The Data Stack Show - 233: The Power of a Triple Threat in Data: Business, Engineering, and Strategy with Solomon Kahn of Delivery Layer and Top Data People

Episode Date: March 19, 2025

Highlights from this week’s conversation include:Solomon’s Background and Journey in Data (0:38)The Importance of a Triple Threat Data Person (5:14)Sports Sponsorship Analysis at Nielsen (7:31)Challenges of Implementing AI in Business (11:09)Understanding Data Delivery Models (14:18)Innovating Data Delivery (17:38)Modern Data Sharing Framework (19:09)Account Management in Data Sharing (23:43)Data Delivery Systems and Skill Sets (26:08)Practical Steps for Monetizing Data (29:02)Building Trust Through Branding (36:51)LinkedIn Personal Branding Tips (40:54)Mastering the Basics (44:16)Professional Development in Data (48:18)Deep Technical Skills (53:18)Active and Outcome-Focused Approach (55:25)Finding Top Data People and Parting Thoughts (56:44)The Data Stack Show is a weekly podcast powered by RudderStack, the CDP for developers. Each week we’ll talk to data engineers, analysts, and data scientists about their experience around building and maintaining data infrastructure, delivering data and data products, and driving better outcomes across their businesses with data.RudderStack helps businesses make the most out of their customer data while ensuring data privacy and security. To learn more about RudderStack visit rudderstack.com.

Hi, I'm Eric Dotz. And I'm Jon Wessel. Welcome to the Data Stack Show. The Data Stack Show is a podcast where we talk about the technical, business, and human challenges involved in data work. Join our casual conversations with innovators and data professionals to learn about new data technologies and how data teams are run at top companies. We are here with Solomon Kahn from Delivery Layer and Top Data People. We're going to talk about both of those.

Solomon, welcome to the Data Stack Show. Thank you so much. All right. Give us just a high-level overview. And one thing I'd love is for you to just do quick hits on all the different things you've done, because that's gonna be so fun to dig into when we dive in. But give us the flyover.

All right, I'll do the flyover quick hit version. I have been data looted for a while, both at tiny startups where I was the data team of one, two large publicly traded companies where I managed 100 plus person team and owned the P and L of a reasonably sized data business that actually sold data as a product. I worked at startups and I've done every data job there is and excited to talk about that on the show. All right. So Solomon, we talked about a number of things before we hit record here. One of the ones I want to dig in on is this client facing reporting. I know that

one of the projects you're working on right now is trying to tackle that issue. It's complex. It's fun. There's not a lot out there that's like nailing that problem. So I'm excited about that. What else do you want to talk about? I'm excited to talk about all of it. The customer facing analytics applications, data, professional development as a data person. How do you do impactful work? I think there's a lot of things for us to talk through. Awesome. Well, let's dig in and get to it. Yeah, let's do it. All right, Solomon, you gave us the quick hit version of your background. And one thing that I think is so helpful is that you've seen the entire spectrum. One person data team at a baby fledgling startup, two teams of three digit head count teams, and probably, you know, dealing with revenue and P&Ls that are in the bees as we say not the end he had not the bees. I've not heard that

Pnls in the bees. I was slightly under the bees, but okay a lot of M's What was the Silicon Valley thing the three comma club, okay? He has that painting and it's like three commas there like what is this? He's like three comma click. Oh, okay. He has that painting and it's like three commas. They're like, what is this? He's like three comma clip. Yeah. Okay. So lots of Ms, but give us a couple minute version and maybe speak to just a couple of the experiences that were really different. Sure. So I've, I started like what I studied operations research in school, never expecting to use it in a professional capacity. Because back then there wasn't really this industry of data people. And I was working in finance.

I decided to switch to tech because I was excited about all of the cool things that people were building with technology in the mid 2000s. And I fell into data work. I was joking with you before. I think that I was already the only programmer who knew what revenue meant. So unofficially, I was the one supporting all of the business people. And then when this idea of this sort of triple

threat data person of like understood statistics, understood programming, understood the business, and there was this excitement about what people like that could do. I was in a really interesting place to say, hey, there's this new developing industry of data and data science. Why don't you let me do this and build a function of this at the company? And so I did that and grew from there. I was a company called Paperless Post. I was there for a while building up the data team. I next worked at a consulting company that built internal new products at like big Fortune 50 companies.

I did that for about a year and a half focusing on data products. I was at Nielsen for a while in a couple of different divisions, but primarily the sports business. And then I was at a unicorn startup for a bit running data before I embarked on my current adventures of delivery layer and top data people. So I've done a lot.

I've gone from that one person data team doing all of the work all the way to leading big teams. And I have a lot of thoughts on being technical as a leader and all of that stuff that I'm sure we'll get into. Yeah, no, I love it. And now you have the founder badge as well, which is super cool. So what a journey. What a journey. One thing I'd love to ask about and maybe we'll just give a preview because we're going to dig a lot more into professional development later on in the show. But you mentioned the triple threat data person, right?

