Data-Driven Decision Making with Mode Co-founder & CEO Derek Steer
Maggie: Welcome to Build. This is Maggie. Today, I have Derek Steer on the show. He's the co- founder and CEO of Mode, a collaborative data platform one that, full disclosure, I have used for years. In this episode, I get Derek's take on what it actually means to be data- driven. How that's different from being metrics driven and why data on its own is not the answer. I hope you enjoy it. Derek, welcome to the show.
Derek Steer: Maggie, thanks for having me.
Maggie: Yes. Of course. I am really excited because today we're going to get into one of those hot topics that seems to get tossed around as a thing we should be doing, but that I would bet that not everyone is really doing effectively all the time. Especially, if you're like me and you work at a B2B startup, and that's data driven decision- making. So before we get into it, I just wanted to start by getting some definitions. So Derek, if you could share what is your take on what data driven decision making actually is?
Derek Steer: Okay. Maybe this is going to be my first bit of controversy, right at the beginning.
Derek Steer: Which is to say, I think most people use the term data- driven in what I'll call, not the wrong way, but in a very different way than how I think about it. So I want to separate data- driven from metrics- driven. So let's take like, a lot of go- to- market orgs have targets to which they're tied. If you're a sales manager, you expect your team to make a certain amount of calls. And from those calls, you've got your conversion ratio. So you know generally how the funnel will shake out. And you manage your team to hit certain metrics through the funnel, knowing that at the end, you'll have the right yield. That to me is metrics- driven. But a lot of sales leaders would tell you that they are data- driven, or marketing leaders or whoever it may be. Go- to- market orgs who say they're data- driven, but what that means is that they're looking at metrics, they have standards that they need to hit, and they're doing their best to do it. That's very different from using data to solve problems and then taking the results of that and we're incorporating it into the way in which you work. So my favorite example problem is how to price and package your product. This is something that requires research and a lot of that should be quantitative research. That's a data- driven way of doing it, to go and research how individual segments use your product. How engaged people are with different pieces of the product. And start to form an opinion based on data of where you should draw the lines for, say, your entry- level tier, your mid tier, your top tier, that's data- driven decision- making, or data informed or whatever you want to call it. I don't really care about names for it. But I do want to draw this delineation between watching metrics all the time and making sure that you're hitting them and using them to drive an operational cadence versus doing analysis to help with decision- making.
Maggie: Right. That's good. I was going to ask you if there was a difference between data driven and being data informed. Because as someone in the product space, I definitely feel the idea that I could look at data all day, but that doesn't necessarily mean I'm doing this right. And the other question I had is, how does qualitative data fit into this, if at all?
Derek Steer: Well, let's talk about the notion of doing this right first.
Derek Steer: For these two different types of things, doing it right means different things. So when we're talking about looking at metrics, doing it right just means having them visible and holding people accountable and making sure that you have a good cadence around that. So for those go- to- market teams or across an entire company, the notion of tracking your OKRs and making sure that you're doing the things you think you need to do in order to hit the OKRs, it's about visibility and the constant drum beat. For analysis, you're really only getting value if you use it to contradict your own opinions. That's not quite right. There are plenty of things you can do to get value from data exploration. But if you're using it as many do, just to confirm the things that you already believe, then there's not much point in doing it in the first place. And I think that's a common pitfall. Is that folks really try to bend their analysis in the direction that they want it to go. I found that the stuff that tends to be hard to argue with and pretty helpful is sometimes really broad sorts of things like opportunity sizing. So if you need to make a decision about what type of product you're going to introduce, understanding how people use your existing product today and saying... For example Search, okay? So Search is a product that is tough to understand. If you want to look at Search patterns within a web product, it's going to be a different answer as to what to do if no one ever searches or if a lot of people search one time and then never again. Or if you have a subset of people who search all the time, but most people try it once and never again. All of these point to different issues you might have with your search experience and then different things that you would do to fix them. That's I think a really good targeted way of using data to inform what you might do.
Maggie: Right. And something that's interesting is that I love this distinction between metrics and analysis, because I think it's sometimes a little bit intimidating to say, I need to be data- driven. What I'm interpreting what you're saying is that I need to be good at asking questions. And then I need to be good at understanding how to interpret the information that I'm getting. Because just looking at a metric, doesn't really tell me anything. But if I have a set of questions and then I have data, then maybe I could reason something out.
