The Data Stack Show - 262: Beyond the Dashboard: Why the Next BI Revolution is Human with Ryan Dolley of GoodData
Episode Date: September 17, 2025This week on The Data Stack Show, Ryan Dolley joins Eric and John to discuss his unique journey from playwriting to leading product strategy in the data industry. The conversation explores the evoluti...on of business intelligence (BI), the growing influence of AI on analytics, and the shifting skill sets required for data professionals. Key topics include the challenges of adapting to rapid technological change, the importance of embracing engineering practices in BI, and the need for continuous learning. Listeners will gain insights into how AI is transforming data roles, why storytelling remains central to analytics, practical advice for thriving in a fast-changing industry, and so much more. Highlights from this week’s conversation include:Ryan’s Journey: From Playwriting to Data (1:05)Making a Living as a Playwright (3:02)Transitioning to BI: Night School and First Data Jobs (4:12)Storytelling and Data: The Art of BI (6:22)Early BI Work: Data Warehouses and PDF Reports (8:33)Moving from Utilities to Consulting (13:03)Building vs. Implementing: Product Strategy Lessons (16:37)The AI Shift in BI and Analytics (18:41)Automation Anxiety: The Human Side of Data Change (22:16)The Evolving Role of BI Experts (25:18)Adapting to Change: Learning Code and Experimentation (29:34)AI and the Future of Embedded Analytics (33:38)Capturing Intent: The Value of Modern BI Interfaces (37:03)Bridging the Data and Software Engineering Gap (39:13)The Historical Divide: Data vs. Software Engineering (43:06)Organizational Challenges: Where Does BI Belong? (46:05)Reflections on Self-Service BI and Value (48:46)If Not Data: Ryan’s Alternate Career Paths (49:04)Final Thoughts and Takeaways (50:17)The Data Stack Show is a weekly podcast powered by RudderStack, customer data infrastructure that enables you to deliver real-time customer event data everywhere it’s needed to power smarter decisions and better customer experiences. 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. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.
Transcript
Discussion (0)
Hi, I'm Eric Dots.
And I'm John Wessel.
Welcome to The Datastack Show.
The Datastack 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.
Before we dig into today's data.
episode, we want to give a huge thanks to our presenting sponsor, RudderSack. They give us the
equipment and time to do this show week in, week out, and provide you the valuable
content. RudderSack provides customer data infrastructure and is used by the world's most
innovative companies to collect, transform, and deliver their event data wherever it's needed
all in real time. You can learn more at RudderSack.com. Welcome back to the Datasack show.
We have Ryan Dolly on with us, and we have a ton to talk.
about, Ryan, thanks for joining us.
Yeah, I'm thrilled to be here, guys.
Thanks for having me back.
All right, give us just a brief background.
Where did you start your career and then how'd you end up in data?
Yeah, sure.
So I started my career as a student in playwriting and acting, actually, but I couldn't make any money.
And so I transitioned to data about 15 years ago and have been just going full steam at it ever since.
I'm the VP of product strategy, a good data, which is an analytics platform.
I'm the co-host of the Super Data Brothers Show.
show. And I really have spent my career focused on kind of the intersection of data and the
humans who consume it. Right. So that's really what drives me. Awesome. Ryan, excited to have you on the
show. One of the things that I've been thinking a ton about recently is workflows. It's about how
data, what data people are currently doing in their daily workflows and then like what they
will be doing, you know, in the next couple of years as we see AI get more incorporated into those
workflows. So I'm pumped about that topic. What are some of the things you want?
to cover. Yeah, I'm looking to hearing your guys' thoughts on kind of the experience of what it
feels like to work in data right now. I think there's a lot of excitement, but there's also a
little bit of fear or trepidition about like what do our jobs look like in the future. And I think
it's mostly good. It's exciting. So I'm looking forward to exploring that with you guys.
Awesome. Well, let's dig in. Awesome. Let's do it. Ryan, welcome to the Datasack show, a fellow data
podcaster so that's always super fun yeah that's unique it is yeah thanks for having me guys
a big fan of the show so I'm happy to be back awesome well okay I want to start out we
broached the subject when we were just chit-chatting before we hit record but this is so crazy
to me okay you started your career as a playwright and I wanted I actually want to hear about your
transition from that you know going to night school getting your degree and then getting into
because it's a really crazy story. But how many playwrights do you estimate do playwriting as
their full-time job? I think probably between 20 and 40 are people who make 100% of their
income writing plays. There's a lot of people who make some of their income writing plays, but not
enough. And so that's where you get the community of university professors in playwriting. There's
thousands of them, right? But people who just kind of pen the paper, I guess.
typing nowadays, writing plays, and that's all their income. It's very few. That's so interesting.
I was trying to think of other professions where there's that small, you know, because there's
lots of theater around the world, but that's a really small concentration of people, you know,
who that's sort of their only source of income. That's absolutely fascinating. Yeah, a play has to be
a huge hit for you to make a lot of money, right? Like a huge hit. And how many can you even think
of off the top of your head? Very few. Yeah. Sure. Yeah. Okay. Well,
Thank you for entertaining that fun fact.
So how'd you go from playwriting, acting into B.I?
Yeah.
It seems like a crazy, I mean, not a normal, you know, sort of,
maybe like over a long period of time,
but he went to school and then right into B.I.
Yeah, listen, it's really hard to make money in playwriting and acting.