So understand statistics, understand programming, and understand the business as well. Give us just an overview of, or maybe a quick summary of why is that so powerful? I have my own ideas of why that's powerful when I've seen it. But in your world, why is the triple threat just the triple threat? Yeah, it's a great question and a great way to frame the question. I I think that the biggest because all of those skill sets existed before.

Right. So why is it? Why is anything beneficial when you get it into just one person? It's because you don't deal with all of the loss of information that happens when different teams and people work together. And so when it's all in this one brain, you're able to instantaneously understand nuance and understand how to do things in a way that takes like a larger organization that's a lot more challenging.

And so what happened was that not that you couldn't do many of these things before, but because data people are often their own product managers and are often their own DevOps teams and are often their own. So having a single person be able to do this, let you do things that you weren't able to do before. And for any individual data person, like you said,

I'm sure we'll talk about this more later, but the more, the more domains you understand, if you just understand SQL versus you understand your business and your industry, you're going to be able to operate far more effectively. Sort of the more you're able to add in there because there's talked about the triple threats. There are quadruple threats, quintuple threats, people who have a lot of these really deep skill sets and you put them together and that's how you become really valuable as a data person. So I'm curious about maybe a practical application of this and the one that comes to mind valuable as a data person.

like, wow, that's super cool. It sounded like they like analyzed all of the TV footage and like to recognize brands. So I want to talk about that. And then let's talk about that, like triple quadruple threat thing. Like how did that apply in that context? Sure. So yeah, it's a great, it's a great topic. So when I was at one of the divisions at Nielsen

where I worked is the Nielsen sports division. They focus on sports sponsorship data. And so what happens in sports sponsorship is you have companies that are buying Jersey patches for the NBA or that are buying the sponsorships on the sidelines or there's so many ways that you do that. And in a big part of the value there is the media exposure that you get during

the broadcasts, because however much time people are looking at the screen, they're seeing your company. And so there are many examples of companies that have focused on this as a brand channel for brand advertising. And so what my division did at Nielsen is we recorded essentially all of the major sports globally, and we analyzed every second of the broadcast. Wow. Or what brands were on the screen. And then we also looked at the value of the media in those games. So how many people are watching and what are the CPMs for if you were to buy a commercial on that broadcast? It's very different when you're watching an NFL game in the U.S. versus you're watching a second or third tier league of football in Europe, where those audiences are much smaller.

And so we aggregate all that data. And then the teams want to look at it a certain way to support their ad sales. The brands want to look at it a certain way to support their ad sales. The brands want to look at it a certain way to support their portfolio. The leagues want to look at it a certain way. And so that's that was our business. And that is so interesting. Yeah. And in terms of how it all comes together, I'll give a great example. So one of the things that we did was and one of the things that I was responsible for was we, and this is pre-gen AI, like this is sort of like computer vision AI. Right. But we shifted and I know you have a lot of engineers here, so they'll

appreciate this. We shifted our analysis from essentially manual. So we had an operations team that was a thousand people plus drawing boxes around logos, sports content in order to actually generate this data set. It's like thousands and thousands of hours. That most of the week, all of her, all of professional sports across the world, right? Big ones. And we were only able to get so we use this algorithm. Let's look at it's like a way of looking at brands to find them. Like comparing images. So it's called SIFT. It's like it's like a way to see, OK, I want an exact match to this image, even if it's bigger or smaller, find an exact match. But sports is not good at finding exact matches because the Jersey patch

when you turn like this, even if it looks different, right? And so we bought an AI company that was very good at computer vision and we acquired the company. And then we had that team work with us to build a computer vision version of, you know. And by the way, we still had a large operations team supporting us afterwards, which is one of the things that people don't appreciate about AI is you have to the biggest challenge when you actually run a real business on AI is how you navigate the 5% of time that things are not right. Yeah, and in some cases you actually need a substantial effort in order to do it. It's still worth it, but you do need a substantial effort to manage that.

So, and for y'all, that was direct. That was actual dollars that are tied to managing that 5% in the long tail. Oh yeah. Okay. Like a customer gets a response. It's not super helpful from an AI agent. It's like, okay, I mean, that kind of stinks and happens enough, maybe they're not a customer anymore,

but you're charging people based on the results of this data, right? Not only are we charging people, but like they're charging. It's like, like we're supporting all of the- You're at the bottom, you're like a platform. Exactly, we're supporting a business. And if our numbers are not right, then that can have substantial implications for that business. And so we end there.

This is a whole other this is more of a media podcast topic on how do you how much is like how to how important it is that these numbers are right in audience measurement and stuff. But I mean, Nielsen invests a lot in that and we did invest a lot in that. But to your original question, like how did being a triple threat help with that? I cannot tell you how important it was to because like any big initiative like this, there were moments where the whole thing almost just collapsed, right? And didn't happen because nothing is ever perfect in any data system. And so being able to make a decision where you understand the business implications of we're rolling this league out with this model or we're not rolling this league out with this model, we're used, we're going to, we need this.

You know what? This is not actually working for this particular sport because we can't build a big enough training set to make it work. Or this one, yes, I know the sales team is nervous, but this one is actually good. And there are countless decisions that have real business implications

as you work on a project like this. And not just from a leadership side, from an actual building it side. When you're a data person building these models, understanding the business context of here's how sports works and here are here's the way that these leagues do sponsorships. And so here's how we need to think about what the models are looking for and where we trigger some sort of manual effort, et cetera.

