Episode 164:

How The GTM and Data Teams at Snowflake Work Together with Travis Henry and Hillary Carpio

November 15, 2023

This week on The Data Stack Show, Eric and Kostas chat with Travis Henry and Hillary Carpio of Snowflake. Travis works as the Director of SDR Operations while Hillary is Head of ABM. During the episode, Travis and Hillary discuss the importance of data in their marketing and sales strategies. They delve into the types of data they use, the challenges of data overload and the importance of a strong partnership between marketing and data teams. They also share their experience of writing a book about their unique approach at Snowflake. The conversation further explores the role of sales and marketing in targeting different personas within an organization, and more.


Highlights from this week’s conversation include:

  • The Unique Perspective of Practitioners (2:10)
  • Account-based Marketing (6:30)
  • Sales Development Representatives (SDR) (8:05)
  • Descriptive, People, and Engagement Data (11:38)
  • Data Overload and Actionable Data (14:20)
  • Working with Data Teams and Internal Data (17:52)
  • The relationship between business and data teams (22:27)
  • The importance of collaboration between marketing and data teams (24:17)
  • Travis and Hillary writing a book (25:33)
  • The taxonomy of personas (34:23)
  • Bucketing and grouping people in data systems (35:37)
  • Account-based marketing and sales alignment (39:00)
  • The data-driven approach and reliance on technology (44:25)
  • Managing complexity in data and account-based marketing (45:35)
  • Adapting to change and evolving data artifacts (51:58)
  • The importance of understanding the business (54:58)
  • Collaboration between data and go-to-market teams (55:56)


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.


Eric Dodds 00:05
Welcome to The Data Stack Show. Each week we explore the world of data by talking to the people shaping its future. You’ll learn about new data technology and trends and how data teams and processes are run at top companies. The Data Stack Show is brought to you by RudderStack, the CDP for developers. You can learn more at RudderStack.com. Welcome back to The Data Stack Show. Kostas, today we have a treat. And that’s what we’re going to talk about with some very data driven people who are not data practitioners or data professionals in their day job. So we’re going to talk with Hillary and Travis, who work at Snowflake on the go to market side. Hillary runs their Account Based Marketing function, which I can’t wait to dig into. And then Travis runs their sales development representative function. And after we had Brendan on the show, who talked through who talked with us about sort of being a data consumer, we thought, man, we need to get more people on the show. And Brendan introduced us to Travis and Hillary, and their amazing, amazing relationship with their data team. Pretty amazing how they developed that relationship. So definitely something for anyone, either on the go to market side or the data side. Definitely, a must listen. And I think the things that I’m most interested to dig into are to hear about their views on data. So you know, if you’re going to run an Account Based Marketing Campaign, or you have a sales development representative, you know, trying to understand all these data points about an account, you know, on the data side, we send data to those teams often, and then they consume it, but we don’t really get to see their consumption of that data. And so I think, being able to sort of go over the wall and get more insight into the specific practical things there would be really great. So that’s what I want to dig into. How about you?

Kostas Pardalis 02:10
Yeah, 100% I think we have a very unique opportunity here to hear from practitioners how actually, data drives value for them. And do that in an organization that does it at scale. And not only that, but we’re talking about like, an organization that sort of like a vendor, like facial Lucien that is used for that kind of problems, right? So there’s, I’m like, I think very unique perspectives here, like both in terms of the scale, like how innovative like the they are, like in terms of their marketing or like sales, but also, how they in a way they use what they evangelize out there to sell the product, but they do like internally, right? So yes, I think like a very, like unique perspective, and very unique. Like, I would say that like a very unique opportunity to learn things like, if we had someone who was only a vendor or someone who was only a user, I wouldn’t be able to do that. So yeah, I think it’s going to be a very interesting conversation. And we’ll try to do a couple of different things with them. Like, obviously, like, tried to learn about how the data is used, and why do like a little bit of data modeling, maybe, like, see how you represent all these information that marketing and sales need for their, for the day to day job and learn a lot about like the people actually, which is also like important, because it’s not just the people will be talking about waste day. It’s also like, always the hours or like, the marketing people have to execute, right? And they like to use information or like to do their job, like effectively, and see how something like this can happen, like at scale incorrectly. So yeah, I think it’s gonna be like a really interesting conversation.

Eric Dodds 04:15
Let’s do it. Every data engineering analyst’s dream is live data modeling with a business user. Let’s dig in and do it.

Kostas Pardalis 04:25
Yeah, let’s do it would be fun.

Eric Dodds 04:29
Hillary Travis, welcome to The Data Stack Show. We are very excited to chat with you.

Hillary Carpio 04:35
Great to be here.

Travis Henry 04:37
Thanks for having us.

Eric Dodds 04:38
All right. Well, we both of you are at Snowflake. And we’ve talked tons about Snowflake throughout the life of the show. But this is really exciting because you both work on the go to market side of the house of Snowflake. And we’re going to talk about the ways that you use data. So why don’t we start out? Can each of you just give a brief background and introduction and tell us what you do at Snowflake Hillary. Let’s start with you.

Hillary Carpio 05:06
Yeah, so my name is Hillary Carpio. I lead the Account Based Marketing function at Snowflake in addition to customer growth marketing, starting in 2019, and have helped grow the team from five individuals to 20 in North America and 30 globally. So really been part of the scale and have enjoyed the wild ride. This Snowflake has been acquired by a private company through an IPO. So excited to continue talking about how we use data, and it’s everything to our business as you could imagine.

