Al and Benn Stancil discuss Mode, data, leadership and analytics

UPS Go.

You're listening to Making Data simple, where we make the world of data effortless, relevant, and yes, even fun.

Podcast listeners. Al Martin here. Thank you. Appreciate you showing back up again. Hopefully you're enjoying podcasts and welcome to Making Data Simple. Today my guest is Ben Stancil. We're going to have a topic or topics of data leadership analytics. Ben is the Chief Analytics Officer at Mood and if I could give a little introduction and then I'll turn it over to you, Ben. He's an accomplished data analyst with deep expertise in collaborative business intelligence and interactive data science. Again, he's the co founder, president and Chief Analytics Officer for Mode. He combines the best elements of business intelligence, data science, and machine learning. He's held analytics positions at Microsoft and Yammer and worked as a researcher for the International Economics Program at the Carnegie Endowment for an International Peace. That's a mouthful. So welcome. How you been doing? Good, good, good, yeah.

Excited to be here.

Where are you located?

Brooklyn.

Brooklyn, Nice. If you wouldn't mind, give us an introduction on your experience and what brings you to the Chief Analytics Officer at moog.

Basically started my career, as you mentioned, at Carnegie Endowment for International Peace. Very long name. It is a mouthful for sure. My role there was doing economics research for policy, essentially. And so the way that think tanks work is there's this kind of bridge between academia and policymakers where you are trying to provide some kind of academic backing to different policy ideas. I was there in the height of the financial crisis in 2009, so we were essentially writing a bunch of papers about the Fed should do this, policymakers should do that in response to that crisis. It was a very different world than what we traditionally think of as like analysts and data scientists and especially in kind of the tech and business worlds, but was shaped kind of similarly. Like the idea was basically look at data, come up with advice, make recommendations to decision makers, and kind of make your case for what it is that people should be doing. The difference was we were writing like the advice for what people in Congress should do who are taking lots of advice. Most of it is not from think tanks. Most of it is from like, you know, sort of the political operatives that are kind of guiding how they make decisions. And so I like enjoyed the work and the process of thinking about those problems, but didn't necessarily like the idea that I was 10 steps removed from anything actually happening. You're sort of just yelling into the void. I left that job to a startup doing a more traditional, just like analytics role on this company Called Yammer. And not long after I got there it was acquired by Microsoft. So I did like analytics, sort of internal company analytics, slash data science, kind of whatever you want to call it for a couple years there and then met a couple folks who we had built some internal tools to enable that team to. The three of us left to go to Start Mode which was building kind of a productized version of that and then that from joining there. Like my background was as an analyst you can make up whatever title you want when you start a company. And that was the one that seemed to make the most sense. And my role at Mode has been pretty varied but has always been centered around the idea of like okay, how do we make sense of the data we have? How do we think about the problem? Trying to solve that kind of thing.

Very nice. So tell me, did anything actually happen? I mean I've always wondered that about think tanks. Meaning I know it's got to be frustrating. I mean whoever has got the money, I got to believe in most cases. But think tanks actually have the data and the hopefully non biased feedback. But anything you could put your finger.

On the one that I worked for, they had programs in a number of areas. There were a couple that were ones that really mattered and they tended to be kind of wonkier things that they weren't the sort of things that were driving like the political conversation. One of the programs that Carnegie had that it was well known for was a nuclear non proliferation program which is also a mouthful. It was basically like trying to convene leaders in sort of nuclear. The nuclear weapons space to figure out how do we make sure that we are like disarming nuclear arsenals and things like that. In that it mattered because it was this very, you know, there aren't that many people in the world I think about that sort of thing. Carnegie had some of the people who really mattered. They were able to get the people together. Did policies change maybe a little bit on the edges? Certainly I think it does. But that was people who, who were really making those decisions and stuff like that. On the stuff I was doing which like economic policy in the height of financial crisis. No, it does not matter in those cases think tanks are a little bit indulgent where they are people typically kind of on the edge of retirement who want a place to sort of have a soft landing where they get to go and hang out with their friends and talk on panels in D.C. and kind of pontificate about the kind of academic problems but aren't actually really doing a Whole lot to change it.

Just switching gears ever so slightly. Did you ever go to Capitol Hill to present any of your concepts?

