00:00:04 Tim Gasper
Hello, everyone. Welcome. It is Wednesday. We have drinks. We have folks who want to talk about data. It's time for Catalog and Cocktails presented by data. world. It's an honest, no bs, non- salesy conversation about enterprise data management. I'm Tim Gasper, longtime data nerd, product guy at data. world, joined by Juan Sequeda.
00:00:23 Juan Sequeda
Hey, Tim. I'm Juan Sequeda, principal scientist here at data. world and okay, here we go. Today we're talking to the guy that if you have never heard him, seen him, read about him on LinkedIn, it's literally the definition of living underneath a rock. Because even if you don't follow him, all this stuff needs to show up on your feed. I mean, assuming you work in data stuff and that is the one and only Ethan Aaron, finally we get to have you on the podcast inaudible so long.
00:00:47 Ethan Aaron
I'm excited to be here.
00:00:49 Juan Sequeda
And yeah, so just sadly, I heard that you sadly spent too much time on LinkedIn and also-
00:00:59 Ethan Aaron
Decision, but it's sad at the same time.
00:01:02 Juan Sequeda
But you also spend a lot of time with people in the real world because you're the organizer of the low- key data happy hours in New York City that has spawned other places to go too. So, Ethan, how are you doing?
00:01:12 Ethan Aaron
I'm good. Right now, I am not in New York City. I'll be back in time for the next low- key happy hour in a couple weeks. Right now I'm in North Carolina, so I'm working remote from the beach for a couple weeks with my family. So enjoying it. It's hot but I'm enjoying it.
00:01:26 Juan Sequeda
Awesome. Well, let's kick it off on our tell and toast. So what are we drinking and what are we toasting for?
00:01:32 Tim Gasper
Yeah, what are you drinking, Ethan?
00:01:34 Ethan Aaron
So I just went into the fridge and the only thing I could find was a, not your father's root beer. It is alcoholic root beer. That's all I got. Yeah.
00:01:43 Tim Gasper
I've never had a hard root beer before. Are they good?
00:01:46 Ethan Aaron
It is unbelievably sugary. It's good. It tastes like a root beer is the answer. Yeah.
00:01:54 Juan Sequeda
Put some ice cream in it too. That may make even more sugary like a hard inaudible.
00:01:58 Ethan Aaron
That actually sounds great.
00:01:59 Juan Sequeda
How about you Tim?
00:02:01 Tim Gasper
I'm drinking some Glenmorangie, some Lasanta. So pretty good scotch. What about you Juan?
00:02:09 Juan Sequeda
I am at home right now and I have this bottle, it's called Ilegal Mezcal. And there's a story behind this is that it originally comes from... It's called Ilegal because they were literally bringing it in into the US from Guatemala and there is this little bar in Antigua, which is... I mean Antigua, Guatemala is a great place to go visit and a bar called Café No Sé where they would actually be making this. And so I had friends who actually lived in Antigua and they would bring in stuff. So I met those folks here in Austin and I think it's now mass- produced, but actually pretty good.
00:02:45 Ethan Aaron
Now it's owned by AB InBev or something.
00:02:46 Juan Sequeda
Yeah, I don't know it's like that. It's supported by Ilegal Mezcal brands in Miami. Yeah, whatever. Its funny story is they would actually... So here in Austin there's a bar called Clive, it's still there I think in Rainey Street and inside Clive there was a little house that they had built and it's a replica of this Café No Sé from Antigua and that's where were-
00:03:07 Ethan Aaron
That's cool.
00:03:07 Juan Sequeda
So they were basically replicating the bar in this little town in Guatemala where they're bringing them. Anyways, so good memories. Love it. And toasting for what? What are we toasting for?
00:03:19 Ethan Aaron
Community.
00:03:20 Tim Gasper
Business value.
00:03:21 Ethan Aaron
Community and business value.
00:03:21 Juan Sequeda
Community and business value. Cheers to that.
00:03:22 Tim Gasper
Okay, cheers.
00:03:22 Ethan Aaron
Cheers.
00:03:25 Juan Sequeda
So our warmup question today, other than data, what's your favorite thing to rant about?
00:03:32 Ethan Aaron
Ranting about? So there's ranting and then there's talking about. In the data world and on LinkedIn, I only talk about data and business value and whatever's going on. In personal life, I also love fixing up houses. So I never post about this on LinkedIn, probably should because I have some fun stuff I'm building. But I have one house right now where it's a thousand square foot house and we have torn everything out of this thing. We ripped all the drywall out, all the siding off, had to replace every single window in this house, the kitchen, the bathroom. And to me, I love both going through the process but also whenever I can actually working remote from Ohio where I have some family and physically fixing up houses. So I can talk about that for ages. But that's my other thing I can go very, very deep into because I just love it. I love the physical aspects-
00:04:19 Tim Gasper
That's awesome.
00:04:19 Ethan Aaron
Not being at a computer.
00:04:20 Tim Gasper
You do a home flipping or anything like that or...
00:04:23 Ethan Aaron
I'll fix things up and then rent them out. So not flipping them to sell but flipping them or improving them to rent.
00:04:31 Tim Gasper
Oh, that's awesome.
00:04:32 Ethan Aaron
Yeah.
00:04:32 Juan Sequeda
That I did not expect. How about you Tim? What do you like to rant about?
00:04:41 Tim Gasper
Other than data, which is a lot of my ranting too, I like ranting about, especially space stuff. The latest developments with SpaceX and things like that. Annoying my wife with it, I'll be like, " Oh man, they're about to launch the next rocket." And she's like, " I don't care, why are you telling me that?" But I'm going to rant about it anyways.
00:05:03 Ethan Aaron
How often are they launching rockets? I feel like it's every three hours at this point.
00:05:06 Tim Gasper
They're launching the Falcons pretty often now, but I'm really excited about the Starship. That's the one I keep waiting for.
00:05:14 Ethan Aaron
Yeah.
00:05:14 Juan Sequeda
Well, I like to rant a lot about wine, that's one thing. But one that thing that maybe people don't know is James Bond movies. I am a hardcore James Bond fan and I can go in and say or talk about all the different actors and why-
00:05:31 Ethan Aaron
Which actor is the best?
00:05:32 Juan Sequeda
Sean Connery is the best and I think Daniel Craig is great, but the way they framed the movies that he did was to refocus it for a new generation and they really forgot about the hardcore James Bond fans. So him dying and him getting love, that's not what... I want to have the dumb, stupid, there's this one person trying to villain, taking over the world, that's the stuff I want to go do. Not him going off and falling, no. I've literally wanting to write essays around this.
00:06:02 Ethan Aaron
Yeah, when are you starting the podcast?
00:06:04 Juan Sequeda
I could start a podcast.
00:06:05 Ethan Aaron
The James Bond weekly cocktails and James Bond podcast.
00:06:09 Tim Gasper
There should be a special episode that we do that's just ranting about James Bond. That's all it is.
00:06:15 Juan Sequeda
We're six months in. We're about space or about flipping houses, whatever. Actually the next time we meet up at some conference we should actually do that. I mean, perfect.