Travis Henry 05:34
Yeah, and Travis Henry here, I joined Snowflake just after Hillary right at the tail end of it being a private company. So already a very big success story. But my role is sales operations and enablement just for our sales development team. So a bit of a unique role, but one that’s catching on, especially with really high growth b2b software companies to enhance and elevate the role of sales development. So the journey here, it’s been about 70, some odd STRS when I joined over three years ago to over 250 STRS today, so very much a study in scale and using data to go on that journey.

Eric Dodds 06:12
Awesome. Well, let’s start. I want to cover just a couple of definitions that each of you mentioned. So pillar, you mentioned Account Based Marketing. And for those listeners who may not be familiar with exactly what that means. Could you just give us a high level overview of Account Based Marketing? Sure,

Hillary Carpio 06:30
yeah, traditional marketing goes out to a very broad audience kind of like fishing with a net and catching you know, those fish that swim into the net and going down from there and qualifying and doing outbound etc. Account Based Marketing is more like spearfishing. So my team is responsible for identifying with the sales team, which accounts are most right to the laughter at any given time. And they’re doing very customized and personalized marketing campaigns to those accounts, in order to get them in the door on the marketing side and then coordinate with Travis’s team to get the SDR on the sales development side going on the outbound piece so that we can bring in the companies that have the highest potential for Spread the most fit, and you have most likely the fastest deal cycle based off of timing.

Eric Dodds 07:15
Yep, so an example of that would be, instead of sort of just launching a campaign, maybe around a snowpark, and sort of capturing general interest in ML stuff, you may go after a specific, you know, FinTech company. And like the ML team at that company with personalized type marketing. Correct. And

Hillary Carpio 07:37
We do have a broader demand generation team that does the broad, more broad based marketing and addresses are workloads and we’re able to do what we do because we have their support in their content generation and their broader coverage. So we’re a complement to that broader marketing function.

Eric Dodds 07:51
Very helpful. All right, and then Travis, define SDR for us. And so define it. And then can you just give us a sense of what you know, what does an SDR do every day as part of their job?

Travis Henry 08:05
Absolutely. So SDR stands for sales development representative. And you can think about it like a specialized role that’s focused on what we call the top of the funnel, or finding new business opportunities for Snowflake. So contrasting it to the b2c world that some listeners may be familiar with, or we all know, you know, Amazon, I get that great, cool pair of shoes coming in my Instagram feed, I do a couple of clicks, and I’m checking out of Amazon and I get a box to my door a few days later, it’s very low consideration, purchase, not a lot of complexity, probably don’t need any humans involved in that. When you start to think about selling a product like Snowflake, or really, any b2b sales motion that’s expensive, highly considered, requires a lot of people in the organization to say yes, and for no one to say no. It takes a long time to do, you start to get into a place where it makes a lot of sense to specialize the sales motion. So if you think about the campaigns that Hillary’s teams run from an account based perspective, you know, we’ll start to gather interests, we’ll start to see that there’s maybe some traction in this account, some awareness in this company, for a solution like Snowflake. The SDR role is really purpose built to monitor those signals, to understand them, to put a story together and to engage buyers in that account, to ultimately compel what we call the hardest part of closing any deal. And that’s finding the deal in the first place. So that’s really the SDR team comes into play. And what that allows you to do is, not only does Hillary get more conversion and return on investment from her campaigns, our closing sales executives, you know, folks who were 1020 30 years in the data industry, they’re focused on working deals and working deals to close and presenting and negotiating. We’re really that in between kind of force multiplier function.

Eric Dodds 10:00
Very interesting. Okay. So it sounds like each of your functions. You know, there’s in some ways, it sounds like there’s a collaboration. But when we think about a campaign, you know, sort of like pushing, in some ways pushing information out to this account, right? And the STRS are sort of digging into these accounts. In each case, though, each of you seem like you need a lot of data on accounts, right? And when we say account, I’m kind of referring to that as a business or a company. Is that accurate? Yep. So when you think about all this data that you need about an account? I mean, what data is that? Travis, you mentioned, there are multiple people involved in this decision? You know, so maybe you need information on different roles that are associated with this business? But can you just paint us a picture of, you know, in your ideal world, what’s the data set that you can use to do your job? Well,

Travis Henry 11:04
absolutely. And I’ll start maybe with the broad buckets, and then maybe pass it to Hillary for some examples of what we used in the past. But it’s pretty cool to go to market in 2023, because it’s a much more data driven motion. And data is that much more important of an asset for our teams. And to give you some general categories here, first, if you just think about accounts, are the businesses, the entities out in the world? Do you want to go sell to having data about those entities is super important for your business, think all the way back to pitch decK to venture capitalists for your great startup idea, right? You’re trying to put together data about, you know, there’s this many companies in the world with revenue above this size, and these industries that we think we can go sell to and you can kind of size a market using data. So there’s kind of in my mind, that descriptive elements, so firmographic data, what are the companies? What do they look like? How many employees? Do they have those basic elements of data? Then there’s the people data aspects. So who are those people in the businesses? And how big is that data science team? Or that data engineering team? And what does that relationship map sort of look like? And what do those buying groups look like? And I think the third broad bucket is what I would call engagement data. This is definitely the noisiest data set, the one with the most rapid change over time. But examples of that would be, you know, how many individuals from a company are registering for Snowflake summit during the summer in Las Vegas, right? That’s a level of engagement that we can measure from an account. You can think all the way into very technical details like product telemetry, if you have a free trial, people are getting in your product, they’re taking action in the product, monitoring that kind of engagement.