Having those sort of business skills combined with statistics skills, combined with engineering skills, you literally can't do it without all three things. And you really do need that all sort of as you're doing. Even if your main job might be building the engineering systems or your main job might be building the models,

if you don't know the other sides, you can't do your job effectively. Yeah. Yeah. I was reading recently about just some people who were at companies like. Apple, Microsoft, et cetera. And it was sort of behind the scenes stories around one example was the Microsoft office for Mac suite, which is hugely problematic for a long time.

There are all these things, right? And so the person who actually sort of solved that problem, it was a force of will. They had a ton of context and they sort of saved that from being just a complete disaster in the way that you're talking about, right? But it was because they were a really unique sort of multi-threat person. So yeah, okay, I want to switch gears really quickly, Solomon, but extend this conversation a little bit into delivery layer. So you were delivering data to a lot of different companies.

I think it's kind of interesting, you were like, we're charging people, they're charging people, sort of the it's turtles all the way down type of thing. But it wasn't just your own internal analytics, right? It wasn't like you were running an analytics team that, okay, we do these reports and all that sort of stuff. You're actually packaging these as assets in a number of different ways for a number of different customers. Was that where you really started to think about delivery layer?

Yeah, I mean, I mean, it was I had done some work with customer facing data in my sort of previous consulting jobs before joining Nielsen. But overall, what one of the things that I think until you've been running a data business, you don't appreciate is that data assets are often valuable to different people in different ways. So the sports one's a great example. What the teams look for, you can take the same data, right? How valuable is any impression of a sponsor on TV? And you can take that same data set and you could see how many people have different ways

of wanting to look at that. So a brand will want to look at it from sort of a portfolio management perspective and look at are my investments in these different teams getting me various different results. The team is going to look at it from a totally different way and they need to see that data totally different. So if you just have a team's product, you've got nothing for brands except a data export, right? You need an entire way to deliver data to all of these different constituencies. And if you don't, all of the money that you're investing

in creating that data set, a lot of it just goes to waste because you're only commercializing a fraction of what you could because you're limited by the delivery of it. And I feel strongly that and frankly this this is a larger problem than just for data products. I think data products feel it most acutely because they're the ones that it's like a clear path to revenue. It's we've got this data asset that costs us a lot of money to have.

And because we can only deliver it to teams and we can't deliver it to agencies and we can't deliver it to brands and we can't deliver it to athletes, we only get a fraction of the money that this data set could provide to the market. But even broadly, the modern data stack is excellent at all of the tools in the modern data stack work together very well and effectively as long as data stays inside your company. But the minute that data needs to leave your company, it falls apart in a couple of different ways. And I felt like nobody is actually trying to innovate in the delivery of the data, right?

Nobody has ever said, I have this new idea for a data table. It's going to be better than the existing data table. And that's where my customer value is going to be. It's like, no, the data in the table, that's where the value is. It's not like the new kind of table. So that's where the idea is. People are pretty into the new kind of tables though, nowadays too, right? Yes, it's true.

It's not a perfect analogy. But from a customer, they do not care. Yeah, you're out of this. So shout out to Iceberg if you missed it. Yes. I'm talking about the visual tables, not the backend data table. So and actually, John, you've dealt with some of this as well. So this is a question for both of you. dealt with some of this as well.

in a format like how do you do that? Like what's all in walk us through? OK, how did you do that? Do you want to start with how it was done historically? Would that be fun? And then like kind of move into. Sure. John, you tell us how do we do it? How do we do it? Historic. All right.

So you log into the server. Usually a window. Usually a Windows server, so remote desktop. Yeah. SSRS. Yeah, well maybe. Or I was going even further back about you. Log into the server, you have this script. The script runs on Windows Task Scheduler

and it launches the files to multiple FTP locations and that's what you do. Yeah, if you go way back. And then obviously there's different iterations, but yeah, I wanna hear from you, what's how it's evolved and what you guys. Yeah. Yeah. If you go way back and then obviously there's different iterations. But yeah, I want to hear from you what you how it's evolved and what you guys are doing now. So I've got a framework that I like to use to sort of help take this ephemeral

concept and make it a little bit more concrete, which is the way that I see it. There's only five ways the data leaves your company. It is either in a web application that somebody logs into and sees charts and graphs in a dashboard. It's via a lookup API where somebody goes and crafts an API requests and they get data back and then they use it or store it or whatever. It's in a big file that goes to an S3 bucket or an FTP site, and there's a cron job somewhere that runs a query and just distributes it. It is in an email where someone sends you an Excel file or a PowerPoint or something

where there's just data inside that. And then there's data to data direct connections. So this would be like a snowflake data share, where you're actually just getting data directly in your database. And what I find is that the general tools in the market play in one lane. And if you are a business,

you need to play in multiple of those lanes. And that is the biggest challenge that I think people find right now in the market. That's where delivery layer is focused. Many people think, oh, you're sharing data, so you start at the bottom of the raw data. It's actually the opposite. Delivery layer starts at the top. It is full-on web application, authentication, user account management, data permissions,