And that was ultimately, you know, that was okay when I was 22 and 24 and even 26 a little bit,
but by the time 28 rolled around,
I was like, hey, if I was working at a job doing data entry, actually, and I was looking
around the department, and everyone in my department had been there at least 10 years.
I was the new guy, and they all hated each other.
And I just saw that, like, if I don't get some skills that where I can get out of this,
like, this is my future.
And luckily, that job was at Northwestern University.
And so I was living in Chicago at the time, and I was able to get a crazy discount, you
Northwestern's not a cheap school.
Oh, yeah.
But the employee did, yeah, I mean, it's really cost a ton, but the employee discount was great.
And so I went to night school to study information systems and kind of cobbled together
a course track, focused on, you know, kind of a mix of general computer science topics
and topics related to database.
There were no analytics or BI courses you could take back in 2008.
It was all database stuff.
It was just database stuff.
So I took database stuff.
And then I happened to know somebody who connected me through with a company in Madison, Wisconsin.
So when I was wrapping up that night school program, I applied for grad school at Northwestern,
and I applied for this job in Madison.
And when I got the job offer, I took it and moved.
And it wound up being the best decision I ever made.
I met my wife there.
It totally changed the trajectory of my life in so many incredible ways.
But that was really it.
And honestly, what attracted me to BI is, look,
you think about all the things you could do in technology.
And BI is really the closest to storytelling.
And that's what I had studied, right,
was what moves people, what motivates people.
And being able to kind of apply that lens in a business setting
by looking at data and then saying,
what's the message in this data?
And how do we communicate it in a way that will move business
people to take the right action was what was appealing to me?
Yeah, it is interesting. I hadn't thought about that because if you think about, you know, if you think about modern movies, you know, especially, I mean, we watch a lot of animated movies because I have young children.
Yeah, me too.
You know, but it's sort of, it's not that there are messages in there, right, but it's sort of transporting you to a different place and sort of creating a world or whatever.
But, you know, and I know the same can be true of playwriting, but there's sort of an integrity that you need to retain with, like, the character.
and what you're trying to communicate and everything.
Form of data integrity.
Exactly, right?
There really is there, you know, which is super interesting.
And it's pretty complex.
If it's a complex.
Right.
So, yeah, that's super, that is super interesting.
There's a DAG involved, in fact.
Yeah.
Well, you think about, like, when it comes to entertainment,
take movies, you know, it's hard for people oftentimes to even pinpoint what it is
that makes a movie compelling, right?
But it is the stuff you're talking about.
talking about like there's an almost it's an art but there is to some degree of science behind it
and it's and it has to do with you know the characters the world are they consistent are their
motivations clear do you understand the motivations right what do they i mean the art of drama is
like two people want something and only one of them can get it right and and that's ultimately
what it all boils down to and you know i mean business is a full contact sport and oftentimes
We, us and our competitors both want that next funding round or we want that deal with that Fortune 500 company and only one of us is going to get it, right? And data is part of the story that you tell to lead to that outcome. Yep. I love it. I love it. Okay. Dick, so your first job, you go into B.I. Yeah. After night school, what were you doing? What were you working on? And actually, maybe even start with what would your definition of B.I. be back then? Yeah.
That's a great way.
Oh, yeah.
Because, you know, it's easy for us to look back and say, oh, I started in BI, right?
But what we mean by BI now is very different, you know.
Guys, this has changed so much since 2010.
What I was doing back then was I got hired onto a team that it was a combination of a data warehouse and BI team.
And so we were building the company's very first enterprise data warehouse,
using an extremely rigid kind of waterfall project management approach where we were working on that data warehouse for, I don't know, four years.
I had joined kind of towards the end of that process, really, but they had been working on it for a long time.
They had invested a ton of money, and all to this magical day when we're going to flip the switch, and the data warehouse is going to turn on, and then suddenly value will appear.
And then the BI layer on top of that was really about the distribution.
distribution, honestly, of PDFs more than anything.
It was like how do we, you know, how do we create the form factor back then was it looked
like a piece of paper.
It was a standard report and it had a header at the top and a footer at the bottom and
a table of numbers in the middle.
And that was predominantly what we built.
Dashboards were, like you could build dashboards in the BI tooling of the era.
Most companies didn't.
They were, if it's hard to believe now, but they were a cool new thing that, like, if you were rolling out dashboards, that was pretty cutting edge at a lot of companies in 2010.
Now it's cutting edge to cut dashboards.
Exactly.
Like, that's, you know.
They had to move back to exactly what you said, like, tables.
Yeah.
Well, and it's, you know, the funny thing about it in some ways, like a table is quite easy for a lot of the AIs and expert systems we're trying to develop to interpret in a way.
that a dashboard is not. And so there is some degree of kind of going back to basics here,
but it was just such a different time. I mean, it was like, look, the job of data analyst,
I don't know how many people even had that job title in 2010. Even that, like you might be
a supply chain analyst or a financial analyst, but the idea that your whole job was just to analyze
data rather than to be a supply chain expert who happened, who was really good at Excel.
That's kind of what it boiled down to back then. And everything was done by a centralized team.
Nobody in the company could produce these reports, but the people on my team.
There was no, hey, I'm an analyst in a department and I have access to the BI tool and I can build my own stuff.
That was, it's not that was an unheard of pattern, but it wasn't widely adopted and certainly not where I was.