I did not. I was like, you know, early 20s at this point. Nobody's asking me questions. The people who were the scholars. Think tanks have a kind of weird structure where they're basically. There is no middle management. It's a bunch of young people right out of college, right out of grad school, who are kind of research fellows, which is what I was doing. And then it's a bunch of again, very senior people who are kind of the face of the think tank. Those folks very much did go to Capitol Hill fairly regularly and they would be the kinds of people that are not in the hearings that you would see on tv, but the kind of hearings that's just kind of happening for the committee that there's three of the members are there and 10 of them are not. And the interview about some random, very wonky thing about a bill that matters a lot to a lot of people, but nobody's really paying attention to the people I worked with had to occasionally do that about. It would be things like bank capital requirements. I think this was when they were doing some of the Dodd Frank legislation. It would be things about some of the minutiae of that kind of bill.

Well, it's interesting though. I mean. Well, very interesting. It sounds like it's interesting work. And you've got a little bit of a political background in here, huh?

Very interesting. You're working on interesting problems. And my role was because it was an economics program, it was very data oriented and I liked thinking about problems in that way. It was just difficult to do it. Sort of realizing that the specific thing that I do doesn't matter. I removed from any conversation and stuff like that. So I think that was sort of where some of the motivation came from to go into something else. But I wanted to continue thinking about those sort of problems, thinking about problems in that kind of way. And so that's why sort of the tech analyst actually ended up being a little bit more of a natural fit that it may look like on the surface.

That's kind of what the second reason this interests me to get that background is because I see a trend. I don't know this is anything new really, but there's a trend of going from economics to data analytics. And it looks like you made that transition. And it seems pretty organic the way you're talking that hey went to analytics. It just was a natural transition. I don't know. It always seems natural at least from my standpoint, economics to there was no data engineering, et cetera. Tell me how you got from point A to point B.

It partly was that I went in that job when I started to realize I wanted to leave. And think tanks are sort of like that again because there is no sort of middle management. Nobody gets permanent promoted up through think tanks. It's not really a career. It's sort of like a go learn some stuff and then you're expected to move on. So it was like I joined it with the expectation that I would not be there that long. When I was looking for new jobs, I basically wanted to do something that I thought would be. I liked the econ, I liked the math, the data side of things. I like thinking about problems in that way. One sort of natural direction that people could go in that is finance, which I think would have been interesting. And I talked to some finance folks. Another direction is basically just management consulting, which had sort of less appeal of for. I knew people who'd gone to like the McKinsey's and Baines and it's like, you know, you sort of get ground away flying out to Cincinnati to help some, you know, random company make a bunch of decisions about who they're going to fire. And I had a couple friends who were from the Bay Area or who had moved out there for jobs in tech. And so I was like, they all really liked it. They all basically said it was fun and it was a good time. You know, the jobs were rewarding, they sort of all the things you were looking for. And so I actually started looking for jobs through some of those networks of people I knew. Like, okay, what do you know who do you. That kind of stuff. I mean, that's how I landed the job at the company that I got was I had a friend whose sister worked there and was like, put me in touch with a recruiter and then you kind of end up having conversations that way and that sort of stuff. So I wanted to work on the same. Like I knew again, the data stuff was interesting to me. I wasn't terribly focused in exactly what because I didn't know the space very well. And especially if you're not in any kind of tech world, like it all seems very foreign. I was basically just like applying to jobs, trying to figure out what they meant and eventually had enough of a conversation to get a little bit of a sense to the point where I was like, oh, this could be pretty interesting. But I honestly didn't have a great sense of what that job was going to be like until I got there, I like understood the basics of oh, we're trying to help you make decisions. But it wasn't until like I was working there that I really understood what, what the point of the role was.

I don't think then you arrive ultimately as a co founder of Mode. So tell us about Moat.

So Mode is a analytics tool for analysts and data scientists to be able to create and share work. And it works as basically like a platform for those folks to be able to do the analysis they need to answer questions and then be able to share reports, dashboards, PI tools, essentially with other stakeholders on the business. So the idea is if I'm, if I'm a data team, I probably get a lot of questions. Some of those questions I want to answer with the stakeholder directly is like, all right, let's figure this thing out together and kind of a deeper research. Some of those questions are questions I want to be able to build you a tool so that you can answer it, like a BI tool or a dashboard or whatever. And Mode is a tool that supports all those things. So it's essentially a way for. It's like the tool that the data team lives in for either distributing other tools to be around the company or for being able to answer the kind of strategic questions that they get asked. You know, when a CEO is asking like which product we build next or which market we expand into or those sorts of things.

Does Mode have its own analytics built in or is it a platform that integrates other BI tools?

It is a BI tool itself. We can integrate with other pieces of the stack. We don't integrate for instance with like tableau because we typically people choose to use us or tableau for reasons. Sometimes people choose to use both or whatever. And like that's a whole other conversation. But for the most part we see it as, we aren't trying to be a platform that sits underneath BI tools. We are the way in which you consume data.