entitlements, et cetera, charts and graphs, and then APIs. And then that sort of gets you both the visual access that you need as well as the programmatic access. And that's where delivery layer sits. Yep. That makes sense. Yeah. Yeah. From my experience, this is a very hard problem for companies. I guess even harder when you introduce like other protocols, Yeah, from my experience, component of the modern data stack is that direct data share piece. I think there's a lot of people that are on these modern Snowflake, BigQuery, Databricks, whatever platforms

that don't realize if you've got customers, clients, vendors on the same platform, it's pretty trivial to share datasets with them. However, there's still a fair amount that has to align for that to work out. You're both on the same system. Sometimes you both have to be in the same region on the same system. So there's still stars that have to align. But having something like Data Layer makes a ton of sense as far as being able to do that data sharing. Yeah, that's that's that. That is the challenge. That to me is where many companies that because that well, hopefully data doesn't accidentally leave your company.

Right. If it's leaving your company, there is a very important reason that it is going somewhere, whether it is product focus, where this is your product, like you have a market intelligence product or a benchmarking or a data product, or maybe it is supply chain or customer reporting portal or whatever it is. It's not an optional thing. And what I find the challenge and delivery layer does less in the sort of like database to database, like I can query your database directly. But database to database like I'll quit. I can query your database directly, but for companies, the challenge with that

is you generally can't control what data warehouse your customers use. And so it works great when everybody is on the same warehouse and then it gets challenging when you need to support different customers on different warehouses, as you said, in different clouds, et cetera. Yeah, which you just have to get to like three or four. Like maybe you get lucky, oh, we need to share data with this one person. Like, oh great, we're on the same platform.

But like the second, third or fourth, like you're going to hit one really fast. That's like, oh, we're not on the same platform. Yeah. So that's one challenge. And then the other challenge comes around and this is dependent on the data stack or the data product, but it's around permissions and account management. So one of the things that is misunderstood about this problem of sharing data externally is that most people think that the problem is like the charts and the graphs or the APIs. And that's actually not the problem. The problem is in the account management

and permission systems. And that is where so much of the complexity lives. And it is drastically underestimated any time an engineering team or a data team goes and builds one of these, that's they're shocked to find that 50% of the time is spent on the permission systems. Again, some use cases, like if you're just like sharing all your data and it's a one

off for everybody, like some use cases don't fit this, but most of the time that's where people really get stuck. Yep. No, that makes a ton of sense. I could say I can think of a lot of, in any time you're, so one of my previous lives, we were a third party agency essentially, we would share data with clients and then the clients would want to share that data with their customers.

That's a permissions nightmare. like if it has this client ID, like everybody can see this, but if it has a customer ID for that client, that's a different combination of permissions. And then, oh, by the way, all the data is really messy too. So you can't just like... And isn't there like a... There's certainly an audit component here, or I guess in the data world, you could even think about it as lineage, but really you're talking about like an audit trail almost, right?

Which is also really hard but Solomon is that tell us about your experience the data layer and sort of looking through that right because if it's if you're selling the data and there are legal questions you kind of need to be able to trace the audit trail right? Yeah yes so and this gets into one of the things that I find in this whole like data processing, like who does what is you, you end up with a sort of clear point between like data creation and data delivery, right? So for data, so, and this is a, this is sort of my own way of thinking and splitting the world into two camps, right? So, and we talked about this a little earlier where delivery, you're trying to do a good job, but you're generally not trying to reinvent the world

for delivery. Whereas when you create data, like you can have very complicated data products that you need to look at a lot of the lineage for how it got created. And then the, oh, and then the, but when you're just looking at auditing the delivery, it's very simple.

Did this customer look at this dashboard? Did this customer look at this API? So that's where, but I do think it's an important split because the skillsets of all of the people that are creating that valuable data are extremely different than the skillsets of the software engineers that you would typically need to build the systems to deliver the data. And so it's sort of just a distraction of you need to build your own delivery system whenever your business is more complicated than being able to just deliver the basics. Yep. One question I have, Solomon, you've obviously thought a lot about this and have been involved

in businesses that sell data products, but do you think that there are a lot of companies out there who are missing an opportunity to drive additional revenue by monetizing data? Yes. And by the time this airs, it will have come out already, but I'm working on a blog post about this. Definitely. It's one of those things where there are a couple of different benefits for companies that are thinking about offering data as a product.

For the people that are listening to your show who are mostly sort of data people, the one of the best benefits is that it very quickly pays for your data team. Right? Like you, you probably only need to increase your total like, cause data businesses are probably going to be a percentage of your total business is not going to, but some companies are exceptions, but they generally know their exceptions. But if you're like, can we get anything for data? You can cover your whole data team with two or three percent extra in revenue oftentimes. And so for data leaders out there who as the industry is

facing this crisis of confidence about our data teams pulling their weight, it's really nice to be able to point to a couple of million bucks that are coming in to say, well, it's sort of obvious, right? Yeah, that's step number one. Number two is that data businesses actually have a lot of great qualities about them, similar to SAS businesses in that they are typically extremely high gross margin because you've got the data. And so as long as you have an effective way to deliver it, then you are able to drive revenue at very high margins.

