user interface of a database or it's migrating a website. Those are technical challenges to solve. And those are typically the reason why you get contracted to work with your client. But you still have to do quite a bit of work to figure out what the real ask is there and what is causing the need for them to have hired you in the first place. And so, the first thing I would do if I was walking a client through this is I would start by asking who the most technical senior lead in the organization is. And I would ask how they are managing data today. I think it's really important, to be honest about the state of data in your organization today. The work that we do designing data governance is very forward-thinking in a lot of ways, but you need a foundation to build upon. And I think people need to be honest about the state of that foundation in their organization. So the first thing I would do is find that most-senior data leader who is responsible for making decisions about data and owns the data strategy because that person is tasked with figuring out how to use data in a way that is going to benefit the business writ large. And so, data governance is a big part of what they are tasked to do. And so, in the first instance, what I would do is I would host a workshop with the client where I would ask them to do a few things. They would start by answering two questions: What is my company's mission statement, and how do we use data to fulfill that mission statement? These are very baseline questions. And the first one is so obvious and simple that it might be a little bit off-putting because you're tempted to think, as a senior leader, I already know what my company does. Why do I need to answer it like this? And you need to answer it like this because just like we often get contracts to work on particular technical problems, you'd be surprised by how many senior leaders cannot articulate their company's mission statements. They'll talk to you about their jobs, the tools they use to do their jobs, who they work with on a daily basis. But they still aren't ultimately answering the question of how their job, how the technology they use fulfills a bigger organizational need. And so, without understanding what that organizational need is, you won't be able to articulate how data fulfills that mission. And if you're not able to explain how data fulfills your company's mission, I doubt you can explain which servers your data lives on, which file format it needs to be converted to, who owns which data sets, where they originate, what your DevSecOps processes are. So answering those two questions about the company mission and how data is used to fulfill that mission is the first step. The second thing I would do is ask this senior leader, let's say the chief data officer, to define the data domains within their organization. And when we talk about data domains, we are talking about the areas of the business that are the key areas of interest. This can also be the problem spaces that your organization addresses. It also can have a hand in how your organization is designed as is; in other words, who reports to whom? Do you have sales and marketing within one part of the organization, or are they separate? Do you have customer success as its own wing of the organization separate from product? However your organization is architected, you can draw lines between those different teams, departments, and the domains that your organization works in. And then, most importantly, you want to be looking at who leads each domain and has oversight over the data in that domain. This is a really important aspect of the work because, as mentioned, stewards play a really key role in upholding and executing data governance. You need data stewards across non-technical and technical roles. So defining not just what the data domains are but who leads each domain in a senior role is really important to mapping out who your data stewards will be and to architect your first data governance council. And then, finally, the last thing I would have them do in the first instance is map out a business capability map showing not only what their data domains are but then the sub-domains underneath. So, for example, you have sales, and that can be a business capability. But then, within the sales data domain, you're going to have very different types of sales data. You're going to have quarterly sales, bi-annual sales, inbound leads versus outbound leads. You're going to have very different types of data within that sales data domain. And you want to build those out as much as you possibly can across all of your data domains. If you are a small organization, it's common to have about four to six data domains with subdomains underneath, each of those four to six. But it varies according to each startup and organization and how they are structured. Regardless of how your organization is structured, there's always value in doing those three things. So you start by identifying what your organization does and how data fulfills that goal. You define the core data domains in your organization, including who owns each domain. And then, you take that information about data domains, and you create a capability map showing not just your core data domains but the subdomains underneath because you're going to use all of that information to architect a future data governance program based on what you currently have today. VICTORIA: I think that's a great approach, and it makes a lot of sense. Is that kind of, like, the minimum that people should be doing for a data governance program? Like, what's the essentials to do, like, maybe even your due diligence, say, as a health tech startup company? LAUREN: This is the bare minimum of what I think every organization should do. The specifics of that are different depending on industry, depending on company size, organizational structure. But I wrote this book to be a compass that any organization can use. There's a lot of nuance, especially when we get into the production environment an organization has. There's a lot of nuance there depending on tools, all of that. And so I wanted to write