00:06:24 Ethan Aaron
The other thing I can rant about, I haven't rant about this in a while, is energy efficiency. It's related to fixing up houses, but it's all the sustainability. How do you save money on that type of stuff? I went way too deep into that world before I got into data and I'm like a lead accredited professional, which is I've read way too much on how do you save money from energy efficiency and-
00:06:46 Tim Gasper
Oh, wow. You're lead accredited too. That's awesome. Wow.
00:06:49 Juan Sequeda
Lead accredited?
00:06:49 Tim Gasper
Yeah.
00:06:50 Ethan Aaron
The other one I can ran about is flying airline miles and loyalty inaudible. Oh, I can go-
00:06:57 Tim Gasper
Having a test.
00:06:58 Ethan Aaron
Give lectures around it.
00:06:59 Tim Gasper
Because you just speak alone, yeah.
00:07:00 Juan Sequeda
Let's get into this space because I think people are going to start tuning on inaudible.
00:07:04 Tim Gasper
People watching are like, "When they going to start the episode?"
00:07:07 Juan Sequeda
Here's the inaudible. Ethan, the honest, no bs thing here is that we haven't really prepared anything with Ethan, but I don't think we really need to go prepare because we can just go talk and talk. So what we discussed was you've posted a lot just in the last week, a lot of posts, I'm just going to name some stuff out here. Stop asking which tool should I use? The smartest people in the data can identify two things, levers that drive revenue and cost. One of the most impactful metrics to run an outbound sale organization is meetings book per week. You can increase the value of your data team by end of day today by finding a chart of low value and deleting it. Am I the only one who hates the term semantic layer? If your company doesn't have a data team, why not create one? Measure the impact of your data team and as data team, their work falls into four buckets. Anyways, this is just a handful of things that you've written last week. So honest, no bs, what do you want to rant about today, Ethan?
00:07:59 Ethan Aaron
Business value. I think we just go deep into business value. Why? How? All that type of stuff. I've had some phenomenal conversations this week alone with leaders in the data world and I think there's a lot of people on the journey towards business value and everyone's like, " Yeah, we do need to refocus on that." But then you also meet people that are experts at this. I post about it, but there are people in the weeds where they're teaching their teams how to read a P& L because it's important to the success of their data team or they are helping the company prioritize their strategy and forecast their business because it puts them into that right seat to then add value. So I can talk about this stuff for ages. Where do you want to start?
00:08:42 Juan Sequeda
You pick one. All right, actually, let me go pick. Let's start with what we discussing before. Obvious, the four quadrants, high value, high- low and then effort.
00:08:55 Tim Gasper
How to think about value, yeah.
00:08:57 Ethan Aaron
So we were talking about this before we went live. So I was catch up with Adley over at SimpleTire, I think it was last week. And he mentioned this and it's one of those things where it seemed so straightforward. I up until that moment was like, " Focus on high value stuff." That was my narrative, that's what I was posting about. And he was like, " Yes, focus on high value stuff but within the high value stuff, focus on the low effort initiatives." That to me was a small realization. It's like, " Yeah, that seems so basic." But it's a very big thing to think about when it's like, okay, cool. You as a data team hopefully went through a whole process of what are the high value initiatives for the company, what are all the initiatives? How do you prioritize them? Finance is a really good example. It's easy to get caught up trying to turn a... And this is another one of my posts, to trying to turn your accounting systems from Excel into a data warehouse and that is extremely high value. That is not low effort, that is extremely high effort. So I think the ability to both identify the high value things, that's step one. If you don't do that, you're going to waste time on low value stuff. And then also take a step back and be like, " Cool, we have 10 options here. These are all critically impactful to the business. They justify your job and your tech stack maybe depending on how much stuff you bought." And then within that, start with the easiest ones if not, you're a waste of resource. So it seems so simple, but a lot of people just jump in and they start writing SQL queries or they start writing some-
00:10:29 Tim Gasper
It seems obvious and I have a perception that if people could see the quadrant and could see their projects laid upon it, they would be, "Well, of course, I would focus on these things." Right? But I perceive, I'm curious if Ethan, you're you're thinking the same thing here, that where people get the most tripped up is they're not actually looking at their projects or potential projects in the backlog and asking is, this high effort or low effort? And is this high value or low value? The initial assessment is not happening really with teams. They're just running with whatever's urgent and whatever the boss says or whatever, right?
00:11:05 Ethan Aaron
Totally. And I think the reason behind a lot of it is it depends on the culture that you've set with your company and the expectations that you've set. Because a lot of people are like, " Yes, I want to work on the high value, low effort stuff." And then their boss comes to them is like, " Hey, I need you to work on this tiny... Change the color of this bar chart." And if they don't have the ability to say no because they didn't set those expectations, they don't have a higher value thing to work on, they're going to go change the color of the chart. When in reality setting those expectations up front of like, " No, no, no, we're going to say no to a lot of stuff. Be prepared. I'm going to tell you no so that we can focus on the top things." Saying no, I think, is the biggest problem that causes people to end up in the wrong quadrants. It's not that they want to be there, it's just they don't know how, especially early in their career or early in joining a company how to say no. And I think the biggest thing there is really just being able to set the expectations that if you don't say no, you're going to waste a ton of time.
00:12:08 Juan Sequeda
So we got a good question here. What are some of the criteria you use to define what might be easiest?
00:12:17 Ethan Aaron
I'm curious to get your take and I'll give my take.
00:12:23 Juan Sequeda
Okay. I mean-
00:12:23 Ethan Aaron
Am I allowed to ask you questions?
00:12:24 Juan Sequeda
Yeah, of course.
00:12:26 Tim Gasper
Yeah, bring the question.
00:12:29 Juan Sequeda
Obviously the answer is that it depends, right? So I'm thinking right now what you should go do is that literally go start putting in... Just throw the quadrant. You should actually have this in your office, right in the hallway, whatever, right? Have in the quadrant and actually put the stuff that people want to go do. And I think when it comes to cost, right? It's the amount of people that you have to go do this, the time that you estimate that it's going to go take and then all these things I think you should continue to ask why. Why? Why? And then if the why becomes a technical, you have to keep why? Why? Until you are able to go tie it down, tie it directly to some of the business units and you end up going, " Is this something that the COO needs? Is this something that the CFO needs? Something that the CEO needs?" Right? That's how we start figuring things out. And then the other thing is that you need to know what are your organization's top level objectives, strategic objectives, your OKRs, however you want to go do that. And you need to know that what is a direct path? To what are those objectives? And if first of all you don't even know what those objectives are, then that's a problem. Go figure that out., right? And if you find that as being indirect or so many paths to that, then there's missing there. So anyways, that's my quick rant right there.
00:13:45 Tim Gasper
Yeah, I think that's pretty good.
00:13:47 Ethan Aaron
What else would you add, Tim?