And on the opposite side, if customers find value in your data and start implementing your data into their systems, you typically have low turn rates. So between high recurring revenue, high gross margin, the same reasons people love SaaS businesses from a valuation perspective, they like the data as a service businesses. I got to ask this follow up then. I don't think I've ever talked to a business that didn't believe owners out there, people may be listening of like, I'm an analyst, like I think we have valuable data. Like what's like a practical step to,

we're not even talking delivery layer yet, but just to get to like, oh, this is monetizable, this is like useful. Yeah, my best advice on this, and it will work for most cases of data businesses, some are gonna be different, but like talk to your existing customers. That is the like lowest, that is the easiest thing. If you've got an idea, it's like, of data businesses, some are going to be different, but like talk to your existing customers.

That is the like lowest. That is the easiest thing. If you've got an idea, it's like we've got all this data that that is a byproduct of what we do and we're serving customers in a specific industry. If you can connect some insights that your data can uniquely provide with some way that your customers can benefit from it, then go talk to them and see what they say. And they will pretty quickly. Well, they're always going to say, oh, yes, that's great, but they won't necessarily always pay for it. So that's the

concern. So you need to actually, this is a very hard earned lesson for people that are thinking about this. So when you're sort of validating whether customers want data and insights, never just ask them, do you want it? Make them give you something in order to get it. Otherwise they're not actually serious. But if you can get that, if they're like, oh yeah, of course, I'll do a pilot with you. I'll sit down with you for a couple hours, whatever,

or I'll pay you some amount. If you can get some real confidence, then there's nothing easier and you already have an existing relationship with them. And the way you pitch it to sales teams or product teams internally is this is good for them to have these kind of conversations with customers anyways. So it's sort of you align all of the incentives for everybody and you do it. And then if it's a good idea, it should become clear. And if it's not a good idea, it should also become clear.

Yeah. Yeah. I was thinking about, do you remember, I can't remember if you were on this episode, but we had a woman named Katie Bauer on the show. And she led data at a company called Gloss Genius, which is like software for salons and spas, right? And so you can kind of think it's a business and a SaaS for business and spas, right?

It's a business and a SaaS for business for that industry, right? So it's like your CRM, your booking tool, your whatever. We were talking about data sciencey type problems, right? How do you machine learning and other things to create these really good experiences experiences and anticipate what the customer needs and all that sort of stuff. We were asking her about all these things and then we were like, okay, that's cool. What's the grand vision for the business and how you're going to use data? It was actually surprising because that's a pure SaaS business. That's B2B, we're selling SaaS, you're using that SaaS to build a relationship

build loyal relationships with their customers. That's actually the huge win for the business, which is interesting. The DNA for a SaaS business versus selling a data product is pretty different, even if you think about user experience and the sales cycle and all that sort of stuff. So speak to that a little bit, a SaaS business sort of becoming a data business. Yeah, sure.

Well, firstly, I know Katie and love talking about this example. So I think I always think about this as forget about your internal way of whether you consider this a data business or a SaaS business or anything. What's the customer getting? What value are they getting? And how is data playing a part in that? But really, it's like data is the sidekick to their success. The product is the sidekick to their success. So in the case of something like this, it's very obvious to me that if you have a big network where you can see all the data of what's working or not working with various different salons

and then give actionable advice to the people that are using your software that can help them make substantially more money as a result, you are operating in a different level than just a scheduling application. that can help them make substantially more money as a result, you are operating at a different level than just a scheduling application. Right? And so that's a perfect example of taking what might otherwise be considered a commodity business, where, oh, you're just a schedule. I don't know exactly what their product is, but if it's CRM plus SMS, some marketing, whatever it is, but because you've got this deep expertise in data moat, you can do more

than anyone else. And I think that's actually a good model to generalize across what a lot of different businesses can do with data because commercializing data can sometimes be like an add-on to the enterprise package where people get industry benchmarks. And yes, it might not be its own product line, but it's the reason that you can put in the RFP that your biggest customers will choose you over the competition.

And it's where are the customers getting value and whether data plays a part in it. But I believe that it's a mistake to be sort of too siloed in the way that you think about it. Yep, yep, I love it. Okay, speaking of business and creating value, you have a great story about how you went out on your own and started Delivery Layer. And we'll talk about top data people You have a great story about how you went out on your own

and started a delivery layer, and we'll talk about top data people in a minute, but you had this idea to start a business and you had a very clear idea of, okay, here's, I have a marketing and distribution strategy for this business. So tell us that story, because it's a great one. Yeah, so pretty much my strategy was to develop an audience on LinkedIn and to be seen as someone who knows a lot about data products, right? And the data industry in general. And

this is something that I think is, I thought this was really important. And for a reason that, I'll tell you the reason. One of the things that I did in my various different data leadership responsibilities was I put together a list of all of the SaaS tools from a security perspective that we were using throughout like an entire division. And what I found was that every single one had a brand. Every single one was a brand. And it made me appreciate so much more the value of brand for purchasing

and trust for purchasing even at the earliest stages. And especially for what delivery layer does, which is delivery layer offers a product that is mission critical to our customers. This is their product. So it's something that requires a lot of trust. And I felt like I needed a way to have people understand the depth of thought that I put into this product and that the depth of expertise and to give people a way to develop that sort of trusted brand without delivery layers, a bootstrap startup. So without VC funding, spending millions of dollars on ads and sales teams and you don't have a patch on an NBA jersey. Exactly. I don't have a patch on an NBA jersey. One day. Yes. And so so.