a book that anybody could use regardless of industry size, team size, all of that information. I would say that those are the essential first steps. And I do think that is part of the discovery process is figuring out where you stand today, and no matter how ugly it might be. Because, like we've mentioned, there is more data produced on a daily basis than ever before. And you are not going into this data governance work with a clean slate. You already have work in your organization that you do to manage data. And you really need to know where there are gaps so that you can address those gaps. And so, when we go into the production environment and thinking about what you need to do to be managing data for quality on a regular basis, there are a couple of key things. The first is that you need a plan for how you're going to govern data throughout each lifecycle. So you are very likely not using a piece of data once and never again. You are likely using it through several projects. So you always want to have a plan for governance in production that includes policies on data usage, data archiving, and data destruction. Because you want to make sure that you are fulfilling those principles, whatever they are, throughout each lifecycle because you are managing data as a product. And that brings me to the next thing that I would encourage people working in data governance to consider, which is taking the data mesh principle of managing data as a product. And this is a fundamental mind shift from how big data has been managed in the past, where it was more of a service. There are many detriments to that, given the volume of data that exists today and given how much data environments have changed. So, when we think about data mesh, we're really thinking about four key principles. The first is that you want to manage your data according to specific domains. So you want to be creating a cloud environment that really accounts for the nuance of each data domain. That's why it's so important to define what those data domains are. You're going to not just document what those domains are. You're going to be managing and owning data in a domain-specific way. The second thing is managing data as a product. And so, rather than taking the data as a service approach, you have data stewards who manage their respective data as products within the cloud environment. And so then, for instance, rather than using data about customer interactions in a single business context, you can instead use that data in a range of ways across the organization, and other colleagues can use that data as well. You also want to have data available as a self-service infrastructure. This is really important in data mesh. Because it emphasizes keeping all data on a centralized platform that manages your storage, streaming, pipelines, and anything else, and this is crucial because it prevents data from leaving in disparate systems on various servers. And it also erases or eases the need to build integrations between those different systems and databases. And it also gives each data steward a way to manage their domain data from the same source. And then the last principle for data mesh is ecosystem governance. And really, what we're talking about here is reinforcing the data framework and mission statement that you are using to guide all of your work. It's very common in tech for tech startups to operate according to a bigger vision and according to principles that really establish the rationale for why that startup deserves to exist in the world. And likewise, you want to be doing all of your production work with data according to a bigger framework and mission that you've already shared. And you want to make sure that all of your data is formatted, standardized, and discoverable against equal standards that govern the quality of your data. VICTORIA: That sounds like data is your biggest value as a company and your greatest source of liability [laughs] and in many ways. And, I'm curious, you mentioned just data as a product, if you can talk more about how that fits into how company owners and founders should be thinking about data and the company they're building. LAUREN: So that's a very astute comment about data as a liability. That is absolutely true. And that is one of the reasons why governance is not just nice to have. It's really essential, especially in this day and age. The U.S. has been quite lax when it comes to data privacy and protection standards for U.S. citizens. But I do think that that will change over the next several years. I think U.S. citizens will get more data protections. And that means that organizations are going to have to be more astute about tracking their data and making sure that they are using it in appropriate ways. So, when we're talking to founders who want to consider how to govern data as a product, you're thinking about data stewards taking on the role of product managers and using data in ways that benefits not just them and their respective domains but also giving it context and making it available to the wider business in a way that it was not available before. So if you are architecting your data mesh environment in the cloud, what you might be able to do is create various domains that exist on their own little microservice environments. And so you have all of these different domains that exist in one environment, but then they all connect to this bigger data mesh catalog. And from the catalog, that is where your colleagues across the business can access the data in your domain. Now, you don't want to necessarily give free rein for anybody in your organization to get any data at any time. You might want to establish guardrails for who is able to access which data and what those parameters are. And the data as a product mindset allows you to do that because it gives you, as the data steward/pseudo pm, the autonomy to define how and when your data is used, rather than giving that responsibility to a third-party colleague who does not have that context about the data in your domain. VICTORIA: I like that about really giving the people who have the right context the ability to manage