I was at a utility company.
So, you know, right there, utilities are cautious about adopting new technologies and approaches.
Yep.
Okay.
So then fast forward.
So what was your next move after that?
And then how did you eventually wind up at good data?
I was really frustrated working at a utility company.
I was really, I mean, I just, I saw that there was a lot of cool stuff going on.
And what specifically drove me to make the move was I went to, we were, the product we worked
with at the time was Cognos, which was a really huge BI tool in the 2000s and maybe the
first early 20 teens.
Well, I mean, it might surprise people, I mean, how many people, how many companies,
and he's probably still run it.
Oh, yeah.
It's so, so I'm, I got to be really good friends with a lot of people on the Cognos product
team.
I became, IBM has this program called IBM champions for the people outside of IBM who are the
greatest experts and advocates for their technology.
And I got that award five years in a row.
Wow.
So, so I got to know the product team really well.
Yeah.
And so I think, I mean, it's like many thousands of customers still and a lot of the
largest customers on Earth still have huge Cognos deployments.
Yeah.
So it's, I talk.
talk about it like it's gone, right? But it's really just, I don't work out of it anymore. It takes a long time for that entrenched stuff to go away. Yeah, yeah. Well, listen, if what you need to do is send out 10,000 customized PDFs to people, like, there's still no better tool in the world. And I mean, you can, you can, in fact, probably make an incredible living doing Cognos Consulting. Oh, yeah, you can. In fact, I think you may, you can maybe, like, there was a, if you go back six years, there was like a lull, but a lot of the people who are really great at that generation of BI tool,
are approaching or in retirement.
And so if you're an expert in those tools
and, you know, General Motors or Procter & Gamble
someone to come in and help maintain and build on them,
that's actually a high-demand skill.
Yeah, ride cognos into the sunset, me.
That's right.
Anyway, sorry for your enough.
Oh, yeah. So I got, listen, I got,
I just got frustrated because I went to a user group
and IBM was kind of showing the new version of the software
and I went back to the company I worked at
and they were like, well, it's on the capital project schedule
for five years from now.
And so I was like, okay, so I'm going to be using this version for the next five years.
I got to get out of here.
Yeah, yeah, yeah.
Because I just, like, I got too passionate about the work, you know?
And so just the pace of work, I could see that there were big changes coming to VI,
and I was a little worried that they were going to pass me by.
And so I actually started a blog, and that blog got the attention of a consulting firm based
out of Chicago and they hired me on as just a consultant there. And then I eventually became a
partner at that firm. Oh, cool. Yeah. Yeah, it was, and I'm still really good friends with those guys.
But it's the same story. I could see that there was another wave of innovation coming. And I decided
in 2020, 2021 to move over to the vendor side of things. And so that kind of takes me. It went to a
startup called Count, which makes a really cool BI Canvas. It's like the world's only multiplayer
or BI tool. I don't know how to describe it. And then from there to good data where, yeah, where I'm kind of, I'm kind of like in charge of, I'm almost like an executive individual contributor where what I focus on is go to market strategy, helping the product team and then representing the company in public. So Big Data London's coming up next week. If you're going to Big Data London, come say hi. I'll be presenting there with the good data team. Awesome. Okay. I want to sort of skip over the first move, you know, where you saw
the next version of Cognos, and it was going to take five years and see you said, okay,
there are changes happening. So then you started the blog. That makes sense. And then you moved on
from the consultancy. What big changes did you see at that, you know, at that juncture? Yeah. I mean,
I think that was in the kind of the just before the explosion of the modern data stack, right?
And so there were a couple things that I saw happening. One was that the innovation,
on the data front end had really stalled out.
Tableau is a great dashboarding tool.
Power BI is a great dashboarding tool.
But I was just asking myself,
like there has to be something other than this, right?
Like, this can't be the end state of the front end of data.
And so that really set me looking for what that could be.
And then there was also, you just, I mean, I could see coming,
you know, that was also when the lakehouse was a cool new idea.
obviously that's a pretty standard pretty standard approach nowadays and then and I got really
interested I started hearing about this thing called DBT yep and so kind of that that confluence of
events really made me say okay I and also I wanted to build something right I had spent so much time
I had spent so much time implementing and you develop really strong opinions about what
tools should do you know yeah yeah it's and so then I said listen I've gotten good at implementing
it's time to build, and that was the final thing that pushed me to move into the startup world.
What's the most surprising thing to you having transitioned from, you know, a customer implementer to a builder?
Like something where maybe if you look back to your Cognos days and you're at a user group and you're like, come on guys, you know, and then now you're on the builder side and you're like, oh, oh, look, there's a lot.
I will, if you're listening to this, I want you to imagine the vendor you're really frustrated with right now.
Okay, just picture it in your head.
And now keep in mind that they have hundreds or thousands of other customers who are equally frustrated, but with something else, right, that you're doing.
And you know that when you don't work for a vendor, but the complexity of juggling vision and customer requests is really, it's a
a really delicate balance. Yep. And it's hard to appreciate. Even as I describe it, it probably
seems like, well, yeah, like obviously that's what it is, right? But the surprise for me was really
in the complexity and delicacy of that process. Yep. There's so much that goes into what features
should we implement and why and for whom and on what timetable. And so that surprised me. And I'd say
the other thing that surprised me is just how often you wind up being wrong, you know?