00:13:48 Tim Gasper
I would just say that, so my background is especially in product management and so I always think about when you're building product to what makes it easiest, right? And I think there's the work to be done. So there's that, right? Here's all the tasks I need to get done and how long I think it's going to be. But then you need to have your multipliers, right? It's like complexity, right? Are there dependencies on other teams? Are there dependencies on politics or buy- in or things like that? Well, that's make it harder. And then on top of that is uncertainty, right? Whenever there's uncertainty or error, that's going to create even more time and more complexity on the thing. So that to me is always feeding into the easy equation. What do you think Ethan?
00:14:32 Ethan Aaron
It's not a math equation is my answer. It depends on a lot of different factors. I think for most companies, the one thing that drives everything is time. It's how much time is this going to take your team to deliver? Unless you do something really wrong, just actual outflows of money to buy tools or technology should not be the biggest blocking item for this. Most technology is not nearly as expensive as your four data engineers. They're going to cost you more money. So it's really time. And I think the idea of time depends on what is the ask? How many charts? How many insights? What type of data? Et cetera. But it also depends on what you have in place today. So if you already have in place data vault and all this advanced stuff where you can just add that thing and it flows through, maybe you can do it in a day. If you don't, maybe that same task could take you six months. If you have talent, that's great at building dashboards but really bad at building pipelines, maybe building a pipeline will take you six months, but building the dashboard will take you a day. So I think a lot of it just comes down to lay them all out. To Juan's point, put them all on the board and hopefully you don't have a hundred things on the board. If you do, slim the board down, get to the top 10, get to the top 15. And then just sit there and have a real conversation with the people around the table and say, " Hey, we're tasked with this, estimate it based on weeks, estimate it based on days. How many weeks do we legitimately think it'll take us to deliver on this?" That's the only thing that really matters. And then if there's a massive price tag, I would note that also, most of the time it's time. So sit there and just be like, " This one will take two weeks, this will take one week, this will take whatever." That's the only thing I would use to prioritize most of this work, is how many days or weeks will it take you to deliver? And is there a really crazy price tag associated with any of it?
00:16:23 Tim Gasper
Yeah, price tag associated as well. What about on the value side? Right? When you see either in your own organization or the organizations you're working with, what do you see as, oh, they're doing a good job of assessing value?
00:16:41 Ethan Aaron
It's a really interesting concept. So I think before you jump into the value of a data team, you first need to understand what is the point of that data team. Because data teams, if they're trying to optimize for multiple things, they're going to end up in a very odd spot trying to prioritize, if I can save someone two hours a week and time verse I can provide 10% leverage to the CMOs marketing budget of 10. They're two fundamentally different things. So I tend to bucket things into four categories as data teams. So you should only have one priority as a data team. It's either analytics, building dashboards that provide leverage to executives. It is automation, just how many hours can you automate away every day or every week for the business. Product development, in that case you have a product you're selling, there's money, revenue, and cost associated with that product. Or risk mitigation. Risk mitigation is more difficult to quantify, but it's a very real way of creating value. So first you have to understand what's the point of your data team. Inside of LiveRamp when I was running the data team, I was actually running the analytics team. So my job was dashboards for leaders. There was another person whose job was automate tasks. For me, I had one goal, it was provide leverage to executives. The other person had one goal, automate away manual tasks. So know which one is your top priority and then within that most data teams are analytics teams, the job is dashboards. Again, it's one of those things that's not perfectly quantifiable, but the best way to do it in my opinion is go to the executives, find the CMO, find the CRO and be like, " What are the top two things for you? How do they impact the business?" Give me a two sentence anecdote. " Hey, this is currently taking two people to make this decision. It's taking four days a week." Great, quantify that. That's your anecdote alongside of it. And then go back to them and be like, " Are you sure? Would you vouch for that? Would you spend the budget to actually pay for that if we succeed?" Or if the CRO is like, " Yeah, if we can get these insights, we can get 10% more leads that lead quantifies." The executives want you to do work for them as long as it's super high value. They should be able to justify how it's going to impact the business and then ask them to confirm that that's the case, otherwise... So that's how I would think about is you're only talking 10 or 15 things, it's all relative priorities. Figure out who is willing to say, " That's the biggest impact to the business and I will vouch for it." And use that. It's not a mathematical equation.
00:19:05 Tim Gasper
I love that you're mentioning this vouching for it and really confirming it thing. And I'll tie it back to product management again too, which is people respond differently when you ask them like, " Oh, what do you want?" Right? And of course, they tell you everything they want, right? But if you ask them to start making trade- offs, well, would you pay for that? What if you only had 10 points and you could only spend those 10 points? How would you spend them? All of a sudden you get a very different answer.
00:19:30 Ethan Aaron
Yeah.
00:19:31 Juan Sequeda
So this ties into one of the things I saw last week, I was at the CDOIQ conference where it was all about... If I go back last year when I was at the CDO conference, it was all about business value, right? This year, the very specific business value was being a profit center. So you would see the data IQ top 10 CDOs were there and literally there's a panel with the top one, two, three CDOs, right? From Colgate, from Chevron and forget. And they're all about how can we generate more revenue? Where are we identifying we're leaving money at the table to make that we're not just saving money actually making you do that. And one of the approaches, somebody I forgot who it was, called out say if you say that you're saving amount of time and to whatever group go tell them and say, " Okay, I saved you x amount of time, what would you accomplish with that time?" Oh, we would accomplish y. All right, so let's go accomplish y. And that's what you should be pushing for. It's all about having these partnerships with folks. And I think another great thing, it's come up several times with folks that I talked to, but I want to hear it more, is that we really need to go partner with more of the finance. And somebody gave a great quote, which is, I cannot evaluate my own homework. So if I'm saying I'm doing this and I'm saying I'm bringing this value, you should actually have somebody else evaluating your homework. How well did you do? And that's partnering with finance right there.
00:21:12 Ethan Aaron
Yeah. Yeah. The other thing, I think those are a lot of tactics people can use to actually quantify value and it's either external revenue, internal accounting, and effectively someone actually saying, " Hey, that's worth x dollars to me if you do it." The antipattern I see, the way I see people fail at this is they turn it into a math equation and they assume that if they get some value for one executive, some value for another and some value for someone else, that if you add it all up, it matters. If you solve a third of a problem for three people, it's not a solution to any of them and none of them will vouch for any of it. They're not going to give you any budget, they're not going to say you succeeded. So it's one of those things where when you're thinking about value, think about the person that's going to vouch for the value and make sure that they actually will vouch for the value. Do not end up in a world of, yes, we have this spreadsheet and the spreadsheet shows all the value. Unless you can point to the people on the leadership team that are very, very bought into those numbers, it's a bad spot to be in as a data team. Yeah.
00:22:14 Juan Sequeda
That's incredibly insightful right there. I really, really like that. When you say you can save a third time to three different people, it's different to save you time for that one team directly.