Derek Steer: Yeah. And to that end, if you were a early stage company hiring your first head of analytics or data science, the things I tell you to look for are their ability to ask the right questions and then their ability to interpret information, which is of course, a hard thing to interview for. But if you go to talks by the foremost people in this field, they'll mostly tell you the same thing, which is that a really great problem solver with simple tools is going to do a much better job than someone who's not quite so good at thinking through problems but can do all of the fanciest modeling in the world.
Maggie: Right. Again, I want to circle back because I think as a PM, a lot of the common wisdom out there is you got to talk to customers, you got to talk to customers. So how does that fit into this world of being data- driven, if at all?
Derek Steer: Well, what I'll say about the job of an analyst is not just to crunch numbers, the image of the nerd in the corner who doesn't talk to anyone is really wrong. The best people I know at this job are students of their businesses and they go talk to other people across the floor. They don't necessarily have the time to go and talk to customers directly, but getting information from the customer is super important. I'll give you an example from my time at Yammer, which is where I worked before starting Mode. At Yammer, one of the most impactful things that our product team ever came up with was really simple. So for anyone who doesn't remember or know, Yammer was a really early entrant into this social networking for enterprise space. The space that Slack and Microsoft Teams now dominate. But Yammer was more of a Facebook style threaded messaging product. And the way that it worked was anyone could just sign up with their company email address, invite other people from their company and start messaging about any kind of work stuff. That's really what it was for as opposed to Facebook, which is your personal productivity. Well, whatever. Facebook's inaudible at least at that time. So some PM and I forgot who it was, but someone had this novel idea that, if we were to let folks know who's new in their organization, someone new signs up, if we made it clear who's new, then other people are going to interact with them. They're going to have a good experience and they're going to stick around. And so that's going to improve retention. And sure enough of everything that Yammer ever introduced from a clearly measurable metrics perspective, this is one of the best features that was ever introduced. And what it was in practice really was just like a little black bar across the bottom of a new person's photo with white text that said, new. That was it. Just a clear label.
Maggie: That's amazing.
Derek Steer: Yeah. It stuck around for like a week. It was so good that it made everyone sit up and take notice. All of the product managers said," Oh, okay. Well, now that we know that there's a really clear lever here, getting interaction in your earliest experiences on the platform is going to make you more likely to retain. What are all the other ways we can think of to make that happen?" And so you started to see product managers really gravitate towards anything that was going to facilitate conversation with new people. It's also relevant to customer success, who after we sign a deal with a new customer, they're going to go to their champion and say," Hey, if you want this thing to be successful, and we know your reputation hinges on you making Yammer successful, if you want that to work, you should make sure to personally interact with new people as much as you possibly can." Because we know that that's going to help them to stick around and generate good conversation. Same story with our sales team and our marketing team. We build go- to- market stories around making this thing happen. And it results in a much better outcome for the business with everyone understanding, okay. We've taken this thing that we've measured and we've broadcast it out to everyone. Now, in terms of getting feedback from the customer, it's not just that we did this analysis and realized this thing retained better and then broadcast to everyone. Our research team went and followed up to try and understand why and what specifically was about these conversations that were helping loop people in it. And I think that's the role that the customer needs to play is, their voice is important. It's not really practical to interview every customer about everything. So start with data, understand the broad trends and then confirm your hypothesis or reject your hypothesis after talking with people individually.
Maggie: Right. Yeah. I love that. Just using that as a check against the data that you do have, and making sure that you're interpreting the data correctly with anecdotes or at least customer conversations.
Derek Steer: What I'll say too is, we use Chorus and really love it. And we use it way outside of just our go- to- market organization. So Chorus is a little thing that sits on our Zoom calls and records them and creates transcripts. And our sales managers love it because they can work with account executives to help them ask better questions, refine their pitch and so on and so forth. But we've got groups in product and engineering who are listening to these calls too, and hearing directly from the customer. And so through tools like that, it's not practical for our product team to sit in on every customer call and they don't have to now. Now our account execs can say," Something interesting came up in my call. I've got the transcript and the recording. I can ship this off to our product org. They can understand in the customer's words, what the issues are. And then we can use that as a jumping off point for analysis." So I think with technology that exists today, we now can get the customer insight at the beginning and end of the process rather than just at the end.