Eric Dodds 13:02
Hillary, anything to add to that?

Hillary Carpio 13:05
Yeah, so on the campaign side, our job is to take all of those different pieces of data and make them make sense in a single signal. So from the product telemetry side, you know, that’s a goal of ours from the third party data side. So what are people doing related to our business outside of our property, first party side on our property, we work with our data science team and our marketing intelligence team, specifically, in order to bring those into that score that helps the sales team understand who to go after and helps us help them as well dissect what their accounts are interested in.

Eric Dodds 13:40
Okay, so this is the event just talking about those sorts of three buckets. And we covered a lot there. So different sources of data, internal and external. We talked about marketing intelligence, we talked about data science, I mean, there’s you’re touching a lot of different data sources, and even different data teams. It sounds like Hillary, what is it like for you as a data consumer? How do you get this data? Do you define what you need? And then sort of have all these meetings with, you know, these sorts of data producers? What does that experience like for you?

Hillary Carpio 14:20
Yeah, I mean, we live in a world right now in marketing, where we’re in data overload from a commercial standpoint, so I can open my email any given day, and there’s 10 different emails, 20 different emails from providers with why their intent data is different, why their contact data is different, etc. So there is no shortage of input in terms of what data is out there. The challenge is in the accuracy of the data for your specific business, because one data provider might be great for a different type of business, a different set of personas, but not for us. And so, my time is spent distilling, what is the data source, what is the methodology and the forward looking plan with the data from the vendor? And how does that relate to what we’re trying to do? Is it relevant? And then I’m constantly asking myself the question and Travis as well as is it actionable data. So we have plenty of data sources available to us that we choose not to use, because there’s no action to be taken from it. And it puts sales and marketing both and like an analysis paralysis, just having too much data out there as a list of statistics as opposed to a meaningful insight and what to do with it. So that’s kind of where I am, where I’m spending my time and my discretion. And then we’re very fortunate to work with an excellent marketing intelligence team and sales operations and sales intelligence team that can help really guide us, consult us and tell us what’s possible. So I do the dreaming and the ideating. And they keep us in reality, and then also do a ton of innovation to make our data dreams come true.

Eric Dodds 15:49
On the third party side, and this is a really specific question, but really curious. So when you procure data from a vendor, which is common, right, so there’s tons of data vendors out there, when you procure data from a vendor, do you have? Do you work with your data team on sort of what that looks like in terms of ingestion? Or like, where do you mean, some companies just send a giant CSV or, you know, FTP it, you know, but increasingly, we see companies collecting their data into a single sort of repository and Snowflake. For you, as someone who’s procuring that data, how do you work with a data team, in terms of actually receiving it and sort of, you know, operationalizing it from a data perspective upstream of you?

Hillary Carpio 16:36
Yeah, so all kinds of stuff, the business case of what data we need, and why and what we’re trying to do with it. And then the intelligence teams will help us understand how they map together and how to ingest them. Our preference is always going to be to consume data through our marketplace and do data sharing using Snowflake so that there’s no FTP or anything along those lines. So we’ll work toward that if it’s not available, or if it is available, that would be our preference to use our own product.

Eric Dodds 17:04
Yep. Yeah, it makes total sense. The marketplace is really sweet. For sure. Travis, can you speak a little bit to the internal data, so let’s just use product telemetry data as an example. So that can actually be quite complex from a data standpoint, to get to aggregate, you know, we do some of this at RudderStack, of course, on a much smaller scale than, than Snowflake, but, you know, I just know from experience those data teams sort of collecting that data that that’s a non trivial exercise. But you as a consumer, what does that look like? I mean, what data points do you want? You know, when you’re trying to look at an account, and you want that product usage data?

Travis Henry 17:52
It’s a great question. I think it relates really well with Hillary’s point around data overload. So I think the first thing that always comes to mind, from a consumer standpoint, internally is like too much, you get too much data pretty quickly in terms of what’s useful. And the challenge is actually for our data teams to distill down the signals that matter, or specifically the moments that matter, think about in a customer’s journey, using our product or going through the trial. So it’s more an exercise in reduction and focus than it is giving me as much data as possible about everything everyone at this company is doing with the product. And for us, you know, that is kind of an interesting, open ended conversation that we have with our data teams, and with marketing, which is like, if you are always trying to put yourself in the customer’s shoes, right, if I’m using the product, and maybe I’ve signed up for it. But I haven’t started to use any credits inside of the trial. That’s probably a moment that matters to us in the sense that I kind of abandoned my journey there. And for some reason, I continue to use these resources. So maybe, is that because there’s a knowledge gap? And now we can have SDRs provide very targeted, you know, helpful information or use cases or some queries to get started with? Or is it because that individuals, you know, maybe no longer interested in going with a competitor? So it’s very much around, you know, feature adoption, as well as I would say credit consumption specifically in that product example. But I think the even more interesting part of how we work with data teams is that dialectic of educating the data teams on the business and giving them the definitions around. How do our teams operate? Because you think about internal data, you know, outside of the product example. Definitely SDR teams, Hillary’s team, pretty much all go to market functions now. Are you using a crazy tech stack to do their work, whether that’s, you know, end user facing, or it’s behind the scenes kind of capturing the exhaustive activity, that internal database, right, that internal set of data of what users are doing, is only becoming more valuable and robust. So just to give you a concrete simple example, most companies will track statuses on individuals, at companies that we’re interacting with. So, hey, is anyone communicating with this individual? Are we actively trying to reach out to this person? Have they responded to us or have we disqualified them is not a good fit, you know, things as simple as that. You start to multiply, hey, we have seven statuses for all the people we’re engaging with. And we have 2 million people in our database, and we have segmented by account. If you can educate your data teams on what those statuses really mean, and how they fit into the process, you can start to really build some interesting value added business relevant models and insights from a data team side. And I think that the opportunity for data teams is to move from purely order takers, for lack of a better phrase, into strategic partners who understand the business and can recommend, Hey, have you thought about a contact propensity score to predict who’s most likely to take a meeting? Because it seems like that’s your choke point? In the customer journey, things like that? Hmm.