So I thought that LinkedIn was a good platform for data people who are felt like I because I have a number of unique experiences in the data world. And I felt like a lot of the people that had developed really big audiences on LinkedIn were of the archetype of sort of people like people working in the industry that are a little bit more junior as opposed to people who were more senior interesting or harder and lessons and not that there aren't people doing that but I feel like there there are not that many people doing it as sort of open as I felt like the LinkedIn audience wants.

It's for a lot of reasons that we can get into this. My first controversial post on LinkedIn was about how if you have a real data job that's not in sales and marketing, you probably should not be a LinkedIn influencer. I love that. I had some of the influencers messaging me that were like, I can't believe you would write that. You're so wrong. You should be careful who you are talking to in this mark. Whatever.

It's like you wouldn't want anything bad to happen to your tiny startup type of thing. The red one. Oh, yeah. But but but I just I felt like that was important. And so I started posting and I grew an audience. And yeah, that was my marketing. I have a question about sort of one of the last statements you made. Maybe this is slightly more personal, but you know, one thing is that you're posting

on LinkedIn as yourself, right? And so it is personal. And so in those situations, when you think about an influencer, it can kind of get tied into your identity, which is maybe why your spicy take on, which actually I don't even think is spicy. It's probably just like, this is actually, but you know, I can sort of hit close to home. I love how clinical you approached it, right? You're like, okay, I'm going to start a business. I need a distribution channel. Brands develop trust. I can't put a patent in NBA Jersey.

So how do I do this? Right. And it's clinical. I mean, I'm not saying there's not like personality there, but your approach is pretty clinical. How do you balance that on a personal level? Right. Cause it is you. Yeah. I guess the answer is that I just decided in advance that I would develop whatever thick skin was necessary

in order to do this thing, and that's what I've done. And I've been through. Pretty tough situations at various different points in my business leadership career, right, Like leading a sports business when COVID hit and there were no sports going on was absolutely brutal experience. And so I felt like I had developed to the point that I could navigate the sort of ups and downs

of social media, the social media, sort of people who like and hate what you say. And yeah, I mean, still I had to develop sort of a new set of skills around it. But it was- Yeah, no, it's so great because I just appreciate it in that, and I know I keep returning to your, your post that made a bunch of people angry,

but one of the things that an advisor told me once, because we were talking about a number of different things, and it's like, look, if you are really good at a craft, like astoundingly good at it, you're gonna always find work, because people will know, whether or not they know about you on social media, the people who work around you, and tell other people, they're just gonna be like, that person,

if you need this done, like this is the person, right? Because they're so good at their craft. Right. And it was great. And yeah, just love the entire thinking around that. Okay. So what are the LinkedIn tips? What are your top tips to become an influencer? Yeah. Yeah. I saw I've had a bunch of startup type friends ask me this. And by the way, just to be clear, for the people that are listening to this that are individual contributor data people or data team managers, when I say don't be an influencer on LinkedIn,

you should still develop a personal brand, but just do it in all of the traditional ways that won't make you, won't land you in a perilous situation in your job, right? You go give conference talks, participate in data Slack groups, go to meetups, meet people, you won't land you in a perilous situation in your job. You go give conference talks, participate in data Slack groups, go to meetups, meet people, all of that.

There are a lot of ways to develop a personal brand that don't involve the CMO who's about to lose their job because they're not hitting their numbers, seeing you futzing around about how good of a data person you are on LinkedIn when they're waiting for that system that's behind. So that is, or they just see you posting and think immediately you're looking for another job. So go do all of the traditional stuff. If you're in sales and marketing, LinkedIn is great. I'm in sales and marketing now as a startup founder. So I was a successful data

leader doing a lot of things and data for a long time. And I literally never posted on LinkedIn once. And maybe I posted once, but that was it. All right. For the people who are in sales and marketing and want to be on LinkedIn. And I think I told you this before, I have the worst possible advice for you because it is the advice, the advice that you know is true, but that you do not want to hear, which is it's just consistency. I grew an audience by posting every single day pretty much for two and a half years.

And many times I posted things I thought were going to be like insanely amazing and they got nothing. And sometimes I posted throwaway posts that got huge exposure. But there was no like one viral thing. And I've tried to be very specific about my posts to make them not go viral for the wrong audiences. So I actually don't post on a lot of topics that I know would grow my audience, but that aren't like super data specific

because I find that it sort of dilutes the LinkedIn brand and then LinkedIn sends your posts to your followers. And if you're posting mostly on data and your followers are like, want care about your opinions that work from home in tech, then it gets it's. So that's my strategy. It's not a great strategy. It's a great strategy in that it works. It's not a great strategy in that there's no trick.