Maggie: Yeah. We do exactly the same thing. We use Gong, but similar workflow. And I think that's probably one of the tools that we least expected that the product team would be obsessed with, but we do tape reviews and yeah. I love having that level of access.
Derek Steer: Yeah. Anyone who's not doing listening parties is missing out.
Maggie: Yeah. A hundred percent. Another thing that I'm curious about is how teams set themselves up to do this process better. One of the things that I learned early in my career was, and full disclosure, I've used Mode at many companies and currently use it today. But one of the things I found myself doing, especially, when I was a product manager, was just doing a lot of exploratory looking around. And asking questions. And getting answers. And asking better questions. And getting answers. But I'm curious how you see teams set themselves up to be even start accessing data effectively.
Derek Steer: It's such a broad question and it depends so much on the existing DNA. What I'll say is there's a thing that a lot of folks miss, and not just companies, but individuals. It's a trap that's very easy to fall into, which is, you've got an analyst or someone who knows a little bit more about the data than you do at your company. You go to them after having looked through some kind of elementary stuff. So in the product world, you'd probably use a mixed panel or an amplitude or something. We often see Mode deployed alongside or as the successor to something like a mixed panel or Amplitude.
Maggie: I like Mode just straight up. I don't like the abstraction of the other tools. I'm much more of a fan of like, I want to get in there and I want to play around on my own. But yeah. I see your point.
Derek Steer: Yes. A PM after my own heart. There's lots of ways to get started, okay? So maybe you've got a dashboard set up in Mode, for example. But there's something that is set up to kind of give you a filtered view already, whether you set it up in Mode or whether it's in Amplitude or something else. That is just going to say like, okay. Here's the high level stuff you need to care about. And then you have a question and it's time to start doing the followups. The big trap I think people fall into is that they are prescriptive in their followup questions. And this is the number one data team value killer. Is if you have a culture where people just go and request specific like, hey, data team. Can you help me get this metric in this way? And the data team is just like, yeah. At your service. Totally. Sure. The value that they can provide and the missed opportunity we'll get from a good team there is, hey, I know you're asking for this thing. Can you tell me more about what you're trying to solve? Are you trying to go in this direction? Because I happen to know that this data might be a little bit more closer to what you want. They can anticipate your follow- up questions. They can just do a better job of understanding the problem if you give them the proper problem, as opposed to just a request for a metric. And I just see this go wrong over and over and over even internally within Mode.
Maggie: Yeah. I was going to ask you. I have two things, one, it sounds like exactly the same thing that I would say to a product manager who's treating their engineering team as ticket takers. When really what you want to do is engage them in the problem that they're solving, because of course, they know the domain and they know how to answer your questions really well. So I love that parallel. But then the other thing is you are running Mode and hopefully your team is the best at making data- driven decisions. So how do you make sure that your team is following all these principles?
Derek Steer: Well, I'm guilty of this myself, even in this role. And part of it is because I think I'm good at this. It's funny. So my wife is a design leader. And I've seen in places where she's worked, where the founder is a designer themselves, it's a different flavor of challenge. And I'm sure that working in data at Mode is the same flavor of challenge, where there are these founders who have deeply experienced in this particular job and have an idea of what it should look like. So I'm probably more prescriptive than I should be, and I pretty routinely. Fortunately, our team is really, really, really good and will just tell me, no, Derek, that's not what you want. Just let me help you out here. So that happens pretty regularly, but the real answer to how we solve it is we've got a good idea of how to hire people who are capable of recognizing when to push back and then pushing back in the appropriate way. And saying, hey, I know you asked me for something that's real basic, but let's get to the bottom of it for real.
Maggie: Kind of related to that. Another thing that I've seen, I know I'm definitely guilty of is going off and doing some exploration and trying to get a better sense of the data that I have. And then I find myself redoing the same analysis over and over again. And every team, as our company is scaling, everyone's doing their own thing. How do you think about how people can scale data- driven decision making without just everyone redoing each other's work?