How do you define platform then? That's where I went with that because when I was thinking platform, I thought maybe it's more integration of further tools. Your definition of platform?

Well, there are probably two versions of that. There's like the Silicon Valley marketing messaging platform, which every product is a platform because it sounds better than product.

That's why I ask.

Yeah, you launch a company, you're supposed to be the next platform for a thing. And so I think it's like a little bit of just a catch all term for product. Basically the probably proper definition Is is something about like you are facilitating other people building something on top of it. I think there was a Ben Thompson definition of this that was something along the lines of the value of the ecosystem is considerably more valuable than your particular service. Or I think it's like a Microsoft bit in there somewhere. But yeah, Mode is not that in the sense that it's not like an app platform that people build stuff on top of it. It's much more of a. It is a tool that brings together a bunch of other analytics products into a sort of a centralized place. Like, but we have built them. So it's a platform in the sense that like the G Suite is kind of a platform for a bunch of different types of products. Mode is sort of similar to that, but it's not like a platform. And probably the.

So it'd be similar to Tableau or.

Cognos, similar in sort of structure. We focus on more modern data teams. Data teams that are more sort of code oriented, data teams that are writing SQL, Python, those kinds of things as much as the ones that are doing sort of the traditional BI work. But it is in the same vein as those.

So I'll come back to Mode. But you're the Chief Analytics Officer. No, we got data people that are listening to this podcast. But as the name suggests, I try to keep it simple. What is your definition of the Chief Analytics Officer? I mean, what do you do every day?

Three of us who started it, there was our CEO, there was our technical co founder. I know it was me. Technical founder's job was to build products. CEOs job was to go out and do all of like the talk to potential customers, talk to investors, be the face of the company. My job was as someone who was not an engineer and was not out doing like the fundraising stuff, it's kind of like why are you here? Which is a reasonable question. But my background was in analytics and so my role has been fairly varied. It has essentially been moving around to different things that are. What is it that we need someone who understands the customer to do? In some cases that was just support. In some cases that was leading the product organization. In some cases it's been leading marketing. So my day to day has been essentially doing those things where we need like a voice of a customer. We need someone who understands the way that they work. So in that case, I don't think it is reflective of a. Whatever a Chief Analytics Officer title typically is. To me, the more general title. And this is kind of where my job is headed to some Extent is one of two things. It is either the person who is the highest up data person in the organization. So essentially managing data science teams, maybe data engineering teams, kind of responsible for thinking about data infrastructure, how a company basically uses data as an asset. The other version of it is sometimes it is more of a CTO shaped thing where it's kind of okay, you are thinking about the sort of furthest forward analytics problems, trying to be kind of the executive teams analyst, thinking about like the strategic stuff that the company needs to be thinking about that you want a high level sort of analyst to be focused on. The former is much more common, but I think we're starting to see a little bit more of the latter.

How big is MO today? I mean it's a true startup, right?

It's close to 200 now.

Pretty good. It must be growing fast. I'm going to come back to mode in a second. But now as you described the chief analytics officer, the interesting thing about your answer there is you have some very specific points on what you think the I don't know if the analytics officer should be or at least your career in analytics should be because you wrote an article on Substack about the role of a chief analytics officer and you talked about rising up through the ranks of the organization. A lot of times the right level of data analytics is not sitting at the table with the C suite and you had some very specific point of view there. Could you talk to that?

Analytics and data science is still a relatively new field. It's something that it has a history of business analysts and BI and those sorts of things. But as the field that has become kind of its modern incarnation, it's relatively new. It's on the order of 10, 15 years old. We don't have a great sense of what those career trajectories look like. If you are someone who gets a data scientist job relatively young, like where do you go from there? Like where do you end up in 30 years? There's not really a great answer for that. We're in other fields. We do, you know, if you're a marketer, you basically rise up to be a cmo. If you're there's another C level places you could end up or whatever. For all these other jobs for analytics, it's kind of like where do you go? It's like okay, you become an analyst, a senior analyst, you become like the leader of a team. Sometimes you'll get to a vp which is usually this kind of tweener between sometimes they're on an exec team, sometimes they're like a step below. They report to the CFO or cto, your exec manager kind of thing, at that level. There are two problems with that to me. One is that essentially it forces the career path to be management and like executive leadership, which I don't think is necessarily appropriate for an analytics role. Kind of similar to how engineers have career paths where they can become principal engineers, they can become staff engineers, where they aren't managing teams, but are just like super senior engineers building sort of the hardest technology, solving the most complicated technical problems. But still, as an IC who's very senior, you don't have that path. As an analyst, what ends up happening is like the exec team and the leadership teams need the same level of analytical support as every other team. If you have an analyst that's working with the marketing team to figure out, you know, how do we attribute ad spend to signups? You have an analyst that works on that. If you're the executive team, you're trying to make a decision about, like, should we acquire this company? Which is often a decision that is closed off to the exec team. It's not something that's shared very widely. Like it has certain privacy and security concerns around it. Who do you go to? Like, you don't have that person that's like the person thinking about those strategic problems. And so I think to me that that means there's a gap where exact teams should have a role like that, that you could rise as an analyst, where your job is to basically work on the very biggest problems of the business, but not leadership and people, management and organizational management, but more an analyst, but with a much wider and more strategic purview. And the analogy to me, this is like a CTO. So sometimes CTOs and startups are managing the technical organization, but more commonly CTOs are kind of like the head of a research division kind of thing where they run. Okay, how are we thinking about technology going forward? What are the technological trends we need to be paying attention to? How do we architect our products around those trends? They aren't responsible for running an entire organization. And just like making sure a team hits okrs, they're more responsible for like thinking about these things at a strategic level. And I think the same parallel could exist for analysts.