When you're an outside expert implementing these systems, you have very strong opinions
and you think, oh, I can step in and like I will know exactly what to do because I've spent
so much time implementing these systems. But when you actually find yourself in that chair,
you learn that part of the process is placing bets and some of those bets are not going to turn out
and that's okay. Right. Yep, yep, man. Wise words for anyone,
looking to get into product leadership. Okay, John, I've been dominating the mic, because I know
you have a bunch of questions about this. So here's a lead-off topic for you, John. But I'll start by
asking Ryan a question. So that was that shift. And now we're in the middle of another one.
And I know that probably consumes an immense amount of your thinking as far as product strategy at good
data. So what are sort of your core beliefs about the current shift that's happening? First of all,
like what are the major things that are changing? And then how does that influence the way that you
think about product at good data? Yeah. The fundamental shift we're going through is obvious
in a lot of ways. And it's, of course, driven by the advent of artificial intelligence in the
industry, right? And I think that my fundamental belief has to do is really that the front end
of data is going to radically change the way we are, we, we,
present data is going to change, but that it's, the change is going to be a lot more about
how we facilitate communication with our end users and how we take that, and how we turn that
more and more into a two-way process. You know, in BI, a lot of times, you're building,
shipping, and moving on to the next thing. And that is, is totally going to change. And I think
it's going to be a lot more about, you know, co-creation with our consumers, iterating through
problems, and being the front end of a combination of human and expert systems that are running
behind the scenes. And I think a big part of that is that, you know, in BI, we have really
struggled to embrace engineering practices and code. And that's just not going to work going
forward. I mean, it just isn't. These AI tools, language is the second L and LLM, right? And so
this purely UI-driven development process and UI-driven consumption process is not going to hold
going forward. And that's something that we've invested a lot in at good data is building up the
tooling around that as kind of the foundation of what we see going forward as AI becomes a bigger
part of the data front end. Yeah, I couldn't agree with you more on that. But I think it poses a
question. We talked about this before the show. And I'll actually paint a picture. I've got a picture
in my mind from when I first started
in data and IT in general
almost 15 years ago
and then a conversation I had this week
which I think will illustrate this
when I first started
however old I was 23
24 I remember sitting down with
some ops users and some accounting
users to automate some reports
like distinctly remember
they pull up Excel pull up
a chair Windows XP
whatever it was at the time
and actually like watching you know the
process in Excel. And I, you know, was newly equipped with access to the database and knowledge
of SQL and like super excited about automating these reports. But I remember the emotional
reaction of the people like one of one of the people I can remember specifically of like
essentially like, well, what am I going to do? Like what if this gets automated? Like this is what I do
every day. And I've had some other reactions like that now. But this week, I sat down with a data team
and went down that road with like some AI workflows and you know some and this like data is
code and then using AI with BI as code and some other kind of as code things and I didn't get
that explicit reaction but it was going there and it is a different experience for people where you're
used to being the technical person to sit down and automate this other manual part of the
business and it's really like some script flipping yeah to be on the other.
side of like looking around like on this, there's like three, you know, like three analysts
types on this call, like looking around being like, but like if it can do that, like thinking
to themselves, you know, I'm kind of guessing here, but thinking to themselves, well, what am I
going to do? They need all three of us now? Yeah. I think that's a big deal for people. It's a huge
deal. And I will say, look at who's excited about what's happening in data and AI right now.
It's vendors and executives and directors are very excited, right? Exactly. Yeah. And
And how are data analysts feeling?
I don't think the excitement is quite so strong when it comes to the people whose careers
have been built on implementing these systems.
And so I think that's a look, it's a legitimate concern.
I think if you look back in history, the idea that BI people in 2010 had this dream that
we would, through BI, we would eliminate Excel.
And like, did that happen at all?
We had the ability to automate a lot of what end users do in Excel.
And if anything, there's more Excel today than ever.
And so I-
BI is an avenue to download to Excel.
I mean, what's the most popular button in every BI tool, right?
It's the export button.
Yeah, for sure.
You know, I'll tell you a good data, the export button itself is even an embeddable feature.
You can embed just the export button into one of your other-
Wow.
Yeah, like just embrace it.
Yeah.
Go all the way, right?
But so I think that, look, there will be an element of that, okay?
that like, like, I remember having that experience to sitting within users trying to automate
their spreadsheets and them clearly having this visceral, like, this is my job. You can't take this
from me. Exactly. But let's be realistic. There are many elements, I believe, of what if you're a
Tableau expert in your whole career is built on your expertise in Tablo, your ability to manipulate that
Tableau user interface, to get it to look just the right way. I think there's a lot of what you do today
that a machine is going to be doing in relatively short order.
So my question to that person, I get why that's really scary.
My question to them would be, is working for a week or two weeks or a month on a dashboard
and then delivering that dashboard and maybe people use it and maybe they don't?
Is that really actually that great of a job?
It might pay pretty well.
But do you really love it that much?
moving the pills around on Tableau, like, is that really you want to do that for the next 20 years?
I can tell you I didn't when I did a similar role in Cognos.
I got sick of it after about five years.
And then I said, I want to move out of this.