Derek Steer: There are a few ways to do this. The biggest problem is one that exists in any content management system, which is understanding what's cruft and what's valuable. There are things we've done internally at Mode to organize this. And I certainly hope that your customer success manager has been working with you on helping to organize that stuff and outside of the context of a podcast about that too-
Maggie: I'll save my feature requests for later.
Derek Steer: Yeah. Yeah. I'll create a podcast. You can come on that and [inaudible 00:16: 58 ].
Maggie: Perfect. Yeah. I'm just curious about, maybe not even within the context of Mode, but how you help people scale this asking these questions and doing data- driven decision- making, because it just feels like something that's really easy for, okay. Everyone wants to use data. Everyone has to make data driven decisions. Everyone goes off and uses all these new tools that have popped up. But then how do you, maybe it's more like a knowledge management question, how do you help a team not repeat each other's work and actually get it right?
Derek Steer: Yes. So content organization is at the root of this problem in any system, because there are other products in our space that are kind of geared towards solving this problem specifically. And what they say is, okay. Well, you're going to use our product to define a menu of metrics. And those are going to be the metrics that you work with. And you face this challenge of, well, sometimes that menu is not big enough, or you need to put something new on the menu and it takes too long. There are real trade- offs to that and you get this value of like, okay. Everyone measures stuff the same way every time and that feels good, but also miss out on a ton by not going direct to the data, by having this filtering mechanism. The way that we think about it in Mode internally, we have something that we call core concepts that we've been rolling out to our customers with pretty good success. Core concept to us we map them to things like revenue or customers or engagement, things that are the way that human would think about them. We've organized them the way that if you are a thoughtful person who is employed by Mode, you can go into our Mode instance, go to this core concepts folder and see here's all of the things that are blessed by the organization where I know that the data is going to be right. And I can use these as building blocks for my own analysis.
Maggie: Got it.
Derek Steer: That's how we've done that. We have this particular problem where part of our differentiation in the market is that you can do anything. Totally flexible. Write Python, write R, just do whatever you want. And so people do a lot of ad hoc analysis in the system. And so this is something that today we try to guide people on how to organize content to do this. We have some kind of productized things like you need to deliberately put your work into a public space for other people to be able to discover it actively. They'll share stuff in your private space. But if you want something to be discoverable, you've got to make it public, which we view as saying that something is production ready or relevant to other folks. Even just that productization, when we introduced that, it used to be in the original version of Mode that all of the content anyone ever created was just in one big old bucket. And it was totally unworkable. So having those different kind of staged levels of production really helps to guide people to the right stuff.
Maggie: And do you have an opinion about who should be using Mode or doing this type of work? It seems to me like a natural fit in my life as a product manager, but I'm curious, are you seeing more functions that surprise you who are getting in there and asking questions and using the tool than maybe you predicted when you started?
Derek Steer: I'm not surprised that there's an appetite for data analysis across entire businesses. And there really, really is. The thing that surprises me is that finance are the laggards, generally. They're living in Excel land still, but beyond that, and that's a real sweeping generalization, there are exceptions. And in fact, if you were to go to a company that does micro- transactions, so if you think about gaming companies, where people are buying lots and lots of like$1 things, they are super technical on their finance team. So certainly not a universal statement. But in terms of appetite for data, we see data teams organized under almost every part of an organization. And that's what to me makes me even more optimistic about the space than I should be otherwise. Everyone knows that data is hot. What's exciting about it for me is it's hot everywhere. And everyone wants to own it and have it be part of their organization.
Maggie: Oh yeah. I think another sort of area of question I was going to ask you is if you're seeing that change the way that leaders are operating, because I think I learned SQL in my first PM role. And now that I'm moving into product leadership, I find I'm not really doing the analysis myself anymore, but I have to understand enough about how someone on my team might do them. Interrogate them. See if they're right. Ask better questions. And so I think data literacy, it feels like it's sticking around with our roles and becoming just as important in leadership. And I was curious if that is something that you're seeing across the different orgs that you work with.
Derek Steer: If you want to read a more complete answer on this, I wrote something on Quora a while back. So you go to my Quora profile and check that out on just the evolution of roles in the data world.