So does that mean to say that your article, your blog that you wrote, you believe there should be a VP responsible for managing the team's daily operations, and then the chief analytics officer that's sitting at the C suite table is at the bottom line or what?

Basically, yeah. I mean there's, there are reasons why there's like to be a little bit suspicious of that of like how many people do you have in the C suite? Is that too many people? And I think those are fair questions, but it's not to me it's like we should be starting to grapple with those. And that's not a totally unprecedented thing. The VP of Engineering and CTO are often that same thing where a senior VP of engineering might be the person who oversees the entire engineering org and the CTO is. Is also kind of on their own own set of problems that is. Is not necessarily a part of that org. So it would be sort of two people who are thinking about these problems. But if companies want to invest as much in data as they say they do and sort of care as much about that doesn't seem like a reach to me.

Are you saying you'd rather prefer those suit positions than a CTO or in addition to.

In addition to. I mean I think it's. I don't actually know that the VP of data is an exec level position necessarily. It depends on the organization. In some cases it may be, in some cases it not. You know like is a VP of Product an exec level position? Sometimes they roll up to an entire like product design engineering leader. In some cases the organization is so product centric that they have like a chief product officer that is very much an exec level position. So I think that can vary org to org. And it may be that whatever this chief analytics officer's type of job is isn't actually an exec level. Like they're just a step below or whatever. But I think it's valuable to have someone who is very senior who can be a part of sensitive conversations, who can just be experienced in solving these data problems, who's not also trying to juggle Organizational management isn't responsible for all of the operational data people management stuff. Yeah, that. That makes it. You can't do both. You can't be a strategic analyst. Plus running an organization like you have to choose one.

I think today most organizations look at the CDO to do both. Yes.

So I think that's the problem is like they do and I think to me the, the CDO essentially becomes the analyst of last resort for the exec team where they're primarily responsible for the organizational side of things. They're primarily responsible for figuring out how what's our strategies for basically data technology for managing an analytics team for data science teams. Like what are our sort of long term visions for where those things go? And they're basically the person responsible for like executing all that. But if the CEO has a question and is like we have to make this big decision again about say we're trying to acquire a company, they probably go to the CDO and say, okay, like help me make this decision. And that person may end up punting it to somebody else. They may delegate it to like a lower level analyst, I don't know. But it's still the CDO's responsibility to kind of now be the person responsible, like trying to figure that out. And to me it's like again, the VP of engineering and CTO parallel is there where you don't want the person running the engineering Org also to be like, hey, we need to do this big technical rearchitecture. Can you go do the investigation on that re architecture as well? Like they might be in the room when those conversations are happening, but they can't own that. And having someone who is senior left to be able to own that to me is a beneficial thing for a cto. It's also a very beneficial thing on the analytics side.

And I was reading the data mirror the other day and they were talking about in 2012, I think 12% of companies had a CDO. 2018 it was like 70%, which I don't know that I believe. That seems awfully high to me. Awfully high. And they say by 2025, 90%. So I was going to say that, look, I just think many organizations need to start with the CDO and then work from there because don't even have that. But I don't know, I think if 70% then we're making good progress. I just don't see that when I'm talking with clients myself for the most part. Yeah.

And that 70% seems very high. I would see there being 70% of people who have like some data person who may also be like an IT leader. That's kind of the person whose job it is is to say, okay, we've delegated data problems to this person.

I could see that you're right where they just say, hey, you're the CDO today.

Yeah, exactly.