Yeah. Yeah. And to your point, Eric, it's like, I have done these things in the data world. So I like it's like when I knew I would be able to grow an audience because I know I've done a lot of interesting things. And so that that was that that was the foundation of it. I think that step number one is have things to talk about by doing a lot of interesting things and then go. Step number two is talk about them.

Yeah, I love it. Well, let's actually, so one of the things we chatted about along those lines, John and I just loved when you told us, I just posted every day for two and a half years. I mean, obviously a bunch of thought went into it, but it was a consistency. But you said that's actually been a theme that you've noticed throughout your career and the careers of others where

just doing the basics consistently is often one of the key ingredients to success, but a lot of people don't do that. Yeah. So, yes. And this is a, so, and this kind of ties into professional development and data work, et cetera. Being an advanced data person is really about being advanced at the basics. I know that data work, there are new tech and it's always cool and excited and complicated and crazy.

And there are some areas that you do need to be sort of advanced, but you also always have to do the basics. And even when you have these advanced technical skills, you overwhelmingly, your efforts, if they fail because you messed up something in the basics on the business side, not because you messed up something with the advanced technology algorithm. And the biggest challenge for most data people

in growing their careers to a senior level is actually getting really good at the sort of business side of data work, which is mostly having very high level basics, like being advanced at the basics type of skills. I put basic in quotes here because they're not actually basic. There's a fun graphic I've seen of this.

It's like a bell curve. And then I'm like on the left-hand side, it's like Excel, Google Sheets. not actually basic. I think that's so true. I find that like, spend a ton of time talking people down out of complicated solutions and tearing things down to like, okay, what did they actually ask for? What do they actually need? What's the simplest delivery vehicle? What can we cut out to like get 80% of the value or 85% of the value? And like you said, that's the senior job

and that's the job all the way up. Like, and it just keeps like abstracting out to executive roles. I mean, that's, that's the job. Yeah. Yeah. There, you said almost word for word, a lesson I actually learned from a coach. I used to race mountain bikes a good bit. And this guy who had coached a lot of the top athletes in the world was sort of would travel around and do these like regional clinics and see if you could go do a clinic. And I remember this concept, it'll stick with me for the rest of my life, but he said, okay, in mountain bike racing, he's like, there are less than 10 skills and they're all basic. And so he said, the difference between, he said, so actually being advanced in skill is mastering

less than 10 basic skills. And he says, so the difference and he's like, okay, then let's say you master those. The difference between you and a professional is that they can combine them together, right? And do them at the same time. And at the right times. And at the right times, right? And so, but he said it is all just executing. It's mastering the basics and then getting really good at executing them in secrets or

at the same time or whatever. And I was like, man, it was mine. I was like, wow, like that is really wild. And that is essentially what you said, right? Like you do need some advanced skills, but like often the problems are because you screw up something as basic. Oh, a hundred percent. It's funny.

I have another bike analogy that I've used before for this. Also this thing, which is like there's the category of like the triathletes who are like really into the gear. And it's like, oh, there's this new like $25,000 bicycle that's much better than like the last $18,000 bicycle that I bought. And that's going to make me so much better. And then they get smoked by the person who's just like really fast. And it's like, yes, you can have very bad tools that screw you up.

So yes, have your good tools. But for the most part, if you are like a very fast swimmer and you're strong and you have great endurance, that's going to help you more than if you're out of shape, but on a twenty five thousand dollar bicycle. Yeah. Yeah, I love it. Well, let's wrap up by talking about top data people because we're talking now, we're headlong into career development, focusing on mastering the basics.

Tell us about top data people. What is it and why did you start it? Yeah. Top data people is a small professional development program that I started for senior individual contributor data people. And it is focused on the business side of data work. And it's for all types of data people, data engineers, data scientists, data analysts, data PMs.

And it's sort of like a small group. There's a curriculum and there is group calls every two weeks to talk about the topics in the curriculum, as well as situations that people face at work. And it came about because I was doing a lot of content on LinkedIn and I had grown an audience and I wanted to I felt like the LinkedIn form factor was limiting in terms of how much I could teach, right? I have a lot of years of experience, a lot of data people that I've managed. And so I wanted to take all of my less because I've seen a lot about like, oh, I see like over the course of

managing hundreds of people, you see patterns and you see people who are just like, wow, this person was amazing at this. This person was amazing at that. This person was amazing at that. This person was amazing at that. So how do I take those lessons and package them up to be able to share them more broadly? And so I started writing out what is what are these lessons? And then I realized actually, if I was on the other side, some like super long course is probably not going to be that beneficial. Course completion rates are not as high as they should be. But I was like, if I wanted to actually help someone, what would be the best way to do that in the way that also let me do it as sort of part of the marketing channel for delivery layer and sort of make it all work while from a time investment perspective and sort of find some happy medium.