00:22:25 Ethan Aaron
I made this mistake, so I learned this one myself. When I was at LiveRamp standing up the BI team, I went to all of the executives and I interviewed them all being like, " What are the top metrics, top priorities? What can I help you with with data?" And then I created a top 10 list and then I tried to do a global prioritization where it's like, " Hey, what does everyone think is the top priority for the business?" And yeah, sure, it's this, it's that, it's this and trying to come to a consensus. And then you end up walking away at the end of that where sure, everyone agreed that that's a top priority, but the CMO thought it was the second- highest priority and the CRO thought it was the third- highest priority. No one there was like, " That's the most urgent thing. I'm going to champion that and that will change my life so that I can then justify coming back and saying, I want to do the second one." Even if you don't pick the right top 10 for the business, I would pick the right top 10 where you can find a champion who will drop everything and be like, " Yes, solve this problem for me." And then find the low effort ones within that.
00:23:28 Tim Gasper
I love that. I think this is actually a good segue to another post that you made on LinkedIn where you said that the smartest people in data can identify two things. You said the levers that drive revenue and the levers that reduce costs. Can you talk a little bit more about why did you say that? Why is that so important when you think about assessing business value and focusing on the right things?
00:23:53 Ethan Aaron
Totally. So I was talking to someone earlier this week, to your question before the show, how do I come up with content ideas? I'm talking to people and it's not word for word what they say, but it just sparks ideas, I'm like, " Oh, that's really interesting." Another one where it's not that complicated. You should understand how your business works. The reason why that's so impactful is if you think about how dashboards can help or how automation can help, if you understand the levers that increase revenue and increase cost, it allows you to focus on the right stuff. If you're focused on a thing that doesn't increase revenue or reduce costs, how are you going to tie that back to any sort of value? And where this came from is, I was talking to someone earlier this week and they're part of the financial forecasting process. So as the data team, they are involved in the finance process of, okay, cool, what do we budget? What's our forecast? How's all this going to work? And a big thing if you haven't spent time in finance, I used to work in an investment bank. And if you have your model and all the forecasts and all this stuff, and then there's five levers in the model that change everything. Those are the things that matter. How many bookings can you get every month? How many things can you do? If you don't know what those are the things that actually drive revenue and reduce costs, you're going to be spending your time on the wrong stuff. So being able to identify the things that matter that you can actually increase or decrease. Then you can build your dashboards, then you can build automations on top of that. Then you can build products that increase those things. But I think a lot of people start building a dashboard for a person working on a task that's not material to the business. And you do that 10 times and now all your time is gone and you have nothing to show for it.
00:25:34 Juan Sequeda
And what brings up is the thing I always talk about, the data team needs to understand the business. I mean, understanding these levers is understanding how the business works and we can't have that separation anymore or say different ways. Those who actually understand the business, who work close to it, are the ones who really going to provide business value. Otherwise, you're going to go do a bunch of busy grunt work that it's low value work right there. So going back to the first thing is how do we know it's high value? How do you know? You understand the business, you understand what those levers are right there. So that's a critical thing. I love this.
00:26:19 Ethan Aaron
And they change based on the business. If you think about the model that matters to a e- commerce brand, cool, customer acquisition cost, mostly unpaid media combined with shipping and logistics. Those are the two things that probably matter to your e- commerce brand. If you think about SaaS products, its subscriptions and churn are two of the big ones. An interesting question here from Shane on LinkedIn is, should data teams be involved in financial projects? And I have strong opinions on both sides of this. One, personally, I don't recommend for most 99% of data teams or consultants to get involved in accounting or anything that's actually going to represent the books and records for your business. It's high value, but it's extremely high effort to get that stuff done. And most data teams don't have the level of rigor necessary to make sure that everything is perfect. You can't mess up your accounting. So for that one, I would say no. The other side of finance inside of organizations is planning. Okay, cool. Once every quarter you come up with a forecast of the future. It doesn't have to be perfect. It's inherently not perfect. It's inherently made up. And in that case, it's okay, here's the plan, here are the things we have to have in order for this to become a reality. That side of the house is where data teams should be very clearly involved. They should understand what those levers are and how they impact the business and how they flow through that model because that's the focus of the business. That's the strategy, that's the things that matter. So I think there's two sides of finance data, accounting and books and records, I would say as far away from as you can, unless that's your specialty. If you are a consultant focused on that, do it. Everyone else should not. It'll take too long. And then there's the strategy side of finance that I think data teams should definitely be tapped into. They should understand profit and loss.
00:28:13 Juan Sequeda
I think what it's really important is to start understanding what are the standard metrics in your industry? Right? You talk about customer acquisition costs, right? All these are metrics that are going to be the same across what industry now. The implementation, the semantics, we're going to get into semantic later. inaudible. I'm teeing this all up, you see? By the way, we did not plan this all up, but... No, yeah, inaudible. The metrics-
00:28:38 Ethan Aaron
The shirts are backwards or at least we think the shirts are backwards.
00:28:41 Tim Gasper
Oh, yeah.
00:28:41 Juan Sequeda
There's actually that, oh.
00:28:41 Ethan Aaron
We don't know if the shirts are backwards.
00:28:45 Tim Gasper
We hypothesize it's right on the video, but comment. Whoever's watching comment. Are our letters backwards or are they forwards? You let us know.
00:28:53 Ethan Aaron
Yeah. And then comment, do you think the shirts were printed backwards or do you think it's just the video?
00:28:59 Juan Sequeda
And everybody who's listening not seeing the video, they're like, " What the fuck is going on right now?" All right, we're moving back to my rant here. Understand the metrics within your industry, because these are the things that are just standard. This is just general knowledge of your industry you're going to go learn. And then from there you start getting into how is that implemented? How is that defined within your own company? And then you'll figure out there's changes even across departments and businesses and so forth, right? So I think this is one of the critical things.
00:29:30 Ethan Aaron
I have another thing I can rant on here.
00:29:31 Juan Sequeda
Go.
00:29:31 Ethan Aaron
Just now that we're on it. I think the idea of e- commerce brands all have similar metrics is pretty accurate. They all have a similar P& L, et cetera. But if you think about any e- commerce brand at any given point in time for the next quarter or the next year, it's really a question of the hundred different metrics you could use to run your e- commerce brand, for the quarter, the company has a strategic priority or maybe three. That's it. So sure, every e- commerce brand has a similar data model, similar entities that matter, et cetera. But you can't look at a hundred things to run your business every day. It's too much. Everyone will get lost. No one knows what the priority is. And the whole point of having a strategy as a company is every quarter to identify what are the top three to five levers to pull. So maybe this quarter is reduced inventory costs or speed of shipping stuff, maybe next quarter is paid media. And I think it's yes, every company has its own, here are all the metrics you could choose from. But the other point is every quarter you have to pick the top ones and that's the whole point is don't try and solve for the perfect data model for every e- commerce brand that's going to work everywhere. Solve for what is going to move the needle for your business this quarter. Make sure you're aligned with what your executive team has already agreed on around that and put it front and center for the company and delete everything else. Ignore it all because that's what matters. That's the only thing the company's focused on, so.
00:31:02 Tim Gasper
You got to focus and you got to be aligned with what the strategy is at that point. OKRs, a lot of companies do OKRs or something like it, MBOs or whatever it is, right? How much do you think the data team needs to pay attention to OKRs and feed into OKRs? Is it a lot? Is medium amount? What do you think?
00:31:21 Ethan Aaron
Have you read my SubSec article on this?