And I think that, look, what these systems are going to allow us to do, I think I predict that as a B.I. person in particular, your job is going to become a lot more about interacting with the people who consume your
analytics and understanding more deeply, what are they actually trying to do? And then in conjunction
with AIs and other systems, figuring out the best thing to deliver for them. And to me, that's a
more exciting job. I personally like digging into my consumers' problems and figuring out what
they really need more than I like the technical question of how do we make this dashboard look just
the right way. Right. Yep. So I think just really quick on that, I think that the trouble with
the GUI thing and then, you know, and then AI changing a lot of that. And I think I can relate to some
people that are in this weird middle ground because it's really where I was 10 years ago. So
I was doing BI work and got more technical and did DBA and like ops work after that. But you do
if you, and Tableau was one of the, like when it very first came out like a decade ago was one of the tools I
used and you and if you use it enough you do get to this like flow state with the tool yeah it's
like it's special like it really is like and people are like this with excel too and at one point i was
that way with excel probably not as much anymore and there is something like really fun about that
and i think if you're there and like tablo's my tool or excel's my tool or whatever and you have
flow state with that tool and somebody's saying like look the future of this was we're going to
split this. You're going to have to learn all these command line tools and get good at context
engineering and like this, that and the other. And to you, it's going to feel like, I'm not that
person. Or you're going to have to like go interact with users all day and do requirements. Well, you know,
like, that sounds terrible too. I think there are some people in that middle ground that, like,
it's going to be a little rough. Because like they don't necessarily want to like do all the
frontline requirements gathering like business interaction clarifying problems part. And they don't
necessarily want to like really dig in a little bit deeper technically like get good at prompt
engineering context engineering tooling set up data platform stuff yeah and i don't know how to help
that well i would say though i don't i mean and i don't want this to come across as you know as overly
negative but i think that person would struggle in their career eventually anyways because the tools
themselves would change right if you have that sort of attitude where it's like well i really just like
flexing this one muscle that is
larger than most other people in this
sort of very singular context.
And I don't really want to learn the business
and I don't really want to learn new technology.
Like, it doesn't matter.
You can replace AI with anything.
You know?
Like, say the company had decided to like
rip out Tableau and put some money.
Right.
Like it would be a, you'd have to be the same
learning.
Look, Tableau, to be clear,
Tablo itself is going to, is trying to replace you.
Yeah, sure.
Yeah, yeah.
It's not just about the tool.
Yeah, but I think, but there, there's a class of people.
I think we've all probably, like, work with people.
And I think that's going to be a challenge.
And it's always been a challenge, like, for example, like the Cognos thing of like,
man, you got to switch companies a lot, only find ones that use Cognos.
As Cognos may decline some, it's going to be tricky.
But you can kind of do it if you really care about Cognos and just transfer companies
and move to like a utilities industry from more progress.
But like that class of person, which isn't like, not a ton of people that way, but I think it's
going to be a harder for them. It's going to be harder for them where you could have made it
happen. And now I don't know if you can do it. Listen, I 100% understand how those people are
going to feel. When I made the decision to get out of the Cognos world, it was the hardest thing for
me was actually leaving that world behind. Because that was my professional identity. Yeah,
I was a, I was, you know, it's funny to say, but there were a few, there were a few famous
Cognos people in 2019. A friend Paul, who literally goes by the name Cognos Paul,
like he's one of them, but I was one of them too. I would go to a conference and people would
write, because I had a YouTube channel. I focused a lot on Cognos tips and tricks. So people
would be like, oh my God, I learned how to do this thing by watching your YouTube. And it was my
whole professional identity. It's really hard. It's a really hard transition. And so I completely
understand anyone who's looking at any B.I. tool or really any piece of technology that's
looking like it's going to radically change in the next few years and saying, like, I don't want
to do this. I, right? Yeah. A hundred percent feel for them. Yeah. But this is coming. So, you know,
like you, what would you tell yourself if you give that version of yourself advice of like,
hey, you're about to make this transition? Like, you got to do it. Like, is there anything that
comes to mind of like, man, if I'd only known this, it would have been easier? Yeah. I mean,
I think the first thing I would say is, don't stress so much, dude. It's going to be okay. Right.
But I think, like, you know, there are things that could have made it easier.
I think so one thing that I, that may be hard for B.I. people in particular is, and I think we,
I said this earlier, but I like, a lot of B.I. people are allergic to code. They just, you know,
they're just allergic to it. And look, there's no way around learning at least the basics, right?
Luckily, that you have now these AI helpers who are going to do a lot with you.
I don't want to say for you, because you're going to have to learn.
Right.
If you can't tell at all what they're doing, that's going to be a problem.
But they will do a lot with you and help you get there.
So it's just vastly easier.
If you're a B.I. person and you're like, I don't know any Python.
Sure.
Right.
I would tell myself to embrace the coding thing a lot harder and earlier in that process.
You know, and I think the other thing would be to like to really try to spend,
more time experimenting with these technologies. It's just easier and easier than ever. If you're a
B.I person and you hear someone say like something about DuckDB, you know, you can spin DuckDB up
on your MacBook and start messing around with it. Like the barrier to that is so low now. Like just
go for it, right? Yep. And it's going to be scary. But, you know, if you can learn some code and
learn, like I think a lot of the future, you know, I think we're heading towards the future where
there will be some degree of like guided assembly of front-end BI systems, including, you know,
like in-browser databases with in-browser visualization. And if you learn these fundamental
technologies now, you're going to be in a pretty good place, I think, in five years.