Eric Dodds 21:33
Can you talk about how that relationship has evolved? Because it sounds like it’s running really well. And I think for a lot of our listeners, you know, unfortunately, in many companies, it’s sort of a, you know, here’s a JIRA ticket, I want, you know, a list of active users, you know, or, you know, people who have, you know, started using, you know, X feature in the last however many days, right, and you sort of get into this back and forth hell of like, I’m asking you for something, you know, and then it’s like, what is an active user mean, right? I mean, what kind of user isn’t an admin? Is it a, you know, I mean, that’s like, they’re different definitions. How did you form that relationship with the data team? And what does that collaboration look like?

Travis Henry 22:21
Yeah, I mean, hats off to our data team. So starts with really good talent. But I think if I was to characterize the two sides of that relationship, on the business side, for me, and Hillary, I think what’s made it work is not being afraid of data analysts in general, and actually getting really curious about data. Like, tell me more about how that works in our tables and Snowflake. And, you know, how would we shoot this thing over time and measure change and really be curious and open to it, data conversation, rather than shying away from it, which I think a lot of go to market teams do. On the flip side of that relationship, I think our data teams have done a great job of not just being curious, but pushing back and kind of calling bullshit. If I can say that on the show. You can better come from our team, right? Because we have fallen into that trap where we think we know what we need. And we think we know how it should work. And it’s just a list of requirements we handed over. And instead of going in building our data teams, like, let’s get a meeting on the books, let’s talk about what this means. Hey, if you don’t, you know, normalize these different populations of contacts, and you’re weighing everyone against each other, you know, Asia Pacific is never going to get a high score, and everyone in North America is going to get all the good scores, and they start to push back and educate the business. Some of the pitfalls there and they’re not afraid to recommend alternatives. And I think that it is a very human kind of work through it, jumping on calls, giving each other feedback and commentary. It’s yeah, it’s not that advanced or technical, it’s actually very human in terms of how that relationship has evolved over time.

Eric Dodds 24:13
Love it, Hillary, anything to add from your end,

Hillary Carpio 24:17
I’ll just reiterate what Travis said is it has to be a partnership, right, you can have a marketer who has all of these ideas can really make a difference on the go to market side and a marketing intelligence or data team that isn’t on board. And that’s not gonna go and you’ve kind of had the opposite with a data team that is really on board and a marketing team that does not button and that’s not going to go anywhere. So you really need innovative minds or two innovative teams to come together and share the same vision and be willing to push the boundaries and be willing to try things. I think we’re really fortunate that our counterparts like I said, Receive our ideas and are willing to go figure out how to help make them happen. And so it has to be both sides.

Eric Dodds 24:57
Love it. Well I know Costas has a ton of questions. But I have another question, which is a little bit of a detour. But both of you wrote a book, which congratulations, that’s a huge accomplishment. What was it like to write a book? And I mean, really, this is, you know, a lot of what we’re talking about is covered in the book, right on how you built these relationships inside of the company to accomplish some pretty great outcomes. But what was it like to write a book who, who had the idea to start with,

Hillary Carpio 25:33
so we were presented with the option or idea to write a book by our CMO. So after several QBR, Travis and I presented where we were heading, what we’re building, in conjunction with, of course, the rest of the team, she was like, Hey, this is really unique. I think the market needs to know about it. And we jumped on the opportunity to share our ideas with the world.

Eric Dodds 25:53
So cool. Did you ever think that you would be an author?

Hillary Carpio 25:57
It’s always been on my bucket list. But I never had gotten to the point of thinking about what the topic might be, or, yeah, I don’t have that background in journalism, or most of the degree in journalism. And I’ve always loved to write, so I wasn’t sure if it was going to be like on the fiction side, or the business side or what? So appreciate her for keeping that off.

Eric Dodds 26:17
Yeah, I am about to tell you,

Travis Henry 26:20
I absolutely did not anticipate myself ever being a published author. But here I am, of course, getting to share that burden, that journey, and accomplishment with Hillary, which was a really cool opportunity. And just to answer the question directly, for anyone thinking about it, you just gotta put in, it’s like working out, you know, you got to set out your gym shorts, you got to go do your reps every morning, you’ve got to take it little by little over a long period of time. So I think the journey was probably eight months of just heads down, writing, rewriting, analyzing your ideas. But the cool part about it is, you have a lot of good stuff to say, or you think you do, and that’s why you’re writing the book. But it also is a challenging exercise, because not only do you have to crystallize the thoughts in your head, or what you’ve done into words on a page, you also need to make those clear and valuable and understood by an audience as well, which is a big challenge. And I think, for me, that was the biggest surprise, which was just, it helped me think about concepts and the work that I did, in a much more clear and structured way. So it’s kind of a forcing function to go through that. So, you know, if you don’t have a publishing deal, or the funds to go do that, you know, maybe write a little book or get your thoughts out there and publish stuff, because I think, just going through the writing, exercise, and discipline is a really good exercise no matter what field you’re in. Totally.