Makes sense. Let's come back to Mode now. I represent IBM. We've got a product called Cognos. We're pretty proud of that. There's tableau out there. You guys have created Mode. Why do we need another BI tool? What's the differentiation you guys bring?

So there's a few things. One of the big broad strokes version of it is essentially stuff moving to the cloud. There's a new wave of data tools. Lots of them are cloud based. They have to work with other cloud based data tools. This kind of new, the modern data stack, as people say, sort of ushering in a whole series of new tools that work in slightly different ways than old BI tools or any kind of legacy data tooling. That doesn't mean those legacy tools can't work with the same stuff. There is some like circle of life, Silicon Valley type of stuff around. Hey, there's all these big sort of megatrends about moving the cloud, about the things happening with warehouses, the sort of rise of redshift and snowflake and bigquery and those folks, what are the things that change in the ecosystem because of that? And so in that sense, Mode is part of that kind of wave of tooling that is all right, we want tools that are more sort of consumer inspired, that are cloud apps, easy to sign up for, easy to distribute, easy to share, easy to be collaborative, all that kind of stuff. The other side of that and the more specific side to Mode is I took a job at a tech company 10 years ago, basically at a data team to help people solve problems. We weren't solving the traditional problems that BI tools would serve. We weren't just trying to build sort of reports and sort of financial reporting and that sort of stuff. Like we were sitting next to the PMs answering the hard strategic questions about like which products we build and things like that. You can do those things in BI tools, but often they are this mix of kind of analytics and data science that you're actually trying to do where you're trying to like you need a lot of granular access to data, you need like code to be very flexible. You write a lot of SQL, you write a lot of Python, you write a lot of rubber. You're not doing reporting, you're doing this kind of research and analysis. And that is something that is exploded. Like that is now sort of. The new data team is often shaped like that. Like there's a lot of pressure on people to have these analysts whose job it is to help people make decisions. And in those cases you need different tooling than you would get out of traditional bi. You know, you need something that is very flexible, that can work with data where it sits in databases, something that has the power of, of these scripting languages. And so Moat is aiming to serve those folks, doesn't necessarily replace the traditional BI needs. It essentially just like the new sort of demand that a lot of data teams have for the kind of tools that need them.

What would be your secret sauce, your moat or your sweet spot? I mean, basically what I hear you saying is that, look, we're targeting the modern stack. We're targeting cloud. We want to work with data where it sits, flexible. We want to really not just do reporting, but research and analysis to help people make decisions. I got that. I would imagine if, you know, if I talk to my cognos buddies or I talk to my tab buddies, they'll say, yeah, we do that. So what would you say? Hey, this mo. But yeah, the difference is mo does X, Y or Z?

There are two things, three, I guess, that potentially make it different. One is that. And the big. And the sort of secret sauce that we had in the beginning was basically, it is something that understands how analysts work, that it fits their workflow such that it gets out of the way rather than getting into it. And so an example of this is one of the patterns that we would often see is somebody would see a dashboard, say it's on tableau. Something is wrong with the dashboard. Like, there's a number that's down. An exec is freaking out. Some CEOs like, what happened here? Why did nobody sign up yesterday? Somebody fix this. We're like, is this real? Should we panic? The dashboard doesn't tell you that question enough. Like you can kind of poke at it and do the sort of drill downs and things you would do, but you don't get to the root of the problem. As a result, the analyst is like, all right, I have to go do this deeper analysis. They end up connecting to a database, writing some SQL queries, putting data into Excel, making charts on Excel. Sometimes they might do something. They say, hey, this would be easier to do in Python. So they export the results into a jupyter notebook. They do something to figure out what they think is happening. They then take some screenshots into some Excel files and say, hey, here's in an email what we think is going on. The CEO or whoever is like, oh, this makes sense. This doesn't make sense. Go back and forth. You have this kind of iterative back and forth process until you figure out what's going on and then some decision gets made and you fix the dashboard, or if the dashboard wasn't broken, you just move on and that's the problem solved. Within all of that, you end up bouncing around between a ton of different tools. You end up being like, all right, as an analyst I have to write a SQL query here, I got to make visualizations here, I got to share it here. I got to take these screenshots. Mode really tries to streamline that. And it tries to streamline it in a way where it makes it very easy to go from that dashboard to the SQL query, to the visualization, to the Python notebook, to all these things such that you can choose those tools at the right time when you need them, rather than having to say, like, I have to open up a new thing. And so in that sense, like the secret sauce is understanding that there are these workflows for which BI tools don't really help because you have to end up stepping outside of it. And so anytime you step outside of it, it's just like this friction that's. That's very painful.

Is there another way of saying that that mode targets the developer Persona rather than the data analysts itself? Or is that not so those?