their product and their data within their product. That makes a lot of sense to me. Mid-Roll Ad: As life moves online, bricks-and-mortar businesses are having to adapt to survive. With over 18 years of experience building reliable web products and services, thoughtbot is the technology partner you can trust. We provide the technical expertise to enable your business to adapt and thrive in a changing environment. We start by understanding what’s important to your customers to help you transition to intuitive digital services your customers will trust. We take the time to understand what makes your business great and work fast yet thoroughly to build, test, and validate ideas, helping you discover new customers. Take your business online with design-driven digital acceleration. Find out more at tbot.io/acceleration or click the link in the show notes for this episode. VICTORIA: What is it like to really bring in this culture of design-thinking into an organization that's built a product around data? LAUREN: It can be incredibly hard. I have found that folks really vary in their approach to this type of work. I think many people that I talk to have tried doing data governance to some degree in the past, and, for various reasons, it was not successful. So as a result, they're very hesitant to try again. I think also for many technical leaders, if they're in CIO, CDO, CTO roles, they are not used to design thinking or to doing human-centered design work. That's not the ethos that was part of the tech space for a very long time. It was all about the technology, building what you could, experimenting and tinkering, and then figuring out the user part later. And so this is a real fundamental mindset shift to insist on having a vision for how data benefits your business before you start investing money and people into building different data pipelines and resources. It's also a fundamental shift for everyone in an organization because we, in society writ large, are taught to believe that data is the responsibility of one person or one team. And we just can't afford to think like that anymore. There is too much data produced and ingested on a daily basis for it to fall to one person or one team. And even if you do have a technical team who is most adept at managing the cloud environment, the data architecture, building the new models for things like fraud detection, that's all the purview of maybe one team that is more technical. But that does not mean that the rest of the organization doesn't have a part to play in defining the standards for data that govern everything about the technical environment. And I think a big comparison we can make is to security. Many of us… most of us, even if we work in tech, are not cybersecurity experts. But we also know that employees are the number one cause of breaches at organizations. There's no malintent behind that, but people are most likely to expose company data and cause a breach from within the company itself. And so organizations know that they are responsible for creating not just secure technical environments but educating their employees and their workforce on how to be stewards of security. And so, even at my company, we run constant tests to see who is going to be vulnerable to phishing? Who is going to click on malicious links? They run quarterly tests to assess how healthy we are from a cybersecurity perspective. And if you click on a phishing attempt and you fall for it, you are directed to a self-service education video that you have to complete, going over the aspects of this phishing test, what made it malicious. And then you're taught to educate yourself on what to look for in the future. We really need to be doing something very similar with data. And it doesn't mean that you host a two-hour training and then never talk about data again. You really need to look at ways to weave data governance into the fabric of your organization so that it is not disruptive to anybody's day. It's a natural part of their day, and it is part of working at your organization. Part of your organizational goals include having people serve as data stewards. And you emphasize that stewardship is for everyone, not just the people in the technology side of the business. VICTORIA: I love that. And I think there's something to be said for having more people involved in the data process and how that will impact just the quality of your data and the inclusivity of what you're building to bring those perspectives together. LAUREN: I agree. And that's the real goal. And I think this is, again, something that's actually easier for startups to do because startups are naturally more nimble. They find out what works, what doesn't work. They're willing to try things. They have to be willing to try things. Because, to use a really clichéd phrase, if they're not innovating, then they're going to get stale and go out of business. But the other benefit that I think startups have when they're doing this work is the small size. Yes, you don't have the budget or team size of a company like JP Morgan, that is enormous, or a big bank. But you still have an opportunity to really design a culture, an organizational culture that puts data first, regardless of role. And then you can architect the structure of every role according to that vision. And I think that's a really exciting opportunity for companies, especially if they are selling data or already giving data as a product in some way. If they're selling, you know, data as a product services, this is a really great approach and a unique approach to solving data governance and making it everyone's opportunity to grow their own roles and work smarter. VICTORIA: Right. And when it's really the core of your business, it makes sense to pay more attention to that area [laughs]. It's what makes it worthwhile. It's what makes potential investors know that you're a real company who takes things seriously. [laughs] LAUREN: That's true. That's very true. VICTORIA: I'm thinking, what questions...do you have any questions for me? LAUREN: I'm curious to know, when you talk to thoughtbot clients, what are the main aspects of data