Eric Dodds 27:55
Did y’all have some sort of editor to give feedback?

Travis Henry 28:00
Oh, yeah, totally. And that was the other cool part, which was complete non industry, non subject matter expert. Right. So your reviews, like hold on backup us seven last 30 minutes, and I have no idea what the hell you’re talking about. Like, okay, yes, I need to get out of my bubble. And, yeah, pack them. Right.

Eric Dodds 28:22
Very cool. Well, congratulations on the book. That in itself is a huge accomplishment. Costas, I’ve been monopolizing the conversation. Yeah,

Kostas Pardalis 28:33
Let’s find out because you’re asking some very interesting questions. And I’m really enjoying what I’m learning here. And there’s a lot to learn for me. Okay, so I want to start with the concept of the account. And I’ll ask you to do a bit of data modeling actually, and try to model like an account, right? So what’s under an account like if I’m, let’s say, we bring like game design, Jimmy are like DB admin today. And you ask them to start creating like this key mother to represent the account, right? What would you ask them to put on the table there? And we can start with, with Hillary first and then

Hillary Carpio 29:19
put on this thought. I mean, we’re gonna start with our basic demographics and company size, we want to know what industry they’re in. We want to know sometimes how long they’ve been around. All those sorts of pieces, where their headquarters are, how many offices they have, the elements that help us understand if they could be a good fit. And then once we have those basics for demographics, and demographics, we also want to know what they’re interested in and that’s where that intent or third party data comes in. So first side fit, we have a score this base just off of that second side timing. Are they consuming information on our site and across the internet, about the things that we are selling to help us understand if they’re in the market? And I think there’s another side to which is technographic. So are they using tools that either complement a Snowflake tech stack or compete with a Snowflake tech stack that can help indicate whether they would be a right customer? For us?

Kostas Pardalis 30:14
Mm hmm. What’s your take on this Travis? Like from the sales perspective? Like, what would you add there? Come on, we have to make the model like, like, really complex? Here’s

Travis Henry 30:23
a complex bullet proof. Okay, you definitely need a unique identifier, because one of the things in sales and every operation is you look like an entity like Berkshire Hathaway, right? That’s a holding company that owns a bunch of other companies for a global corporate hierarchy like alphabet now, right, that owns Google Waymo, and all these other entities, then you think about, wow, if you’re selling to retail, you know, there’s Walmart in in Huntsville, but how do you go out and map all of the different Walmart locations and start to build out corporate hierarchies? So that’s a big data challenge and a kind of schema challenge of how do you duplicate those different entities? And how do you put parent child relationships around those entities? Because ultimately, that’s kind of fundamental for a go to market team, because it relates to questions like, how many customers do we have? Well, we’ve sold to the Japanese arm of Honda, and we have not sold to the jet, or the American arm of Honda. Is Honda, a customer or a potential customer, and you start to get into these kinds of strategic board level questions with the schema. So I think Hillary nailed it. That’s what I would add on. And then I would also talk about all the complexities where what accounts become, especially in an account based world, which we are obviously, big advocates of So zooming out, you know, there’s account based go to market, there’s product lead growth, there’s B to C, and there’s all these different go to market motions. But we’re very big fans of looking at everything in the world through the lens of that account. So under that account umbrella, you have to build out all of your sales opportunities, your potential deals, how much revenue do we think we can get from this company? Or how much have we gotten? You have to build out the people that were all the people that work at that business? What are their relationships to each other? And then you can start to continue building on that account model where, hey, well, we’re not just selling to them, we’re also servicing and supporting these customers. So what about things like support tickets and customer satisfaction? How do those fold in? And then you think about marketing campaigns marketing lead generation? And how are those driving the account? So it’s actually a really big pivot point, or core, or whatever the correct data term is for that model and understanding your business’s relationship to that business that really all falls under that account. And you can build that out to a very, you know, extreme degree.

Kostas Pardalis 33:10
Yeah, yeah. 100%. And, okay, like, usually, in companies, you have many different people, right. And he doesn’t have to get really big to get to that like, even like, medium sized companies, like even smaller companies, you have many different, let’s say, like personas in there, okay. But they interact with something that is as fundamental as infrastructure, right? Because at the end, like Snowflake is a piece of infrastructure like for the company? So how do you, like, distinguish, like, what types of people you have there? Like, I can think, for example, there’s like, probably a distinction between the buyer and the user. Right? Maybe you will tell me, but I definitely was like, probably probably, like, very crude. There’s probably more to it. I would say taxonomy of like people. And I want to hear from you how you see this taxonomy. You translate from the sales perspective. And then you ‘ll add, like, from the marketing perspective, and try to see if there are differences or like, contact points there.