I think that is one of the sort of things, I think, like the trends I was saying that is sort of the bigger trend is sort of cloud stuff. There is increasingly little difference between those folks. To me, it depends. Like, a lot of analysts now are increasingly people who write SQL and are developers in that sense. They aren't engineers, they aren't people who are building apps, but they are people who, who want to work in code. They want to write SQL, they want to write Python. They're not sort of the traditional, like business intelligence analyst who is kind of an Excel user and Excel user extended. They're people who are much more comfortable in code and things like that. And so there's, there's this now kind of like blending space in the middle. And Mode very much targets those folks. They're not. We're not selling to somebody who wants to write Java, but we're selling to somebody who wants to write SQL and Python and then ultimately selling to their customers. Because when they do that work, they have to distribute out to the rest of the organization so that people can like make sense of those or whatever work it is that they're, you know.

Trying to answer why the name Mode, where'd that come from?

Mode analytics.com was available when we started it. The answer was we wanted something where we could get a website that wasn't like Modly, IO or whatever. We wanted something, so we wanted a dot com. We wanted a name that was spelled correctly. There are a lot of companies, like, especially at the time when we started it, there were a lot of companies that would basically like spell things a little Weird. And we didn't want to always have to tell people, like, oh, it's spelled Mode, but it's spelled M O, I, D, E. It was like a word that was. We could get the domain. It was spelled correctly and it sounded vaguely mathematical and kind of fit into the. To the vibe that we wanted to create.

Can you speak a little bit more about the technology? I mean, what's underneath the hood? Part of that question is, if I'm a client, how do I use the product? I mean, you've kind of. Again, I think you referred to it, but let's hit it again if you don't mind, because I think it's worth.

There are two kind of broader categories of technology that I would say that go into Node. One is essentially glue. Like, we have solved problems that aren't technically that difficult, but they're painful to solve. And they're the sorts of things that bring together a bunch of things into one place. So one of those examples is like database connectivity, connecting a bunch of databases. It's not that hard to connect to one, but it's hard to connect to a bunch in a way where it scales and where you like, do handle different things the way like script works and how databases work correctly. And if you want a tool that can sit on top of all of your data, you have to connect all these things. It provides a nice way of saying, great, we have connectivity, all these things, but it glues these things together in a way you couldn't otherwise. So Mode has built a number of things like that that are nice workflows. Between connecting the data to writing SQL queries and kind of a web IDE to a notebook, to visualizations that are all kind of like components that themselves may not look like something that's sort of new tech. But bringing them all together is a much longer and sort of harder technical problem than it may seem. The second piece of technology that is like much more of the real, like proprietary moat is we've also invested a lot in basically visualization tech. The space between what most BI tools can do with visualizations and what say, tableau can do is very, very big. Tableau has a lot of foundational tech underneath the way that they've actually set up their. Their visualization technology. We have built things to aim to build sort of the similar remote where we think great visualization technology is very important. We think it can make a really big difference in the experience for customers. It's an investment. It's something you have to be willing to make a bet on and something that we have made a bet on. And so I think that's a place where, you know, if you want complex and rich visualizations, you have to build that foundational technology. And that's another place where we've invested where a lot of other tools do.

You connect across like what I would call a hybrid cloud, in other words, multiple clouds within your tool and or even on prem for the most part.

Yes, there are. One of the challenges with connectivity is there's always going to be somebody who has some particular weird configuration where you're like, okay, that's the one that's not going to work. But yes, we can connect to anything, kind of wherever it sits, have ways to connect to cloud infrastructure, have ways to connect to on prem infrastructure. There are the people who have buried their server in a basement somewhere that isn't connected to the Internet in any way. And those we cannot connect to. But outside of that we can connect to most things.

So that leads me to what is your monetization strategy or your go to market? Is it SaaS? Is it as a service?

It's SaaS. So we sell. Most of Moat's revenue comes from selling seats. So say that you're a 100 person company and you've got 30 people who want to use mode. We'll sell, you know, a set number of seats for having that you're basically buying for the number of people who are using. Aligns the incentives between what we want to do and what our customers want to do pretty well. A lot of data companies will charge for things like throughput or charge for things like the amount of data you're processing. Our view is that's a very kind of opaque pricing model and it's difficult to actually make sense of as a customer. We basically think if you're using the product, you must find value in it. It also encourages us to make sure the product works not just for the analyst or data scientist or like the developer, as you said, but for everybody else. Like people have to consume data as well who aren't on those teams like executives need to and operations folks. And we want to make sure we're building a product that's, that's good for those folks. And if we do, then again it's a monetizable thing for us because the more people who use it, the better off we are and hopefully the better off our customers are.