00:31:27 Juan Sequeda
No.
00:31:27 Tim Gasper
No, I haven't.
00:31:27 Ethan Aaron
So I wrote a whole SubSec article on this. So backstory on this. At LiveRamp, we had OKRs and like just most companies I looked at, I'd be like, " What is the point of this?" So I read John Doerr's book on OKRs. Overall, it makes a ton of sense. The whole idea is the company can define OKRs, teams and people can define OKRs. It's really just what's our goal and how are we going to track progress? Where it tends to fall flat is either doesn't align with the company strategy, so it's just process for the sake of process. Or, you do this and you sign up for it at the beginning of a quarter and then you have no way of tracking any of this stuff. Because even your company OKRs, it's like, cool, now we need the spreadsheets to track these things every day that goes by, every week. So to me, I think data teams should replace OKRs. And the way they should replace them is they should say, " Sure, we want to set objectives and we want to track progress." Instead of us defining OKRs and then trying to figure out how to measure it after the fact, I would argue that the company and the data team should say, " We're going to set our goals and we're going to track them. And the end of our OKR process or whatever you call it, is not a list of objectives and hypothetical things you can track, it is a dashboard. And if you can't get the data onto a dashboard, it's not an actual goal that you can track."
00:32:45 Tim Gasper
Wait. Wait. So Ethan, you're saying that we shouldn't spend the whole quarter trying to figure out how to measure the thing that we were supposed to be measuring and acting against inaudible?
00:32:53 Ethan Aaron
Do it during the process. So this is actually how I run Portable. So I'm the data person at Portable, I'm also the CEO. And every quarter we come up with, " Hey, here's the plan. What are the two, three things that we're focusing on as a business?" And then as we're going through that, I'm like, " Okay, cool, that's okay. But if we can't track that either in a spreadsheet or in a tool..." And when I say spreadsheet, I'm not saying we go look at the spreadsheet. I'm saying we track the inputs in a spreadsheet and then we expose them in the same dashboard, everything else gets exposed into... The eyeballs are still on a dashboard. But if we can't track an objective thing in a dashboard, it is not a goal. It can't be because how are we going to measure the success of it? So every quarter we go through planning and that planning does not lead to OKRs, that planning leads to dashboards with metrics that stay up to date either because we're automatically pulling them out of systems or because we are updating them in spreadsheet and then pulling that directly into the dashboard. So I think if you run a data team and every eyeball isn't on dashboards for the OKR process, you lost, or at least you're losing for now. Someone is using a spreadsheet to run the business and your first goal should be build a dashboard that shows all the information in that spreadsheet, even if you have to just go update a spreadsheet every day and keep your looker dashboard up to date or your retool dashboard up to date. At least now their eyeballs are on your dashboard and you can start to automate away the pieces of it. So I have very strong opinions on this. I think OKRs got us halfway there. They just didn't actually figure out a way of measuring progress. That's the fundamental flaw with the process. Yeah.
00:34:34 Tim Gasper
I love this. This is making my business heart pump.
00:34:40 Juan Sequeda
I do want to say that I think part of the culture of organization is that you want to get people excited and getting people excited about, oh, we're here to increase this 80% like that, right? Then people are here for a job. And I mean, this is also a culture way of how you set up your business. But I think part of the separate O and the KR, the objective and the key result. The key result is something that is this measurable thing you want to go do. But I think the objective is okay, if it is something inspirational. We are trying to go do it because you want to really rally the company saying, " This is what we're trying to accomplish and let's get really excited about it." Right? I don't wake up excited on Monday morning to increase or reduces 20%, no. But I do wake up mean I want to make sure that all our customers are... They say they cannot live without us, right?
00:35:36 Ethan Aaron
Yeah. But you also want to be able to see the progress over the course of the quarter-
00:35:39 Juan Sequeda
No, no, no. And I think that's where you have the KRs or the ones that are metrics, and this definitely is things that need to be measured. But those are connected to the objectives that can be a little bit more squishy feeling, but just because that's what drives the culture but drives people excitement around that.
00:35:54 Ethan Aaron
Totally. Yeah. Those are the benefits of the company. The benefits of the company are, everyone can look at dashboards, half of it stays automated, the other half can be updated in the spreadsheet, doesn't matter. The benefits of the data team are twofold of this approach. Number one, you get eyeballs on your dashboards and now you can actually add other stuff in there that's valuable to the business and you have the eyeballs. The biggest problem with most data teams is they don't have the eyeballs of the leadership team or the company. So this is an easy way of getting that. The second thing it does is if you focus your efforts on the KPIs for the business and actually becoming that source of truth for the business, it forces you to delete all the other stuff. And every quarter you don't keep adding more KPIs. If you know anything about OKRs, it's not like last quarter you had three goals, this quarter you add three, and next quarter you add three, and then you have nine goals. Then the company is dysfunctional. What happens is every quarter you pick the top three and you ignore everything else. That is the whole point. And then you track them. And what I think this process forces you to do as a data team is every quarter delete all the stuff that's useless. And now what's happening is your data team doesn't have to keep increasing cloud spend, increasing headcount because you're trying to maintain more and more and more dashboards and models and whatever. You have a scarce resource. It's x number of goals for the business that go on a dashboard or on whatever dashboards. And it gives you the opportunity to say, " I'm going to delete that because it's not a goal for us as a business." So that's the other benefit to data teams.
00:37:31 Juan Sequeda
Goes back to the other post of, you can increase the value of a data team today by finding a chart or dashboard of low value and delete it. So that low value is actually probably the metric that we're tracking before that we don't care inaudible right now.
00:37:47 Ethan Aaron
I did it this week. So in that post I talked about it. Last quarter or the quarter before, we cared a lot about blog posts. Me posting stuff and some copywriters that we had posting stuff on our site. So it shows up for SEO purposes, search. It was great. We were going strong. We know the value of doing that. This quarter, that's not our goal. It's not one of our top three priorities as a business. I looked at the dashboard one day, I was presenting to someone being like, " Look, this is what we do for our KPIs." And I was like, " Wait a second, this isn't one of our goals anymore. This was a goal, showing how many blog posts by month, are we shipping?" And I was like, " This isn't goal anymore. This is a distraction to the company. This is a SQL query and a data flow that I'm maintaining as the data person that is irrelevant right now." So I went in, removed it out of the dashboard and deleted the charts, deleted the SQL queries, turned off my data flow, moving my Contentful data into my data warehouse and it's gone. And now it's cleaner. Now I can just focus on the stuff that actually matters today instead of continuing to accumulate connectors, queries, compute all this stuff indefinitely because most data teams do that. It's just every month that goes by, everything keeps going up. But why? The company priorities don't keep compounding. That's not how the stuff works. Yeah.