Travis Henry 34:23
Yeah, I think we’ll probably have more contact points than differences in our answers. Because what we do is we do that exercise, right? We sell to and work with probably the most of the listeners of this podcast, right? And those are different distinct groups and functions that have different interests and pain points and use cases. And so we do a lot of work and specifically our product marketing teams. In our go to market teams, think about what are those logical personas and how are they different from each other, what do they look like and care about? And we group them together. So data engineers for us, you know, are very distinct from, let’s say, a VP of analytics, right? And different drivers and pain points and different ways we would provide value to those folks. The interesting part, just on the data level, is how do you go about bucketing people and grouping them in your systems and in your data warehouse, it’s actually kind of hard, not that straightforward. Because we collect titles from people, and there’s many creative titles in the world, people really like to have fun with them. So you have to account for all those different permutations of title and then make sense out of them and normalize them. And that’s actually a really helpful project that our data teams have done to support us is, hey, we actually have two axes of the people in our accounts. So we understand function. So are they data science? Are they data engineering? Are they analytics? Are they developers, or software engineers? And then the other axis, is that seniority kind of level? Where are they in the hierarchy that the conversation with an end user data engineer is going to be very different from the Global Head of data engineering, right inside an organization? So we’ve actually normalized those and then we track our engagement inside of accounts, against those kinds of, you know, squares or quadrants of buying groups, you know, how high or low are we in the account? And also, you know, which groups are we most engaged with functionally?

Kostas Pardalis 36:38
And before we go like to show our drivers like, either, like, let’s say, a natural bias towards like, a certain subgroup that like cells care more about like, and while I’m trying to shade here, right, like, okay, I can think of like many people signing up and trying like to use Snowflake, it’s not necessarily that people are agreeing, like, sign the cheque to buy it, right. Probably like in the bigger organizations, the people who sign the checks are like, Never saw like the user, the facials, like Snowflake. So like sales, let’s say focusing more on trying to communicate it, especially from the SDR point of view, right? Like more into the buyer, like, let’s say, like to the person that was doing like to write the check, or it doesn’t matter?

Travis Henry 37:30
It’s a great question. So that’s really, you hit it with the SDR piece, which the SDR plays a big role in translating a lot of that end user education, discovery, you know, folks playing with the product and attending events and learning who, like you mentioned, typically not people signing checks and making big budget decisions inside the organization. And so one feeds into the other. All those signals were probably having conversations with some of those end users, we’re understanding what’s not working in the organization. And we’re starting to map out a story about this business and about, you know, their data journey and their data challenges. And then what that does is that empowers us to go have conversations with the people who do sign the checks, which are ultimately who we need to get in front of. So we can provide a much more educated, relevant resonance point of view to those decision makers, because of those end users.

Kostas Pardalis 38:31
Cool, that makes total sense. And I’ll move to Hillary now. So what about the marketing side of things like this, right? Do you from your side, like focusing more, let’s say on the user instead of the buyer, or it doesn’t matter. And then we can get a little bit more practical and like trying to understand what’s the difference between the two, like from the perspective of a marketeer?

Hillary Carpio 39:00
Yeah, so there’s two different angles, we do sell the full buying centers, and we market the full buying centers. So in Account Based Marketing at any given time, in our larger accounts, is not just going to be one campaign to one account, we’re gonna have a message going to the CIO a different message in the CDO a different message to the DBAs different message to the marketing analysts, right. It’s going to be different depending on who they are. And the goal is that when Travis in the SDR organization makes their phone calls, that’s not the first time the teams have heard of us. And so when the DBA goes up to their boss, or when the CIO gets brought in, or the CFO, they’re not going to Snowflake and why do we need it. We’ve already hit them with a message and a value added super relevant to their job in the buying committee. So we’re tackling most of them now. The other side of that is that we follow the sales team’s lead. So there might be an account where they’re like, hey, ABM we have great engagement at the practitioner level, but we are stuck in finance. So we might go heavier into that. Finance department we might go heavier into a different role depending on what the sales team needs on their sales cycle and where they’re seeing traction and whether or not

Okay, that’s super interesting and sounds very complicated to be honest. I Okay, let’s start from the I’m trying again, like consider me like, like personal LogStash knows like absolutely nothing about like the good market emotions here that are in play. So what’s the hierarchy of like all these different functions inside like an organization like Snowflake? So we have a demand generation for exams, right? Which I would assume that they’re like, on the top of the funnel, like, let’s say someone has never heard about Snowflake, they are probably going to hear about Snowflake for the very first time from an activity that comes from them. Right? Yeah, what comes next is like Account Based Marketing that like after someone qualifies, like an account qualifies you get there. And then at some point, there’s another qualification that’s happening in sales starts or things are more intermixed, like how, like the account moves into, let’s say, like the funnel and who is responsible for each part of the funnel there.

Hillary Carpio 41:14
Yeah, so I’ll share an example that we added to the book that I think describes this really well, which is, if somebody goes to a marketing event that the sales team invited the people to the field marketing team, which is events team set up the agenda and got the room for the demand generation did advertising for an ABM helps with the advertising as well. And that person at the event ate the croissant and listened to the speech, who gets credit. And at the end of the day, we just care that they ate the croissant and got the information, we don’t necessarily care which of those teams gets credit. And when you think about it that way, there’s multiple touchpoints from different pieces and parts of the organization into those accounts that are contributing at any given time. So it’s not a direct one two punch like ABM is Dibley, ABM, first, then SDR one two punch. The thing that I’ll share, though, is that our demand generation team goes and targets our entire account base and our entire database, right, so they’re gonna get all of the named accounts that are owned by sales, they’re gonna get accounts that are owned by sales that our corporate account executive team will follow up on. And then my team and account base is only targeting up to 30% of those. So they’re going after a much, much broader audience. And they’re going to get a lot more touch points intentionally than my team is. And then same thing with Field Marketing, they’re going to be throwing events, and they can be massive, 12,000 people right at a, at a company conference, or they can be a boutique event that has 10 People that are all high level and they’re doing executive events of some kind. So they’re going to touch a different number as well. And so we’re really looking at between all of these different functions, how are we cohesively touching our addressable market? And how are they getting through the funnel together? So it’s not quite as linear as one team goes first. I wouldn’t say in general, a demand generation team is going to touch them first, naturally, and ABM would come after.