that they struggle with? I hear a variety of reasons for data struggles when I talk to clients, when I talk to people on the tech side, either as engineers or architects. I'm curious to hear what the thoughtbot community struggles with the most when it comes to managing big data. VICTORIA: I think, in my experience, in the last less than a year that I've been with thoughtbot, one challenge which is sort of related to data...but I think for many small companies or startups they don't really have an IT department per se. So, like, what you mentioned early on in the discovery process as, like, who is the most senior technical person on your team? And that person may have little to no experience managing an IT operations group. I think it's really bringing consulting from the ground up for an organization on IT operations, data management, user and access management. Those types of policies might just be something they hadn't considered before because it's not in their background and experience. But maybe once they've gotten set up, I think the other interesting part that happens is sometimes there's just data that's just not being managed at all. And there are processes and bits and pieces of code in app that no one really knows what they are, who they're used for, [laughs] where the data goes. And then, you know, the connections between data. So everything that you're mentioning that could happen when you don't do data governance, where it can slow down deployment processes. It can mean that you're giving access to people who maybe shouldn't have access to production data. It can mean that you have vulnerabilities in your infrastructure. That means someone could have compromised your data already, and you just don't know about it. Just some of the issues that we see related to data across the spectrum of people in their lifecycle of their startups. LAUREN: That makes total sense, I think, especially when you are in a startup. If you're going by the typical startup model, you have that business-minded founder, and then you likely have a more technical co-founder. But we, I think, make the assumption that if you are, quote, unquote, "technical," you, therefore, know how to do anything and everything about every system, every framework, every type of cloud environment. And we all know that that's just not the case. And so it's easy to try to find the Chief Technology Officer or the Chief Information Officer if one exists and to think, oh, this is the right person for the job. And they might be the most qualified person given the context, but that still doesn't mean that they have experience doing this work. The reality is that very few people today have deep hands-on experience making decisions about data with the volume that we see today. And so it's a new frontier for many people. And then, on top of that, like you said as well, it's really difficult to know where your data lives and to track it. And the amount of work that goes into answering those very basic questions is enormous. And that's why documentation is so important. That's why data lineage in your architecture is so important. It really gives you a snapshot of which data lives where, how it's used. And that is invaluable in terms of reducing technical debt. VICTORIA: I agree. And I wonder if you have any tips for people facilitating conversations in their organization about data governance. What would you tell them to make it less scary and more fun, more appealing to work on? LAUREN: I both love and hate the term data governance. Because it's a word that you say, and whether you are technical or not, many people tune out as soon as they hear it because it is, in a way, a scary word. It makes people think purely of compliance, of being told what they can't do. And that can be a real challenge for folks. So I would say that if you are tasked with making a data governance program across your organization, you have to invest in making it real for people. You have to sell them on stewardship by articulating what folks will gain from serving as stewards. I think that's really critical because we are going to be asking folks to join a cause that they're not going to understand why it affects them or why it benefits them at first. And so it's really your job to articulate not only the benefits to them of helping to set up this data stewardship work but also articulating how data governance will help them get better at their jobs. I also think you have to create a culture where you are not only encouraging people to work across party lines, so to speak, to work across silos but to reward them for doing so. You are, especially in the early months, asking a lot of people who join your data stewardship initiatives and your data governance council you're asking them to build something from the ground up, and that's not easy work. So I think any opportunity you can come up with to reward stewards in the form of bonuses or in terms of giving them more leeway to do their jobs more of a title bump than they might have had otherwise. Giving them formal recognition for their contributions to data governance is really essential as well. Because then they see that they are rewarded for contributing to the thought leadership that helps the data governance move forward. VICTORIA: I'm curious, what is your favorite way to be rewarded at work, Lauren? LAUREN: So I am a words person. When we talk about love languages, one of them is words of affirmation. And I would say that is the best way to quote, unquote, "reward me." I save emails and screenshots of text messages and emails that have really meant a lot to me. If someone sends me a handwritten card that really strikes a chord, I will save that card for years. My refrigerator is filled with holiday cards and birthday cards, even from years past. And so any way to recognize people for the job they're doing and to let someone know that they're seen, and their work is seen and valued really resonates with me. I think this is especially important in remote environments because I love working from home, and I am at home alone all day. And so, especially if you are the only person of your kind, of