00:39:06 Tim Gasper
Right. I think that's pretty sage advice and I think the vast majority of organizations don't delete enough of their data crap, right? We have one customer that has 450,000 Tableau dashboards and you got to think that they probably don't need all 450, 000 of those Tableau dashboards, right? And they just never deleted any, right? They just kept on accumulating. So I know we want to move on to a couple other topics real quick before we hit our lightning round. But I just want to do one last question or comment on this whole OKR topic, is that I feel like a lot of times with OKRs, they're not set with the data people in the room, right? And so this whole idea of even what can we dashboard isn't really even part of the conversation because the person who knows isn't there. Just curious about your thoughts on that. Do you got to make sure that the data person's in the room?
00:40:04 Ethan Aaron
If I was the data person at that company or the executive at the company and I wanted to run the business with data, actually be a data- driven company, yes, the data person's in the room. If you're not in the room today, you have to fight and justify why you need to be in that room. You have to build the dashboard that is better than whatever they have today in their spreadsheets. It's not that difficult. If they have four spreadsheets and you can ETL the four Google sheets into a dashboard. Now they only have to look at one thing and it's magic to them. Even if you just go manual after, it's not that difficult. But you need to add enough value to deserve to be in the room. And then it's a question of how do you run the business that way?
00:40:42 Tim Gasper
That's fair.
00:40:43 Ethan Aaron
But it comes down to the culture and it's changing the culture. If the leadership team's like, " No, I'm going to use my Excel spreadsheet or my Word document or my actual physical whiteboard or blackboard." It's more difficult. But if you want to be data- driven, the data person should be accountable for building those dashboards. The leadership team should hold them accountable and be like, " Everything we talk about, you need to give me confidence that we can actually track this either manually or automated." And then the data person has a right to be at the table. If not, you're going to do low value stuff.
00:41:16 Tim Gasper
No. Good advice.
00:41:18 Juan Sequeda
Ethan, why do you hate the term semantic layer?
00:41:22 Ethan Aaron
I still don't know what it means, is one answer. And then the other side of it it's what I think it means is a business representation of concept. This is our customer, this is our location, this is our whatever. And it doesn't sound like that's what it's trying to say. I wish it was a business object representation or something where it's just focused on things that matter to the business. I'd rather have that full sentence than semantic layer, which just makes it seem like a super technical term that someone's going to go and write code to solve. When in reality it's no, you have to sit down and actually understand the entities and how they relate to each other inside your business. And then work backwards in terms of what you actually want to pull and how it relates to each other and forwards in terms of the dashboards. I think semantic layer has tried to turn a concept that's really powerful, focus on the business concepts and not the data terms and turn it into a data term, which is my rant on that topic.
00:42:23 Juan Sequeda
I have to say, I am now agreeing with you. My background's all about semantics, so I'm like, " This is my work." But really what you said is, yes, exactly, semantic layer is exactly things that matter for the business. It's a business representation of our concepts, right? Now that's a mouthful, but maybe we should be... And I like your point that you're turning this thing into not technical terminology, which is confusing people, right? So.
00:42:52 Ethan Aaron
My guess is 5% of data people actually have a good understanding of what a semantic layer is, if that. My guess is 30% have a general conceptual understanding. The other 70% have no idea what it means. And if you go talk to a CMO or a CEO or a CFO of a business and you're like, " Oh, the semantic layer is the answer here." That is never something you should say to someone in the business. But if you go to them and you're like, " Yeah, as a data team, we represent business concepts first and foremost, and then we'll deliver the technology behind the scenes to get there." I'd rather do that. It is not a term you should ever use talking to a business person. And if that's the case, I probably-
00:43:35 Juan Sequeda
So we-
00:43:36 Ethan Aaron
Should probably come up with a different term.
00:43:38 Juan Sequeda
Last week I organized our honest, no bs dinners and we were inaudible. It was after the CDO conference and this came up as something that's not working inaudible. Lack of semantics is a problem. But you can't sell semantics to the CFO, we need to have a better story. Call it shared language or something else, right? This is exactly the point you're making. All right. We're in agreement.
00:44:02 Ethan Aaron
Yeah.
00:44:03 Juan Sequeda
I got another question for you. So which tool should I use?
00:44:10 Ethan Aaron
So what's the business problem?
00:44:14 Juan Sequeda
Yeah. So what is your rant for folks who are asking which tools should I use? Should I use Snowflake or Databricks, right? Should I use inaudible-
00:44:26 Tim Gasper
Kafka or Kinesis?
00:44:30 Ethan Aaron
It entirely hinges on what business problem you're trying to solve. So the actual reason I started posting on LinkedIn, my first post ever that I got any sort of engagement was me posting and being like, " Hey everyone, I'm pulling together a modern data stack at Portable. Here are all the tools I'm going to use. ELT, data catalog, modern data warehouse data, behavioral data." I listed 30 categories and I was like, " My goal is to figure out how many customers we have at the end of all of this." And everyone's comments were like, " Oh, which tool did you use? Which tool did you use? Which tool did you use?" I just want to know how many customers I have. Where do we track that information? For us, it's mostly in Stripe. So you don't need a data warehouse, you don't need a dashboard tool, you don't need an ELT tool, you don't need any of this stuff if you want to know how many customers you have. Just log into Stripe, if that's where you store your customer data. So it really hinges on building the minimum viable pipeline or minimum viable solution for whatever you actually need to accomplish and you need to accomplish does... Compound is the wrong word, but it evolves over time. So if you have to accomplish 10 things, you might need a data warehouse. But again, don't keep compounding these things indefinitely because then you'll need 50 different tools, a 50 person tech stack or a 50 person data team, and you'll just keep drowning in work. So it depends on what you need to accomplish at the given point in time. Every quarter, if your KPIs change for your business and the dashboards you use to run your business change, you should be reevaluating whether or not you use tools. I'm not pulling data from Contentful anymore into my data warehouse to power a dashboard that shows blog posts written by month. I went back into Portable and I turned it off because why is that data going into a data warehouse if we're not going to use it for anything right now and I'd rather just turn it... So that's my take on it, is it first have a very, very clear understanding of what it is you want to do and then find the fastest and cheapest way of finding the solution to it.
00:46:33 Tim Gasper
Question to that is, do you feel that this idea, the modern data stack is helpful or does it get people thinking too much about there are boxes and arrows and I'm supposed to fill these boxes?
00:46:50 Ethan Aaron
So when was this? Last Snowflake summit I posted the modern data stack is dead. I've since posted the same thing three times. I think the word modern is what throws a lot of people off here. The term modern data stack makes it seem like a thing. When in reality there's no better term for it because it's really just a data stack. And the idea of a modern data stack made it seem like this flashy thing that everyone needs go buy but if you just use the word data stack, it's like you could have something that resembles a data stack. Maybe it's a spreadsheet, maybe it's a visualization tool plus a database, but I think the problem with the culture around adding modern to the front of it is it makes it the shiny thing in a window where you feel like you have to go buy all this stuff and work backwards from there. What we've seen over the last three years is that's what was happening. Teams are being staffed up to buy this stuff without ever asking why. What are we trying to do for the business? What is the dashboard we care about? Because it's very possible that dashboard you care about, you could just put data in a spreadsheet and put a chart on the next tab and it works. You might be good with that, and if that's the case, delete everything else. You don't need the pipelines, you don't need any of it. So I think people working backwards from the tool. Sure, having buckets and categories make sense after you figured out what you need. It's a great way to be like, " Oh, I need this. I need to get this source into this warehouse and be able to identify it here and govern it here and visualize it." It's cool. I'll pick one of these tools, one of these tools, one of these tools and one of these in service of this goal. Most people work the other way around where they're like, " I need this tool and this tool and this tool and this tool and this tool, and then I'll be able to answer every business question." But they never make it there. They got stuck in the semantic layer and never actually talked to the business people.