But when you say seats, is it like a monthly subscription?

Yeah, it's a. Yeah, exactly, it's like a monthly subscription model.

Can you get an on prem version of it?

No, we do not offer an on prem version. That's fine.

You would have to. I guess the challenge would be, is connecting to data sources that are on prem. But you say you do that as well.

Basically we offer agents that you can install that we can connect to that. There is some Mode software that actually will run on your server if you do that. So we offer. But it's a smaller package rather than you like installing the entire service.

Any client references that you can talk to?

We have a number of folks. I mean there's a bunch on our website to make sure I share the ones that I know are okay to share.

Yeah, right. I get it.

Some of the more notable ones, DoorDash, Lyft, Shopify, VMware, some of the larger customers, Twitch, the gaming company. There's a few, like media companies, like large media companies. Conde Nas, Bloomberg Reddit, folks like that.

Are you venture funded right now? Yes, yes. Series A, B, what we have raised.

Through a D. So we raised a series D in 2020.

That kind of brings back to where we started. I find it interesting that. Are you going full remote or is it kind of partially remote now or what?

We are full remote with the expectation that there will be offices for people to go to if they want to. Mode is a remote first company, I would say, but not remote only. We have been mostly remote really since the pandemic. We've had some, some interludes where we've tried to have some folks in the office. We do have an office in San Francisco that once basically like essentially we've tried to open the office and there's always been the variants that make it where it's like, okay, this probably isn't worth it. You certainly don't want to, you know, put people in a position where they, they might come in and get sick or anything like that. Our expectation would be that we would have an office in San Francisco. How will it affect not having an office? Like, is there a difference between working out of an office versus not? I'm sure you lose some things. Like, you know, I, I'm not going to go full. Like we can't possibly ever get anything done, not in an office. Like, I think the last year, almost two years has clearly proven that's not the case. I personally get stuff out of being in an office. I would like to be able to be in one. You know, it's trade off. How exactly that all shakes out. What other companies do, I don't know. We'll See, but we want to try to. Again, that's one of the reason we want to provide somewhat of a mixed option where if you want to be in an office, partly to get out of your house or partly just to be in a place with a little bit more energy, with some people around, where you can have kind of those serendipitous conversations, then then there's an option for that as well. Honestly, we'll see like the rest of the world, we'll kind of see what happens in two years when hopefully this stuff is fully behind us and we're into something normal, whatever normal looks like.

Then as they say, the new normal. Going back to pivoting back to a tech question here. You wrote an article about the missing piece in the modern data stack. What do you see as the modern data stack? And what is the missing piece? Do you remember that?

And it's not really missing anymore. The definition of the modern data stack, you could call it a bunch of different things. In its broadest sense, it's like data tools that have been sort of created in the last eight years, five to eight years, that follow some general philosophies about. They're usually cloud based, they're usually like thinking about building the stack in a more modular way rather than these sort of really big, huge like end to end tools that do data warehousing and data transformation, ingestion and visualization, like basically trying to break those pieces up. They tend to be more oriented around the idea of analytics being a code first process. So they are code oriented, have version control, things like that. So all those things I think are like, what constitutes modern data stack is basically tools that fit into that paradigm. There's no real clean lines around it. Everybody would sort of agree that like this tool is in it, this tool is maybe on the edge, it doesn't really matter, but it's basically just like the new wave of data tools. To me, the thing that was missing was we have, and this is something that like Cognos does, for instance, that we had built ways to basically ingest data. So there's like the five trannon stitches of the world that will ingest data into a warehouse. There are warehouses, there's the redshifts and snowflakes and bigqueries. There's ways to transform data within that warehouse. So like DBT became kind of the data modeling tool for, okay, we have a bunch of raw tables, how do we turn them into fact and dimension tables that are clean and well governed. And then we have like visualization Tools that sit on top or in other cases. We also have a lot of other things like, you know, data science platforms. There are operational analytics tools, there are data catalogs. There's kind of all of these different ways in which we actually surface data to the people who need it. There is a piece of there in the middle, which is the way that people often consume data is around concepts of metrics. It's like as a sales leader, I want to know our win rate by the quarter or I'm an executive and I want to see our ARR like run rate every month for the last three years. That definition of what revenue is and how we compute it often still lives in a BI tool. It probably lives in something in Cognos that is modeled in Cognos or in Tableau or in Looker or in whatever. However, a bunch of tools want that information. Like if I am doing a machine learning model or I'm doing an A B test, I probably want to know what is revenue. And I don't have access to that because it lives in the BI tool. Like I have to go back to the warehouse and the warehouse doesn't have that. And so to me, there is something similar to what the transformation layer did where it made basically data transformed in the warehouse instead of by the BI tool. And so that was pulled out and therefore accessible to every other tool where you make like metrics similarly accessible to everything else, where instead of metrics living inside of each kind of consumption tool, they can live in the warehouse in some way or on some sort of layer around it. Whereas a BI tool, we can easily access some global metrics definitions. Whereas a catalog tool, you can access the same thing as one of these augmented analytics tools, you can access the same thing. So we don't have to all recreate our own definitions of what does revenue mean in a bunch of different places.