Yep. Yep. Yeah. We're going to take a quick break from the episode to talk about our sponsor,
rudder stack. Now, I could say a bunch of nice things as if I found a fancy new tool. But John
has been implementing rudder stack for over half a decade. John, you work with customer event data
every day and you know how hard it can be to make sure that data is clean and then to stream it
everywhere it needs to go. Yeah, Eric, as you know, customer data can get messy. And if you've ever
seen a tag manager, you know how messy it can get. So rudder stack has really been one of my team's
secret weapons. We can collect and standardize data from anywhere, web, mobile, even server side,
and then send it to our downstream tools. Now, rumor has it that you have implemented the longest running
production instance of rudder stack at six years and going. Yes, I can confirm that. And one of the
reasons we picked rudder stack was that it does not store the data and we can live stream data
to our downstream tools. One of the things about the implementation that has been so common over all the
years and with so many rudder stack customers is that it wasn't a wholesale replacement of your
stack. It fit right into your existing tool set. Yeah, and even with technical tools, Eric,
things like Kafka or PubSub,
but you don't have to have all that complicated
customer data infrastructure.
Well, if you need to stream clean customer data
to your entire stack, including your data
infrastructure tools, head over to
rudderstack.com to learn more.
I want to return to a statement
that you made, and this can steer us a little bit
towards talking about good data.
Yeah.
But you talked about who's getting excited about this.
And we had a guest on recently,
actually a former co-host of the show,
cost us from type of AI. But who said, you know, the interesting thing about this is everyone in
technologies really used to disrupting other people. Yeah. And, you know, they're getting disrupted,
right? And it's like, whoa. Wait a second. Yeah. Yeah. We created the monster and it's attacking us.
But, you know, you said directors, executives, etc., are getting excited about this. Vendors.
Vendors, for sure. Why are they getting excited about it? And I want to hear that, you know,
put some actual feed on that from the perspective of, you know, who's a director who's interested,
in buying good data? Why are they excited about it?
Yeah. Yeah. I think so good data in particular, our lane in terms of analytics platforms
is we're really best when you are building out a suite of analytics products that you're
offering your customers, right? So like this is what you would classically call embedded analytics,
but I think it's moved way beyond what people think of when they hear that term. They think
like, oh, I'm going to eye-frame some charts into my app.
Yeah, right.
Yeah, yeah, yeah.
And that's not what it's about at all anymore.
And so there's really two, I think we're lucky.
The people, because our customers are using us in many instances to generate revenue,
what these directors and executives are seeing is the ability to build exciting new products
that have an AI component that help, you know, we do a lot in financial,
services. So we're talking with someone right now about like portfolio, AI-driven portfolio
rebalancing, right? And just looking at your portfolio and saying, okay, here's where you're,
you know, having the analytics to, first of all, present the portfolio to the end user in a
comprehensible way and then say like, okay, here's what your investing goals are. And here's
where your current portfolio is out of line with those goals. Here's a plan that we can execute
over time to get you in line with those goals, you know, do you want to execute it, right?
That's kind of their vision.
And that's a revenue generating thing.
Either directly through subscription services or a serious competitive advantage versus other
investment management providers.
And so that's really exciting.
What I hear about, too, though, and I think I actually wrote about it on LinkedIn today
because I was a little frustrated with a cold DM I got from someone trying to sell me.
It was like, hey, my AI product will replace your product managers.
And so that's the other bucket that's out there.
Like there are executives and directors who are like, oh, I can cut headcount.
Yeah.
And I think that's probably true to some extent, and it will be more and more true over time.
But I think the people who are going to win are not the people who just see this ability to reduce headcount,
but rather the people who see the ability to produce new and better products and to be more competitive in the marketplace.
And at good data, that's mostly what our customers do.
So I feel really lucky that those are the conversations.
I'm having, that it's not about like, oh, can we get, can we get by with 20% fewer financial
analysts, but it's rather like, how do we use this technology to out-compete our competitors?
Yeah.
I love that.
And I love also how, you know, it's interesting because one of the first really major
pushes in BI when it comes to, you know, when it comes to AI and how it is changing the
space is, okay, you have a chat pod and you can ask it all these questions.
Yes.
But I think hearing about the types of things that you're enabling companies to do where the ultimate manifestation of this is actually just an awesome product experience for the end user is right on.
It's like, okay, the AI kind of disappears and you're able to do something that kind of seems magical, right?
Like, wow, this would have been really hard or would have taken a human, a long time or a lot of effort to do to sort of rebalance your entire portfolio relative to your goals.
And that's really cool.
That's, I love that.
And I mean, you think about, because, yes, every BI tool has a chat bot that you can say,
like, show me revenue over time and it'll generate a line chart, right?
I mean, that's really table stakes.
But the ability I describe, so one of the things that's really exciting to me about it,
actually as a BI person, when it comes to the interfaces we're building.
And in the chat interface, I think, on the one hand, it's easy to say like, chatbots, whatever.
But on the other hand, those chatbots capture something that,
as a BI person, I always wished I had and I never did.
And that is intent, right?
So just knowing what my users were asking
was something that I could, you could never know.
You would sit down with a power user and say like,
please, you know, show me what you do.
But actually the ability to have this auditable,
queryable log of all the questions someone asked.
I mean, it's incredible.