00:48:37 Tim Gasper
I think that's the money comment right there. I got to build the whole thing and then I've got the arc and then I can stick all the animals on the arc, right? You just never know. Start with business. Start with the business problem. The business value.
00:48:50 Ethan Aaron
Yeah.
00:48:50 Juan Sequeda
So we can argue that all these architecture diagrams are just the root of all problems. They're like, " Oh, I need that."
00:48:59 Ethan Aaron
The incentives were pretty broken over the last few years. It was like, yes, everyone wanted it. The other thing that was happening was people that were hiring for roles or promoting people were promoting and hiring based on people's experience with however many of these tools they could get experience with. It made sense for people to say, " Look, I'm an expert in these 15 tools. I've implemented them all. I know how they all fit together." Which it is valuable to understand how to integrate tools. It's extremely valuable in the data world. Being able to do it and then being able to say, " I don't actually need them." I would rather pay someone a lot of money to do the second half of that than a lot of money to the first half. Being able to say, " No, I do not need this." Could save your company a hundred thousand dollars tomorrow by just actually being like, " I know I can do it. I know I can put it on my resume, but I'm going to delete it because I'm going to save a hundred thousand dollars and I'm going to go ask for a $20, 000 raise because I'm actually fiscally responsible." Very little of that was happening in the last few years.
00:50:00 Juan Sequeda
Lack of being physically responsible. Yeah.
00:50:03 Tim Gasper
That's some inaudible advice right there.
00:50:05 Juan Sequeda
All right. Well, look, we can keep ranting about this stuff, but we got to start wrapping up. We got a couple of things in our next segments. So we're going to go off to our lightning round questions.
00:50:15 Ethan Aaron
Let's do it.
00:50:16 Tim Gasper
Oh wait, before that, are we going to do an AI minute?
00:50:18 Juan Sequeda
Oh, yes. I forgot it.
00:50:20 Tim Gasper
Okay.
00:50:20 Ethan Aaron
Is this where you replace me with an AI avatar?
00:50:24 Juan Sequeda
That actually before-
00:50:25 Tim Gasper
It's a good idea.
00:50:26 Juan Sequeda
All right, let's go do the AI minute. So you got one minute to rant about anything you want about AI. One minute. Go.
00:50:38 Ethan Aaron
Right now, I think the easiest place you can see AI impacting the world is content creation. You can see it on blogs, you can see it in LinkedIn now has the create stuff with AI. And we saw this when we were writing content. So we initially were humans writing content. Then we were also humans, but we were outsourcing content creation and paying humans to do it. And then we started realizing those humans were just using ChatGPT under the hood or AI under the hood. And then we saw a drop- off in quality and then we realized that if they're not going to write human grade quality content, then why not just use chatGPT directly ourselves. And now what's happened is our content quality converged 80%, maybe 60% of as good as it was before. It's a garbage is the answer. And I think we're going to see that everywhere. I think we're going to see it on LinkedIn. I think we're going to see it in blog posts and Google and all that type of stuff. And I think it's bad. I think authentic, human, no bs information is going to have a premium as long as you know that's where it's coming from.
00:51:42 Tim Gasper
I worry about that too.
00:51:42 Juan Sequeda
You'll not be replaced by AI, Ethan. We won't let that happen, so. All right. Lightning round questions. Four questions, yes or no? Quick context if needed. I'll go off. Number one, is LinkedIn a better social network for data people than Twitter?
00:51:57 Ethan Aaron
Yes.
00:52:01 Juan Sequeda
All right. Tim?
00:52:02 Tim Gasper
All right. Should the individual data team members, so the people on the data team, be weighing in on the business value of tasks or is that really more for team leaders and managers and leaders to do?
00:52:17 Ethan Aaron
Oh, it's a yes or no question.
00:52:19 Juan Sequeda
No, you can give a little bit context.
00:52:22 Tim Gasper
You can give context there.
00:52:23 Ethan Aaron
I would say no, not directly. Unless you're an expert in finance or an expert in sales, or most of the time you're not. But it's your job to get the information. People aren't just going to hand you, " Oh, this is how valuable it is." Just a product manager has to get the information out of customers, how much would you pay for this? The PM's not the one saying, " Here's how much people would pay." They're the one saying, " I have 10 people that have said that they would spend this much money. Three of them have contracts, two of them would sign a contract. Five of them are iffy." It's your job to do that as the data team. It's not your job to be the one saying, " Here's the number." But you have to get that information, otherwise you're not doing your job.
00:52:59 Tim Gasper
Fair. Yeah.
00:53:00 Juan Sequeda
Third question, is the semantic layer valuable?
00:53:07 Ethan Aaron
No.
00:53:10 Juan Sequeda
So after you're little rant, I thought we were going to say yes.
00:53:13 Ethan Aaron
A business object relationship, whatever we want to call it layer, is valuable.
00:53:17 Juan Sequeda
Okay. It's a business object... Hold on.
00:53:21 Ethan Aaron
We need a better term for it, but-
00:53:23 Juan Sequeda
Okay. No, no, I got the word here. Is the things that matter for the business, the business representation of our concepts, is that valuable?
00:53:30 Ethan Aaron
Yes. Yes.
00:53:31 Juan Sequeda
Okay. You definitely do not like that term semantic layer?
00:53:35 Ethan Aaron
No.
00:53:35 Juan Sequeda
Okay.
00:53:35 Tim Gasper
We need to come up with some kind of acronym, the business objects concept thingy.
00:53:42 Juan Sequeda
Oh, this is-
00:53:43 Tim Gasper
inaudible or something.
00:53:43 Juan Sequeda
I mean, it's SAP business objects, was that right? It's kind of pop up, right?
00:53:47 Ethan Aaron
Right.
00:53:47 Tim Gasper
Every time you say business objects, I'm like, " Is SAP going to sue us?" I don't know.
00:53:50 Juan Sequeda
Final question, Tim.
00:53:52 Tim Gasper
Fourth question. Is the ability of a data team to be a profit center, which we've been hearing a lot more as a theme, is that ability for them to be that profit center overstated?
00:54:08 Ethan Aaron
No. I think you either drive actual profit or you use internal accounting to find profit and/ or ROI on these things. And if you can get someone to say this is worth a million dollars and you spend$ 900, 000 in resources, that's profit. If you spend a million dollars in resource and get$ 900,000 in value for the business, not profit.
00:54:32 Tim Gasper
No, that's fair.
00:54:34 Juan Sequeda
Perfect. All right. Takeaway time. Tim, kick us off.