That's what this is. And yeah, that's kind of how I started it. And it's been really interesting where we got by the time this airs, we'll be starting the second cohort of the program. And I've had some really great people, great conversations, and it's shocking how similar whatever your industry, whatever your job, you're all dealing with very similar type of political company, organizational, technical, very similar across the board. Indeed. We're close to the buzzer here, as we like to say. So maybe it'd be good to wrap up. You have this curriculum,

you're guiding these discussions. Give us just a couple of top things that keep coming up and maybe speak to the people out there who hopefully are actually interested in exploring top data people and joining a future cohort. But give them just a sort of a teaser on what types of things are you talking about? Sure.

Well, so I think a lot of the big ones are how to get into the right kind of situation where you are actually influencing and impacting your company and like connected with executives. And this is like, I know this, you probably think, oh, this is more of a data analyst thing, but even for data engineers, like the type of data engineers that get entrusted to build a hundred million dollar billion dollar type of systems or supporting those types of companies, you need a lot of trust and interaction and depth of understanding from the business. And so how do you develop those high level skills? That's really what this pro this is more for senior individual contributors versus juniors,

because the idea is that similar to how executives in all business divisions get professional development when they hit a certain level, because at that point you need a broad set of skills in order to operate as sort of an executive. Data people are exposed to like executive type problems at way lower in the org chart because you're working to support those types of initiatives and You need the right context and broad skills as well So everybody has their own things that they're working on obviously in the program their own focus areas But I think that the core skills are the same Another one that would be interesting, and I've got a free article on this on a

substack newsletter that I post is what I call the five foundational data skills. So this is like we talked a little bit about like, what are the like, oh, these strong basic skills, but what are the strong basic skills? Yeah, those to me are the sort of five. I got five strong basic skills, which I know we're close to the buzzer. So can run us through them. Brooks Brooks had to drop. Yeah, I had to drop.

And I can take it over. We're going long. We're doing it live. All right. So the five foundational data skills are number one, strong mental models, which is you have an accurate understanding of how things work. And that's split into three different components, which I call like the accu-mens. So it's business accu-men, how much general business accu-men do you have as well as specific industry

and business understanding of your own business. It is systems accu-men, how do the systems work in your company? You get amazing tech skills, but if you don't know how the systems in this new company work, it doesn't matter and vice versa. Anyways, the third one is organizational acumen. Do you know that whether the sales and marketing teams like each other or hate each other? Because if you want to actually get something done in your company politically,

that's important information. Yes. Okay. So that's the first foundational data skill is strong mental models. The second one is deep enough technical skills. And this is important because you need deep tech skills. But this is also the biggest mistake that most data people make, which is the sort of endless acquisition of technology skills and always pointing to technology as the limiting factor when actually technology is often not the limiting factor for you, either from a career development or from a company impact perspective.

Number three is executive level relationship building communication skills. Yeah. It's so hard. So hard. And this is why these skills are relevant for a junior data person, they are relevant for a junior data person. They're relevant for a chief data officer.

Like a chief data officer is also using all of these exact same skills every single day. There's some others around management that they use, but still these are super important. The fourth one I call active, active, opportunistic and outcome focus or active, or active, supportive, and outcome-focused approach, which is you can't just sit around. If you just sit around your toast, you need to be actively executing some vision for how the business gets better with data.

You need to be supportive. Your job is not to be the hero. As a data person, your job is to be the sidekick. The business, it's the chief marketing officer that has a number to hit, or they get fired. It's the chief revenue officer. It's the chief product officer, right? They make the decisions. Your job is to be there to support them as their sidekick. And if you take a different way of approaching the work, if you think the data is the most important thing and these people need to get on the data bus, you are in, you're going to be in bad shape. So supportive. And then outcome focused is this mix of making sure that what you're

doing is actually going to make a difference. Oftentimes it's not and it's not, but you just don't have the skill to get out of what you're doing. And opportunistic is another one, which is the things that make a difference are fewer than you would think. So when you find something that actually makes a difference, you need to make sure it doesn't get lost. So many times people come up with these great ideas that get totally lost, nobody does anything with them. So if you see something that can make a difference, you can't let it get lost.

And then lastly is sort of data, project management and stakeholder management skills. So that's how you organize the work. But that was a rant organize the work. Yeah. That was a rant, but we got it. We got it. Yeah, I love it. It's really good advice, really good advice.

And one thing I'll just, I'll repeat because I think it's so important is it's so easy to blame like the context of your situation. Right? Oh, I'm just stuck on this project that doesn't matter. And it's on you to figure out, like you can get out of that. Like there is a way out. I think that's just so important

because you can feel really defeated, right? I feel trapped in this thing, but there's a way out. So that's great. Okay, really quickly, where can people find out about Top Data People and join a cohort? Yeah, you know what? You can check out topdatapeople.com, but also just follow me on LinkedIn

and you'll see all of the links in the profile. SolomonCon, all O's, and then K-A-H-N, but I'm sure you'll put a link in the show notes or something. We will absolutely put a link in the show notes. Solomon, this has been great. I'm glad we got to go over a little bit because those five core skills, hugely helpful. I'm gonna go back and listen to that for sure. So thanks for joining us. It's been great. teams turn customer data into competitive advantage. Learn more at ruddersac.com.