And in that particular example I gave,
it's like, you know, the end user's ability
articulate or the consumer, I hate saying end-user nowadays, the consumer's ability to articulate,
okay, here's my rebalancing portfolio, and then you can look at it and say, oh, that's too
aggressive for me, right? Or it's too conservative for me. I want to have more in stock and less
in bonds or whatever it is. That goes, and then to do it, that goes so far beyond what
BI tools are traditionally doing that are just focused on that internal use case.
That's just like, how do I get people to charts faster?
That's valuable, but the real value is in these interactive systems that display data,
help you understand data, and then take actions.
Yep.
Let's talk a little bit more about the software engineering data gap.
You mentioned this earlier, John, I know this is something you think about and talk about a lot,
but what are the, you said, getting closer to code is some advice that you would have given
yourself.
But Ryan and John interested in both of your opinions because you're doing this work every day,
what are the specific gaps?
Like, what are the really, what are the big ones and what are, and maybe even zero and just a couple of technologies or tools that you think are the big ones?
Yeah, I mean, so from the front end perspective, I think I would look at learning, I mean, just bluntly, like I would, you need to understand what a YAML file is and how to interact with it.
You should have some understanding of like document storage and retrieval.
you should know React, I would learn, and I would learn Python. I mean, it doesn't have to be React,
but that's the most popular one amongst our customer base of good data right now is React, right?
And the reason I would learn these things is because I think that, I think in pretty short order,
it will not be terribly difficult to develop a front-end application using technologies like these
and with AI assistance. And so understanding them, I think, is going to be quite important.
And if you look at the BI tools in general, they're all adding SDKs and React libraries and all of this sort of stuff.
We got lucky at good data.
We've had them for five, six years now.
But there's a race to add these to the front end.
And there's a reason for that.
It's because probably the future of the data front end is less about going into a BI suite.
And it's more about someone has curated a data experience for a particular use case.
And the end user is going to go into that.
And what's that going to be built on?
It's going to be built on the same fundamental technologies
that the rest of our applications are built on.
Yep.
Yeah. So this is super top of mind for me
based on some conversations I've had recently.
And it's funny because I was walking through
a solution with a team the other day.
And I just found myself like just,
and it was on a video call, so you're like trying to gauge
or you're trying to like share your screen
and also like gauge like people's reactions
challenging it in and of itself.
But I just had this.
suspicion, like, I don't know if they know what I'm talking about. So, like, rewind a little bit.
Like, are you familiar with this concept? Are you familiar with this concept? Got some crickets.
You know, and essentially what it ended up being is, so this context was using AI tool,
AI tools for analytics engineering, essentially. Yeah. So I kind of glossed over things like,
so, you know, you'd set up a linter this way. And I'm like, do you guys know what a linter is?
And then, like, go back. It's like, okay, SQL fluff. This is what SQL fluff is. It can,
you know, format your code, it can, you can set rules for how you want, and just like went
through like, what is a lender? It needed to do that. And then like, you know, fast forward
some more and then talked about like tool calling and MCP servers and had to kind of go down
into that with some rabbit holes, which they didn't have context for that either. So there's just
a lot of things like that. Let's call it environment set up, which in Devland, like, like, let's
say Circa, I mean, it's been over 10 years like when Docker like really became a thing.
there was a lot of like dev experience stuff
especially around that like docker time
where everybody's like reinventing dev experience
around containers and then
further plenty of iterations after that
and with like CICD pipelines
when those like were first to thing
but that missed data
like it almost completely missed data
I don't know why whatever reason
maybe because of all the like vendors
with heavy influence with the GUI experience
that's probably a lot of it
yep so since all that missed data
like as I sit down and talk with teams
like that's often where I have to start
especially for like that analytics and data engineer like some of them maybe are more familiar but especially
like that analytics engineer or like trending toward the analytics engineer and then trend toward
bi like that whatever skill set you call that I don't know what the future of that is
separate conversation probably but that skill set like you have to go back and explain and talk
about some fundamental things to get to like okay and then like now we're at a spot where
you're excited at the beginning of this conversation like hey I want to use AI
to like make things and you're like well let's rewind here and then after an hour and I've completely
lost you this is you know this is where we need to be so that that's a challenge and I think it's
going to be a challenge for a while John you are 100% correct something happened in the early 2010s
where the entire tech industry zigged and data zagged and it is just I mean just things like like
basic CICD on many data teams they don't practice it they don't know how to
to do it. Until very recently, the BI tools. You can version control. We don't have to do
see it. Right. Just version control. Well, you know, I, like, I mean, a lot of
BI tools don't, there's no undo button, right? You do deployment and there's literally no
undo button. Yeah, I mean, I just couldn't agree more. Like, teams have to learn this. This is the,
we are behind in data. And I will tell you, the closer you get to the end user, the more behind
that person is likely to be. And I haven't, you know, I don't know that I have a theory that
explains it, it might be just that the quest for self-service led the vendors and the teams
that implemented those vendors to really make an assumption that everything's going to be
UI in the future and like we can't have any coding because the end user is going to build everything,
the consumer is going to build everything and they're not going to code. Yeah, I do think that's a big
component. I think cost and democratization of technology was a big thing.