Derek Steer: I think that what's happening is, as you described, your average knowledge worker is becoming much more data literate and has to. As the CEO of a business, I know that it's possible to get data on basically whatever we want. There's very little that we can't know if we decide we're going to measure it. And what that means is the people who report to me have to feel that way too and the people who report to them. Because I know that it exists and it can be part of the conversation. There is an expectation that it will be, and it works its way all the way through the organization. And that exists at every company now at the highest levels, which means that everyone else has got to fill it in. And if you're not bringing data to the table when you're making some kind of decision, then your boss is going to be pissed. It just has to be, and it will be the case. And what that means for roles is, I think that data professionals becoming increasingly specialized. It could be specific technical skills, around skills like computer vision, for example, or things that are really advanced. They'll become very deep in their T- shaped kind of way. It won't be enough for them to just be very broad and shallow, because leaders are already going to be that way. They're just already on the way there. You've done some hands- on data work in the past. As a leader now, you have an expectation that folks will come with at least your level of skill, if not more in basically any role.
Maggie: Yeah. How do you prevent a team from over- relying on this type of work? I'm curious, have you seen either within your company or within other people in the industry teams over rely on data and what happens to them and how to know if you're doing that?
Derek Steer: There's certainly a concept of analysis paralysis, and that's the concern. Is how do you know? What I said, just a moment ago, that we can track anything we want. It doesn't always mean that we should. At the end of the day, having information that is actionable is ultimately what matters. The way that I always talk about it with people who are in this field, is that the hardest thing, especially folks who are early on in their careers. If you're brand new to analytics or data science, the thing I'm going to tell you is that it is an art, is knowing when to stop. How deep do you go into a problem before you realize that either you're going to not find anything there because a lot of times, it might be nine out of 10 analysis that's like, well, this didn't really yield anything of note. So how do you know when you're done with it? Do you just keep going until you find something interesting? Well, probably not. And I'm sorry to say that it is more of an art than a science. There's a certain amount of hunch and it comes from business understanding. So my sort of second piece of advice to folks is when you feel like you're stuck and at the point of needing to make the call on keep going or stop, the correct thing is usually to go talk to someone else who is a domain expert. If it's in the product world, go talk to the product managers that you partner with, understand if there's value in continuing to go deeper into understanding something. And if not, then just stop and go do something else.
Maggie: Right. I want to come back to advice, but before I come back to that, you just made me think of something. What's the best example of data- driven decision that you've seen recently? Do you have one you can share that was surprising or interesting.
Derek Steer: So I've talked a bit about our own move to a freemium model. It's its own whole talk. If you look up my talk at Web Summit in 2019, that's where I get into it in more depth. I'll give you the quick one. I don't know if I can actually share publicly the customer's name, I'll keep it anonymous.
Maggie: Good. Good.
Derek Steer: But we had a customer who did some international expansion and found that their product was less sticky with international audiences. Now there's lots of reasons why that might be. And the way that they approached it is engage the data science team, develop a bunch of hypothesis and then go and analyze behavior internationally and see what the deal was. And they ended up having to look into some data that they hadn't previously aggregated in their data warehouse or to bring new data into the picture, but one of the theories was, our servers are in the United States. People are having a really, really bad time in terms of performance and that's what's going to impact their experience. And sure enough, they were able to tear out performance degradation into a few tiers and found that the worst year of performance degradation, the less likely you were to come back. So it really was tied to just the speed of the product. And then the clear decision was, now that they had these metrics, they were able to deploy servers internationally. It's like all in AWS. So it's like, okay. We're going to spin up in a new region. We're going to direct traffic locally. And then we're going to measure the exact same metrics and see what happens and sure enough, huge improvements. So I think the classic investigative find a good answer. It's an example I like, because there's a nice happy ending.
Maggie: Yeah. Yeah. I've talked about this on the podcast in the past, maybe a year or so ago. But I had one of my first job as a PM where we did this whole analysis and we had this whole theory about the new redesign of a page we were looking for. And we shipped this thing and it worked really well. And then I went and ended up talking to a customer just to confirm the hypothesis that we had, because it worked, but I had some unresolved questions about it. And then the person I talked to was like," Well, you made the button bigger and you made it orange." So I clicked on it. And all of the data that I had looked at was just meaningless. And really all I did was just made it easier for them to see the thing I was working on. And that was a really good lesson that I had early in my career, that sort of a counter example to you, or does it even if you do your analysis right, it may not always be the case.