Makes sense. Hey, anything that I didn't ask you that you wish I would have?

No, I don't think so. This has been good.

It has been good. Where can folks reach you and or mode?

So you can reach mode@mode.com mode.com was not available when we started mode, but we got very lucky and it became available a few years after that. So an added benefit of the name, you can also reach me over email. My email is just benode.com, been with one or two ends. But most of the stuff, like the stuff you referenced about things that I've written, that's where most of me yelling at the Internet these days comes from and that's@Ben.substack.com so there's a substack that's basically a weekly blog about these and related things. And also on Twitter, I'm just Ben Stancil, which you can follow along on that sort of stuff if you're interested.

Sounds terrific. We'll put that in the show. Notes. And if I could, just a couple more very light questions. Where do you think Moat will be in like three years?

I mean, the market is really big. The opportunity is really big. I think that the space that we are in is a really valuable one. I mean, our goal is we want to be on the path. We believe that the opportunity is there for Moat to be on the path to ipo. We basically say like, hey, how do we get there? So that's kind of the scale that we see the market and we scale and see the opportunity. So, you know, a lot of work to do to be able to do that. But that's the path we're going to be on.

What do you think is more important? Visualization or actually the deep analytics that go underneath. You had to pick one.

It has to be the analytics. If you do a really good job with the analytics and the visualization is basic stuff, you can still have value. If you do a bad job of the analytics and you make amazing visualizations, you don't really get anywhere. But I think that visualization is a hugely important part of that for two reasons. One, it's an important way to actually understand data. Like the deep analysis cannot be done by just looking at tables. You also have to be able to understand the trends and we understand that with visualization. The second thing is analysis is only good as people who understand it. You have to communicate it. You have to sort of tell the story about it. A bunch of text and tables isn't going to tell a story. So you have to have visualization for that. So it is possible to do it without visualization and it's not possible to do without the deep analysis, but you really need both.

What do you do for fun?

I play baseball.

Baseball it is.

So I pitch. So I can't play softball because they take away the one thing that I can actually do.

That makes sense. I see. Well, how fast do you pitch?

It's on the decline. I mean, I don't know. I like now I probably like top out. It's not fast. Probably like on a good day, the high 70s, 80 maybe.

But you got a good knuckleball or what?

No, I basically I'm Slowly just becoming an old, like junk baller basically. It's like a, like a cutter. So it's like essentially throw a lot of cutters and change ups. I'm not going to throw it by people anymore. So you got to just like have some sort of weird movement on it and hope people don't. Hope people don't pick it up.

Is there a book that you recommend most?

Actually, someone was just talking about someone recently. This is an older book, but I've been like, I've been obsessed. Elizabeth Holmes trial to me is absolutely fascinating. And the original book behind it, the John Kerry wrote Bad Blood book, which is kind of the big reporting about it, is a really good book. It's. It's like an entertaining page turner. It also is kind of people argue whether or not Theranos was like Silicon Valley proper or not. But it's an interesting sort of peek inside of what happens when companies go horribly wrong and like where ambition can take people when it's kind of unguarded and stuff like that. And it's just like it reads like a fiction story that's actually real and the whole thing is nuts. It's a very good book. It's not like something you're gonna walk away from being like, I now know how to be a better CEO or whatever, but it'll teach you what not to do as a CEO, which is don't be a criminal. But outside of that, it's definitely an interesting book.

Well, thank you for being here. I greatly appreciate it. I wish the best to you and to Mode as well. Thanks. Thank you.

I appreciate the time. Thanks for having me on. And yeah, appreciate all the other questions.

You bet. Hey guys, hit me on Al Martin. Talksdatamail.com We'll put anybody on that you feel is worthy and we'd like to hear back from you. Thank you. And for now, I will see you on the podcast. See y'.

All.

Thanks for listening to the Making Data simple podcast where we make data fun. Be sure to visit ibmbigdatahub.com podcast to access the show notes and uncover even more great episodes. Remember, the views expressed here are those of the host and its guests and do not necessarily represent the views of IBM. Until next time, over and out.

Sam.