as well. So let's just think of an example, right? So 2010's, early 2010s, the data warehouse is
just entering the scene. Yeah. Right? I mean, today, it's just like, okay, and, you know,
everyone spends up snowflake, you know, you can just do analytics without even really thinking
about it. Yeah. That wasn't true back then. And so in, you know, you had major vendors who were,
you know, who were sort of driving the warehouse and BI. You had the huge UI push from the major
Vendors, right, because all the other stuff wasn't democratized. But then on the software development
side, you mentioned React, right? It comes around in the early 2010s and had a slow start, but is
open source, supported by one of the big fangs. And, you know, in a couple of years, become sort of
the de facto way to, like, do software engineering. And then by the, you know, by the end of that
decade run, even though the hacker news comments would suggest that there are like wildly different
opinions, you know, on how to do things right. Generally, it's like, okay, if you're building a
modern web application, like most of the world would say, okay, here's how you do it, right? You have
a database, you have these APIs, you have React on the front, you have CICD, you know, we're using
TypeScript, you know, just, it's like, okay, this is just kind of how you do it, right? And whatever
the variations are is fine, but that sort of open source drive for all of that, you know,
even if you just think about the React ecosystem, was a huge force, right?
but it took a really long time for data warehouses and other tooling to get to the same level
of ubiquity at a cost that made it really easy for anyone to do it.
And it's maybe not really there in some way. In some ways. I mean, it is, it's mostly there.
But the other factor, too, that I experienced along the way was this, like, data is odd.
Data at most companies, there's a lot of companies, and it's funny to watch it.
Like, okay, we're going to have centralized team. It belongs in the IT.
work. It's like, okay, great. And then you've got, like, back in the day, you have like database
administrators and some more technical people. And then like, and then you had some BI people,
but they reported in the IT. So they got sucked into that stuff. And we're generally bad at
requirements. Typically. Okay. It's like, oh, this isn't working. They're not getting things done
that are valuable for the business. We're going to rip out some of those people, have them
report to the business. And you're going to split it. You're going to have like, all right, the most
technical people, the data engineer, modern, you know, DBAs back in the day, data engineers today,
like they're in IT, then we have these analysts floating around.
They're in marketing and supply chain and blah, so we're going to do that.
And then there's communication issues with that, with handoffs and blah, blah,
then you're going to like further decentralize it and we're going to buy like a gooey tool,
go no code.
We just want to IT, you just like tell us where the data is.
We'll connect to it and we'll do our own thing in Tableau land.
So you have all those iterations of it, right?
And it's like, no wonder we're in a spot that we're in.
Yeah.
We don't even know where its home is.
Like we don't even know where this thing should live.
Yes, exactly.
That's really the story of the front end, right, of data.
And it's just like the further, the closer you get to the source applications,
the more it seems people are comfortable with this engineering approach.
But, yeah, I mean, look, I sometimes call that the 2010's theory of value for data.
It's like, how can we get as many people as possible to learn how to use the self-service BI tool?
And then it's like an emergent process.
Out of that, value will emerge.
Yeah, right, right.
And it didn't.
Find insights and discover.
new thing. Exactly. They'll just, and it didn't happen. And so we like, you asked earlier,
well, why, like, why did you make the move kind of out of implementation? And part of it was because
I didn't believe in that story anymore. And I wanted to know what was coming next, but it was
pretty clear that people had made massive investments in these self-service BI platforms that
were still really only used by power users. And the idea that every, oh, look, everybody in finance
can just build their own dashboards
and pull their own queries
and all that sort of stuff,
you know, it didn't catch on.
It didn't deliver the value
that people had thought.
And I will tell you,
I talk to a lot of executives
at companies right now
who are looking at the tools,
the BI tools they bought in that era,
and they're saying,
we spent $10 million over the course of a decade on this
and what did we get out of it?
For sure.
Totally.
Well, we are close to the buzzer,
but I have a non-data question
or maybe non-data question
for you.
Okay, so from the stage to data, if you weren't working in data and you couldn't work in data at all, what would you do?
That's a great question.
I mean, it's easy to say, well, I'd just be a successful playwright, right?
It'd be one of those 40.
One of the 40.
It's easy.
Well, with AI, that's really easy.
Yeah.
You know, it's a great question.
I find, I mean, I always wanted to be an author of some sort, right, since I was a little
kid. So if tomorrow morning it was like, you can't do this anymore, I'd probably take a run at
that. That's a hard road. You know, the other thing, I mean, I've been really, this is kind of weird,
but I've been really looking into like, I feel drawn to doing something in the real world.
You know, like, I don't know what that thing would be. Like, I got it in my head recently that
my wife and I should buy the laundromat in town. Yeah. Just run that, you know.
Totally. Or I'll tell you, here's an interesting one. I don't know how my son did.
this but there's this YouTube channel. It's the BBC archive. And he found this, he's seven,
this kid. He found this documentary about the last lighthouse in Britain that was run by like
dudes who lived in the lighthouse. And it was into 70s. So this lighthouse is now fully automated,
right? Like everything else. But there was just something so appealing about this, about watching
these guys like clean the lens by hand and light the gas lamp, you know, to keep the
ships coming in. And so I don't know what it would do, but what it would be, but like something
where, where I'm doing that, right? The modern day equivalent of wiping down that lens,
if you can find it, like, that's the next career. Yeah, I love it. I love it. Well, Ryan, it's
always a pleasure to have you on the show. Love the conversation, and we'll have you back again
sometime soon. Yeah, thanks, guys. It was a real pleasure to be here. The Datastack show is brought
to you by Rudderstack. Learn more at rudderstack.com.
Thank you.