Derek Steer: I cannot overemphasize that an understanding of the customer and the business is critically important to having a good result. And that data on its own is not the answer. To this end, a lot of the companies in our space or companies that have tried to enter our space, particularly with value props like version your data or data sharing, I don't think data is that valuable on its own. Obviously it's valuable in the hands of someone who knows how to ask the right questions, but sharing data is not that helpful. You want to share analysis, at least that's my take. We very rarely talk, it's also like analysis is not really that good of a word. Analysis driven decision making is not.
Maggie: Right. But I get the point you're making that it's more than just that there are numbers. It's that there is a purpose and you are using data to inform decisions in a way that's like meaningful and structured.
Derek Steer: Yeah. But I think even referring to it as data- driven decision- making. The fact that we use the word data whenever we're talking about analysis, it's misleading for most folks. And I think probably pulls people away or it directs them farther away from the value that they could be getting.
Maggie: Yeah. I was actually going to ask you, when you started Mode, if part of their plan was that you wanted to open up access and have all these different types of functions and people getting their hands dirty in data, or if that was something that sort of came as a surprise or organically with the product? Because that's always, as someone who maybe didn't think of myself as someone who might be using a tool like Mode, it certainly became really useful to me in my career. And I don't know if I would have known that when I started.
Derek Steer: The value in Mode has always been about sharing and collaboration. And we've known that from very early on. We did some experiments around a public open source analysis marketplace, all like GitHub, you'd think back to 2013 when we started the company, it was hot and appealing and seemed like a good way to help bring the product to market and ultimately failed for a variety of reasons. But in terms of sharing with folks who are not dedicated analysts or data scientists, before we even launched a product, I wrote this thing called SQL School, which still lives on Mode's website, mode. com/ sql- tutorial. It's now one of the most viewed SQL tutorials on the internet. It gets a few hundred thousand uniques every month. What's special about it is that it's made to level people up from where they are today. So at the time, if you were to go look at SQL resources, you'd get SQL for Dummies book or something like that, what you'd get is a bunch of lessons on how to set up a database, how to get data into it. And it was aimed at software developers. It really was, here's how you set up your SQL backend to your software application, which is now what analysts do. If you're taking a job in this space, you're going to show up to a company and that's already going to exist, so you just need to be great at writing a select statement. You need to be good at using SQL to do analysis and you can skip the database creation steps. Also, you probably don't have application programming experience already. So you're coming from the different angle of probably Excel in most cases, right? inaudible Tableau or something like that, but something that is just a simpler version of SQL in the data analysis world. And so this tutorial that I wrote is really focused on taking folks with that Excel knowledge, teaching them the difference between Excel on a database and then giving them skills to get value from data in a database. And you can see throughout our product that this notion of leveling people up exists from stage to stage. So I think we've done a really good job of when an analyst shares something with you, you're able to either visually explore it, save that off as derivative work and now you've done something that maybe you couldn't do before. You can, with one click, see the SQL that created it and then clone and create your own SQL statement and play with it from there. Or you can borrow work from experts and modify it, which is a great way for people to get started. I think any software work I've ever done, I've just copied and pasted from stack overflow. So in some ways, Mode is like your internal stack overflow for how to work with your database. And then even with Python and R, one of the biggest gating factors to a SQL analyst learning Python or R is simply how annoying is to set it up. Particularly Python is awful to get running on your machine unless you are sitting next to someone who has done it before, who can hold hand through the process. And with Mode it's totally different. You don't have to worry about getting Python running on a machine, you just click notebook. We dump the results of your SQL query into a notebook. And you can write one line of Python to create a plot or to... Honestly, for me, the most valuable thing there is summary statistics. So calculating a median in SQL is really awful.
Maggie: Yes. Can confirm.
Derek Steer: Yes. It's like an interview. It's an interview question is like, hey, can you [inaudible 00:00:33:07].
Maggie: An interview that I would 100% fail. So great.