your role on your team, it's very easy to feel insular and to wonder if you're hitting the mark, if you're doing a good job. I think recognition, whether verbally or on Slack, of a job well done it really resonates with me. And that's a great way to feel rewarded. VICTORIA: I love that. And being fully remote with thoughtbot, I can feel that as well. We have a big culture of recognizing people. At least weekly, we do 15Five as a tool to kind of give people high-fives across the company. LAUREN: Yep, Steampunk does...we use Lattice. And people can submit praise and recognition for their colleagues in Lattice. And it's hooked up to Slack. And so then, when someone submits positive feedback or a kudos to a colleague in Lattice, then everyone sees it in Slack. And I think that's a great way to boost morale and give people a little visibility that they might not have gotten otherwise, especially because we also do consulting work. So we are knee-deep in our projects on a daily basis, and we don't always see or know what our colleagues are working on. So little things like that go a long way towards making people feel recognized and valued as part of a bigger company. But I'm also curious, Victoria, what's your favorite way to get rewarded and recognized at work? VICTORIA: I think I also like the verbal. I feel like I like giving high-fives more than I like receiving them. But sometimes also, like, working at thoughtbot, there are just so many amazing people who help me all throughout the day. I start writing them, and then I'm like, well, I have to also thank this person, and then this person. And then I just get overwhelmed. [laughs] So I'm trying to do more often so I don't have a backlog of them throughout the week and then get overwhelmed on Friday. LAUREN: I think that's a great way to do it, and I think it's especially important when you're in a leadership role. Something that I'm realizing more and more as I progress in my career is that the more senior you are, the more your morale and attitude sets the tone for the rest of the team. And that's why I think if you are in a position to lead data governance, your approach to it is so crucial to success. Because you really have to get people on board with something that they might not understand at first, that they might resent it first. This is work that seems simple on the surface, but it's actually very difficult. The technology is easy. The people are what's hard. And you really have to come in, I think, emphasizing to your data stewards and your broader organization, not just what governance is, because, frankly, a lot of people don't care. But you really have to make it tangible for them. And you have to help them see that governance affects everyone, and everyone can have a hand in co-creating it through shared standards. I think there's a lot to be learned from the open-source community in this regard. The open-source community, more than any other I can think of, is the model of self-governance. It does not mean that it's perfect. But it does mean that people from all roles, backgrounds have a shared mission to build something from nothing and to make it an initiative that other people will benefit from. And I think that attitude is really well-positioned for success with data governance. VICTORIA: I love that. And great points all around on how data governance can really impact an organization. Are there any final takeaways for our listeners? LAUREN: The biggest takeaway I would say is to be thoughtful about how you roll out data governance in your organization. But don't be scared if your organization is small. Again, it's very common for people to think my business is too small to really implement governance. We don't have the budget for, you know, the AWS environment we might need. Or we don't have the right number of people to serve as stewards. We don't actually have many data domains yet because we're so new. And I would say start with what you have. If you are a business in today's day and age, I guarantee that you have enough data in your possession to start building out a data governance program that is thoughtful and mission-oriented. And I would really encourage everyone to do that, regardless of how big your organization is. And then the other takeaway I would say is, if you remember nothing else about data governance, I would say to remember that you automate your standards. Your standards for data quality, data destruction, data usage are not divorced from your technical team's production environments; it's the exact opposite. Your standards should govern your environment, and they should be a lighthouse when you are doing that work. And so you always want to try to integrate your standards into your production environment, into your ETL pipelines, into your DevSecOps. That is where the magic happens. Keeping them siloed won't work. And so I'd love for people, if you really enjoyed this episode and the conversation resonated with you, too, get a copy of the book. It is my first book. And I was really excited to work with the Pragmatic Programmers on it. So if readers go to pragprog.com, they can get a copy of the book directly through the publisher. But the book is also available at Target, Barnes & Noble, Amazon, and local bookstores. So I am very grateful as a first-time author for any and all support. And I would really also love to hear from thoughtbot clients and podcast listeners what you thought of the book because version two is not out of the question. VICTORIA: Well, looking forward to it. Thank you again so much, Lauren, for joining us today. You can subscribe to the show and find notes along with a complete transcript for this episode at giantrobots.fm. If you have questions or comments, email us at hosts@giantrobots.fm. And you can find me on Twitter @victori_ousg. This podcast is brought to you by thoughtbot and produced and edited by Mandy Moore. ANNOUNCER: This podcast is brought to you by thoughtbot, your expert strategy, design, development, and product management partner. We bring digital products from idea to success and teach you how because we care. Learn more at thoughtbot.com.