Kostas Pardalis 43:12
Okay, that’s what’s so awesome. And I always felt like the impression that comes probably more like from the more traditional kind of marketing, like marketing is always like a little bit of like, spray and pray. Practice, right, especially when thinking about demand generation, but ABM is like, actually sounds like it feels to me as the opposite of that. Like, it has to be very focused, like almost laser focused, like to even like to the people that they are getting, like targeted, right? Yeah. And the content, obviously, or like, whatever, like you’re doing there, are they how is this done? Because like, like, traditionally, let’s say all the tooling that most people are like,
exposed when it comes to marketing or like more into the first category than like the second. So how do you choose RudderStack? For example, it makes sure that Eric is going to read the right type of blog post or email or I don’t know, like, what kind of channels you are using. They’re like to deliver, like, the methods. But how do you do that? Because like, to me, it feels like magic to be honest. So yeah, I’d love to hear that.

Hillary Carpio 44:25
Yeah, so both of us are very data driven, but we rely on technology, right? So we work with different vendors that enable the capabilities to target individuals with display advertising, for example. So the little banner ads that follow you around the internet, that we can do it at when we like, put in our audience, it can be the company name, it can be the function, the title, the seniority, they provide us with these parameters that we can choose from and we can build that audience. We also use a vendor that allows us to put a person’s email address in as the targeting and so when we can get to that People level personalization, we can deliver a hyper relevant message. And then these vendors are starting to be able to reveal who actually received it as well. So did they view it? Did they click it, etc. So we rely on them. And then we also rely on our data team internally to take all of those data points and help make sense of it at a bigger level, because we can’t look at individual people all the time for everything. But that’s what it comes down to is, you know, innovative technology. And we tend to prefer to work with emerging organizations that are really building out the newest, best, most innovative technology for us to leverage.

Kostas Pardalis 45:35
That makes a lot of sense. And then, as with like, listening, like to traverse, and kind of like, also, from my experience, like with sales, sales, things like to give a lot of signal rights through the interactions that they have that are also like a very interesting signal, because it tends to be because it is personnel tends to be like very high, like, let’s say, bandwidth kind of signal, but it’s not signal that you can easily quantify. It’s more qualitative in most cases, or it’s really hard to argue the system and make it like electronically available and like, let’s not get into the LLMs. And like all that craziness, but how do you deal with the fusion of so much information coming from so many different directions that, in many cases, you chose to like? Not quantitative information, right. So like the number of signups or like how many times someone clicked a button or how much data they processed? Right? How do you do that? Yeah, there’s multiple sides.

Hillary Carpio 46:45
So on one hand, we have a lot of anecdotal feedback from sales that, hey, I was stuck. And this data or this campaign, or whatever it might be, is the first time I’ve been able to get a conversation in the last two years with this encounter. That type of feedback or quote, data that’s anecdotal is difficult to quantify in any way. And so we’ll capture that feedback and share it in our QBRs and other formats. And then the data points that are harder to normalize that are still data points, at the end of the day, we kind of funnel into an account engagement number. And so we’ll look at, in an account, how many people are responding to what their levels are, and then create a very sophisticated propensity model that our data team creates. And then we also have a more simplified version that if there’s no campaign responders, no SDR meetings, no API page visits, then they’re unaware. If there’s X campaign responders, or ABM visits or SDR engagement, then they’re aware, and then that goes down to highly engaged, etc. And so we do our best to create parameters to make sense of the data, regardless of what format it’s in. Because going back to what I said earlier, it’s useless if you can’t take action on it. So our job is to do our best to simplify it with the intelligence teams to make sure we’re not, you know, doing anything. That doesn’t make sense.

Kostas Pardalis 48:09
Yeah, that makes total sense. And how do you deal with the complexity of having to address like, so many different, like, types of like professionals in their rights like a data engineer, and data scientists, they both one way or another, they’re going to interact with Snowflake, that completely different like users, right? Like they are using? I mean, they think of the world in a very different way, obviously, and have other priorities. They even might have, like, from a marketing perspective, like completely different channels that you can reach out to them, right? Yeah. And from what it seems like things just get more and more complex than being simplified when it comes to data. Infra so how do you manage this complexity, from a marketing perspective, and from an accounts, like, account based like marketing, where your job is actually like to deliver the right message to the right person at the right time? And do it for all of them? So how do you love?

Hillary Carpio 49:09
Yeah, it’s a great question. Because I think that complexity is actually a super power. So you think about, you know, more traditional marketing and Persona based marketing, you’re putting them in those silos that are specific to these different roles. And the reality is humans are complex rules aren’t the same at every company. And so what we’re doing with Account Based Marketing is saying, This person based off of what we know likely cares about XY and Z, and if it overlaps a little bit into other categories, that’s fine. Our goal is to get them what they care about, and might what I was on my team is, you know, our job was to help them get a promotion. So it has to resonate with what’s important to them in their role at their company at that time, or everyone’s too busy to pay attention otherwise. So it has to be that hyper relevance in order to be hyper relevant. You have to be complex. You can’t simplify them down into a single A profile.