00:54:42 Tim Gasper
All right. I had to take one more note there because you said a really good thing. All right, takeaways. So we said that today was going to be rant day and we had a really good rant discussion. I thought it was good. And we got to call out to a lot of really great LinkedIn posts. For all of you out there, if you're not following Ethan on LinkedIn, please make sure that you're following him. We really focus today on business value. That was the thing that you really wanted to center around and that you started off by saying that a lot of folks are on the journey to business value and it's great that I think it's becoming more of a conversation now. But then you meet folks that are really truly experts in business value data, people that are learning all about the P& L. There is next level stuff that we can be getting here or getting to here as an industry, as data professionals. We're not there yet. We have more work to do. And when the first thing that we talked about was your post around as a data team, that work falls into these four buckets. And you really need to start thinking in terms of this quadrant. If as a data team, as a data leader, you want to be achieving value and you want to get there efficiently and quickly. And the four quadrants are one of axis is value, right? And the other axis is effort. And you want to live in the high value, low effort quadrant, that's where you want to live. A lot of people, they'll jump into SQL, they'll get lost. Or your boss tells you, " Hey, can you make the bar chart be red? Can you do a double stack? Can you move that dashboard from Tableau over in a looker?" These are these time waster type things and you got to be focused on business value. You also need to have a culture where it's okay to say no to things that aren't valuable. And I think that's a deep question, that's one that requires a lot of thinking both from individuals and from leaders about how do we foster that kind of culture where it is centered around business value and not just what the highest paid person in the room says, the hippo, right?
00:56:41 Juan Sequeda
Yeah.
00:56:42 Tim Gasper
How do you know what's easiest? Well, it depends on time, depends on complexity, it depends on uncertainty. If you have a hundred things on the board, try to get it down to 10 to 15 and try to figure out, hey, how many days is it going to be? How many weeks is it going to be? Right? You don't have to get too complicated here and don't become too obsessed with the math, right? You just need to be generally in the right direction. And how do you know what's most valuable? You centered it around what's the point of the data team? Right? Is that data team centered around analytics? Are they centered around automation? Are they centered around new product creation and innovation? Are they centered around risk mitigation? And that's going to drive a lot of what does value mean and how you drive it as an organism within this larger community. And so I think that's really important. And then before I pass it to you, Juan, we also talked about smartest people in data can identify two things, the levers that drive revenue and the levers that drive cost. If you understand the levers, you're focusing on the right stuff. If you don't know what they are, you're going to spend time on the wrong things. So you got to really try to figure that out. Juan, what about you?
00:57:46 Juan Sequeda
Well, I think part of that of the levers is knowing the metrics, and that tied the whole question of metrics with OKRs, right? So first of all, you need to understand the metrics of your industry, but also you can't just keep track of all a hundred metrics to run your business, right? This is why companies have strategies in every quarter they pick the top ones, right? And what is going to move the business forward this quarter? So that's when we ended up coming. Interesting, getting to this conversation of OKRs, right? The data teams should replace OKRs, what you're saying. If you can't create the dashboard at the end the old care planning meeting, then you did it wrong. So everyone's eyeballs need to be on that dashboard at all times. And another interesting point is that one of your posts was find that dashboard that's providing low value, then delete it. Because that low value dashboard really you're tracking a metric that was a past thing, which is actually a distraction. Hey, it be costing you money to keep that running. So I think that's one of the stuff. And then it really boils down to the culture if you're truly being data- driven or not. And that's how you know if data teams are in the room, they have a seat at the table, right? If they're truly a data- driven organization, then the data teams are in the room. If they're not in the room, then it's a cultural issue and they really need to show and fight for it and really show the value out of that. And then we had the discussion about our semantic layer, which we now know that the term, the syntax semantic you do not like, but the semantics behind the syntax of the word semantics, you do. You agree? Oh, this is great.
00:59:15 Tim Gasper
This is deep.
00:59:15 Juan Sequeda
So it's the business representation of our concepts, the things that matter for organization, right? Then using the term semantic layer has really turned it into a data term and very few data people actually know what that is, right? I think that's something we can't go sell easily, right? So you can't go sell your CMO like, " Oh, we're going to buy this semantic layer thing and all problems will be solved." No freaking way. And I think there was a comment over here that Jane said we should bring back business objects. I like that. Maybe it's way. And then finally, for people who ask, what tool should I use? The question should be, what is the problem you're solving? What is the business problem? And really build that minimal viable solution that's going to help you accomplish that business problem, right? And so figure out what you need first, then ask for the tools and service of that goal. Usually it's the other way around, and that's the big problem. And I think one of the issues too there is that a lot of people in the last couple of years, right? They've gotten all so focused on all these tools that they would actually get promoted because they knew experience about these tools and so forth, right? Or even hired around that. And that actually has been probably part of the problem. I think that's something that would change. How did we do? Anything we missed?
01:00:24 Ethan Aaron
I think you missed one thing.
01:00:26 Juan Sequeda
Oh, shoot.
01:00:28 Ethan Aaron
Are the shirts printed backwards?
01:00:31 Juan Sequeda
I think somebody said we are showing up-
01:00:35 Ethan Aaron
Yeah, we're backwards on the screen. The real question that no one will ever know the answer to is, are the shirts printed backwards?
01:00:40 Juan Sequeda
Well, next time you find us in the real world, I have many of these shirts and anytime around, I'm always wearing it.
01:00:49 Tim Gasper
Find us at a conference and we'll give you a backwards shirt and hope you enjoy it.
01:00:55 Juan Sequeda
All right. So to wrap up quickly, three questions. What's your advice about data or about life? Whatever. Second, who should be invite next? And then third, what resources do you follow besides you?
01:01:11 Ethan Aaron
What would my advice be? What's the business problem you're trying to solve? Everything should work backwards from that. Who should you invite next? And I can make interest to some of them. Find the heads of data that are ruthlessly focused on business value and just learn from them. They're experts. They're a lot smarter on this than I am. And where do I learn? I spend a lot of time on LinkedIn. I don't really spend any time on Twitter or whatever it's called nowadays. Slack channels, I don't really spend tons of time on. And meetings. I will go on LinkedIn, post stuff, start conversations in the comments or in threads, and then I'll just get meetings with people. 30 minutes, no point whatsoever to these meetings just to learn, to hear about people's backgrounds, hear about how they approach the world. I learned so much from that. That's where most of the stuff I talk about comes from, is just picking the brains of people that are doing this stuff every single day.
01:02:17 Juan Sequeda
Love it. Thank you so much, Ethan. Just quick reminder, next week we have Wendy Turner- Williams, who's actually the former CDO of Tableau and the former VP of data strategy at Salesforce, right? She's actually about to launch something that she's going to announce next week here in the podcast. So you got to listen to that. And with that, thank you Ethan. Super, super grateful that you were here on the show and we just had this awesome rant session.
01:02:41 Ethan Aaron
Thank you. It was a pleasure. Had a blast.
01:02:43 Juan Sequeda
Cheer everyone.
01:02:43 Tim Gasper
Happy ranting with you. Cheers.
01:02:44 Ethan Aaron
Cheers.
Special guests