Derek Steer: I'd probably fail it now too, because what I would do instead is I'd go to Python or R where it's one line of code. Not only is it trivially easy to get a median, but when you do it, it gives you, for free, every other descriptive, like summary statistic about the dataset. You get mins and maxes, you get the twenty- fifth, 70th percentile, you get all of the stuff that you would want as just like one big dump of information.
Maggie: Right. So this was all part of the plan, was just access and democratization of the data and just making it easy? I don't think I realized that. Obviously that's what I was getting out of the product, but I was just curious if that was part of the founding story given how important that's become at least the way product teams operate, for sure.
Derek Steer: It was all intuitively baked in. I don't know that we had verbalized it in exactly that way. And this is one of the things that we've learned because Josh, Ben and I, the founding trio of Mode all come from this space.
Derek Steer: And it's a big advantage in our ability to serve our core audience, that we have lived their lives and understand their jobs. But what we didn't really know is how to break that down for other folks who are not analysts and data scientists. And that is something that we have learned ex- post. I think we knew intuitively that baking a SQL tutorial and doing it in this style was going to be valuable. And that the way in which we implemented Python and R in the product was the right way for our audience in leveling them up from SQL to Python and R. Those were conscious decisions. We just didn't articulate it to the market in exactly that way, but have listened to our customers and the people who are learning within our product and now are better at talking about it.
Maggie: Right. I have so many questions, but I think it's been really interesting to hear sort of.... It's interesting in talking about how to make data- driven decisions. Is we sort of like are inherently talking about your product and the journey of your company. It's just really interesting. But in the interest of time, my last question for you is, what advice do you have for people who want to be better in this field? Who want to be more data- driven? Who, maybe for whatever reason, just are stuck or intimidated, what would you tell them?
Derek Steer: Start with Mode SQL School. I got to work in my shameless plug. Well, if you work at a company with an analytics or data science team, that's the place to start. Go to those folks and ask them, because what they'll be able to is direct you to things that are specific to the job that you're working day to day. And I've generally found that people learn the best when they have something practical to which to apply their learnings. So Patreon is a customer of ours, and they actually took our SQL School online and adapted it using their own data, and have been putting people eternally through it. And they did it first with people who raised their hands and wanted to do it. So there are people in their organization who said," I want to learn more about this." And then they said," Okay, fine. We're going to put together something that's going to culminate in a project. We're going to give you some kind of analysis to do." Yeah. And sure enough, they were able to run these classes and use the materials. And by the way, if there's anyone out there listening who wants to use SQL School, just hit me up on Twitter or shoot me an email, derek @ modaanalytics. com. I'd be more than happy to just give you all of the SQL School lessons so that you can go do this yourself. I want nothing more than for more people to learn data analysis skills. And I think that the organizations who are doing this really well are running programs like that internally.
Maggie: Yeah. That's how I learned. We had, I don't remember what the tool was, I think this was pre- Mode maybe even. But we had one PM who would put together a sheet and she had written out, here's all the basics of all the different things. And then she had set up a little project within our own data at the company. And we went through little classes to learn how to actually apply the stuff. And I think that has served me so well in my career. And so I love that that's open and available, because I think just that access is huge for people who want to move up
Derek Steer: Totally. And look, not everyone's going to have the bandwidth to create a course and teach within their company, but you can get little tips and tricks one- on- one. You can use Mode's SQL tutorials to learn the basics. And then you got to go to the people at your company who know the data the best and ask them for problems for you to go try ways to practically apply it.
Maggie: I love it. Well, Derek, thank you so much for coming on the show. I appreciate you taking some time to chat with me all about everything at data- driven.
Derek Steer: It was my pleasure. This was really fun.
Everyone knows they should "be data driven" but it turns out that just looking at metrics isn't enough. In this episode, Maggie talks with Derek Steer, Co-founder and CEO of Mode, about what it actually means to be data driven. It turns out that the real work is actually all about asking the right questions and sharing analyses, not just sharing dashboards.
Check out the SQL tutorial here: https://mode.com/sql-tutorial/
Like this episode? Leave a ⭐️⭐️⭐️⭐️⭐️⭐️ review, hit the subscribe button, and share the pod with your friends! You can also connect with Maggie on Twitter @maggiecrowley