Kostas Pardalis 50:01
Yeah, that makes a lot of sense. Like, it’s very interesting what you said about like how you like to help them to get a promotion, I think it was like, very important like for like, everyone was one way or another involved in like, getting to markets like take nickel products, I think very rather stand that at the end we have people who, regardless of power mouth like excited, they are all the technology of the end, they are professional, since they have targets, they have the problems there. And at the end, they need to make a living, right? So that’s a great insight. Cool. One last question from me. And this is going to be a little bit more related to data and like, complex desire. We talked a little bit about that, like also with Travis about how complex like modeling on just the account can get right. But I would also assume that as time passes, like the same concepts that you are tracking, or even like the semantics of these concepts, right? They evolve, they change, even because I can do the simplest case, let’s say you add more product lines. So yeah, like when you started with Snowflake, it was a data warehouse. And now it’s a data warehouse. It’s a complete, like platform where you can do so many different things, right? So things change rapidly. And also like more people wondering like, that’s my like, I suppose my opinion is that the business understanding of the world out there, like changes through the experience that it gets by growing right. How do you codify that? And also make sure that you can evolve, let’s say your data artifacts without, like breaking anything, right? Because to me, it was like it feels that’s, that’s hard. That’s not easy.

Hillary Carpio 51:58
Yeah, I mean, on the surface, accepting that things are going to change. I think it’s the hardest part, I think there’s a lot of people in organizations that have worked so stinking hard to get to where they are that it’s hard to accept that might not be relevant in whatever amount of time that’s coming up. So first accept that things are going to change. The second is we work with great data teams. I am not capable of explaining how they continue to evolve our data and how they continue to change it other than never settling. And our CEO has a phrase that he uses about never being satisfied, right, always going for that next thing, always trying harder, or always looking for more anything, it’s that mentality that drives the ability for the organization to come together to drive that change. And then we use machine learning models and other more technical components that that team could speak a lot better to. But it starts with just accepting and knowing that things are going to change. And we have to prepare for that change from the beginning. So when we’re building our models, building our dashboards, building these different things, we have to make them flexible for the future as more data points come in, and other data points go out.

Kostas Pardalis 53:05
Yeah, that makes total sense. And the electronic, let’s say technology perspective, what about like the human like point of view, right? Because all these people like in the sales organization, or like in the marketing organization, they need like to have a common understanding of Watson would say, and I count these, which is the simplest public how as that involves, like, all these people needs to be trained in a way like, right, so from an I’m asking you, because you have like an experience, or at least like upscale and like, that’s, like pretty unique, how to have that ease and like water, like some good, I would say, deep, like how to make sure as a manager of these teams like to do it.

Hillary Carpio 53:46
Yeah, eliminate assumptions, right? If you assume that somebody knows what you’re doing or knows how to do the thing you’re asking them to do, as soon as not going to get done. Everyone has so many different inputs and apps coming from every direction, that it’s constant enablement, it’s reaching back out, you have what you need. Here’s the messaging, here’s the content. What is your priority? Have your priorities changed? So it’s a lot of asking questions, especially with our sales team, to understand what they’re focused on and make sure it’s aligned to what we are. At the end of the day sales are compensated financially, right, based on quotas. And so you can motivate change, if it’s aligned with their quota. If you are looking for change, misaligned, like you’re gonna have a mismatch. Yeah,

Kostas Pardalis 54:30
no, makes total sense. All right, Eric, I’ll give the microphone back to you.

Eric Dodds 54:35
Okay, I was trying to think of a good question to sort of land the plane on what I guess I would say for the listener out there who is on a data team, who is saying, Man, that sounds like such a good working relationship. I wish we had that at my company. What would you say to that person?

Hillary Carpio 54:58
Yeah, I would say You have to learn the business, right? So the things that you give back and the conversations you have to be relevant to the business that you’re trying to drive. And part of that is that no or this data doesn’t work isn’t really an answer that a go to market can can do anything with hearing out how do you help make sense of the data and share the caveats of where it might be, have some hangups where things might need to change, but you have to help the go to market team present that data back to the business back to the sales team. A lot of times these requests come from sales leaders, they come from marketing leaders. And if the data team I’ve had this experience says, it’s not we can’t do anything with it, right? Like, you can’t just go back with that. So if you can help that business leader with the data, put all the caveats on the table, so there’s no you know, hiding anything, then that’s going to help establish that relationship that you’re on the same team and trying to drive the same results for the business.

Eric Dodds 55:59
Yep, I love it. Well, Hillary has been such a fun show. Just so thankful to have you and Travis on to teach us so many things. And congrats again on your success and reading the book.

Hillary Carpio 56:13
Thank you, Ben. Great to be here. Thanks for having us.

Eric Dodds 56:16
We hope you enjoyed this episode of The Data Stack Show. Be sure to subscribe to your favorite podcast app to get notified about new episodes every week. We’d also love your feedback. You can email me, Eric Dodds, at eric@datastackshow.com. That’s E-R-I-C at datastackshow.com. The show is brought to you by RudderStack, the CDP for developers. Learn how to build a CDP on your data warehouse at RudderStack.com.