NEW Tool:

Use generative AI to learn more about

Product Launch: has officially leveled up its integration with Snowflake’s new data quality capabilities

February 29th:

Ensuring Highly Distributed Data is Available to All with W. R. Berkley

Upcoming Digital Event

Learn how WR Berkley & Singlestone Consulting supported this distributed model with modern data practices and a data catalog built on a knowledge graph.

View all webinars

Data That Doesn't Drive Results Is Useless With Alexa Westlake

Clock Icon 62 minutes

About this episode

If data isn't driving results, it's more or less useless. In this episode, Alexa Westlake will share from her experience on how to define and scale a council, create a culture of joint ownership of metrics, and discuss the relationship between outputs and outcomes.

00:00:05 Tim Gasper
Welcome. It's time once again. It's time for Catalog& Cocktails, your honest, no BS, non- salesy conversation about enterprise data management presented by Data. World. I'm Tim Gasper, longtime customer guy, product guy at, joined by Juan.

00:00:19 Juan Sequeda
Hey, Tim. I'm Juan Sequeda, principal scientist at, and as always, it's Wednesday, middle of the week, end of the day. Time to take that break, time to talk about data, and today I'm super excited. We have Alexa Westlake, who's a senior data analyst at Okta. We're going to get diving into real hands- on, rolling up our sleeves about data teams and just how we're driving results. Alexa, how are you doing today?

00:00:45 Alexa Westlake
I'm great. I'm doing great. How are you?

00:00:48 Juan Sequeda
Well, we're super excited for our conversation. And we were just chatting before this, and you also just came back from some really awesome places that we're interested hearing, but with that, let's kick it off, tell and toast. What are we drinking and what are we toasting for today?

00:01:01 Alexa Westlake
Yeah, totally. So I've brought something from my travels. This is Mirto Rocky. It is a Croatian myrtle liquor. Rocky is distilled with local fruit, so it uses whatever fruit in the Balkans is most present in that region, so in Dalmatia and in the Dalmatian Coast, where I just was in Croatia, they make it from grapes, so just pluck grapes. I got this one on a wine tour while I was out of office for my partner's mother's 60th birthday. And then since I just got back, I also have some Gatorade, because I'm still getting those electrolytes in, a little bit jet- lagged, but I am toasting to a corporate culture that treats employees like human beings and not human doings. I feel very grateful to take the time to celebrate my family and life's important milestones. And while our work at Okta is held to high standards, people do their best work when they're at their best, and taking time off to focus on you is a key part of that, so cheers.

00:02:04 Tim Gasper
That's awesome.

00:02:05 Juan Sequeda
That's an excellent, cheers to company cultures that really drive that, as employees, they treat you as you got to be your best and you go do you because you being your best, that's how we're all going to be great as a company.

00:02:19 Alexa Westlake

00:02:21 Juan Sequeda
Cheers to that. Tim, what are you up to? What are you drinking?

00:02:24 Tim Gasper
I'm drinking something a little bit weird today. Not as interesting as you, Alexa, which I'm fascinated. I took some notes,'cause I actually want to see if I could try that at some point, but I'm actually drinking something that is non- alcoholic, technically, asterisk. It's called Ghia and it's got a really strong bitter and also a ginger note. So if you want to try a non- alcoholic cocktail, it's G- H- I- A. I recommend you give that a try, but there's an asterisk,'cause I'm cheating. I actually think it tastes better with a little bit of an orange note. So I added a little bit of orange liqueur to it, Matilda, which is really tasty. So it is not a mocktail, it is a low cocktail that I'm drinking today.

00:03:12 Juan Sequeda
That's Ghia? I've never heard of this. This is really cool.

00:03:15 Tim Gasper
Yeah, it's really tasty. I've been trying more of these non- alcoholic cocktail type mixes, and I find that if something is pretending to be whiskey, it tastes pretty bad, but if it's an alternative thing where it's like herbs or botanicals or things, it's actually quite good. So for those thinking of trying out mocktail type things, that's my advice to you.

00:03:35 Juan Sequeda
Well, I'm actually having a little something, not similar, but I mean, has all the, it's gin, it's just classic Tanqueray, which I enjoy. I didn't have any tonic, so it's sparkling water, but I have some cucumbers in it. I'm in this big cucumber mode right now. I love cucumbers and then I'm ordering a lot of cucumbers, I'm buying a lot of cucumbers, and my wife is like, " What are we going to do with all these cucumbers?" I happily eat cucumbers like this.

00:04:03 Tim Gasper
You're like, " I got an idea."

00:04:04 Juan Sequeda
Yeah. Anyway, I got my veggies in my drink, so cheers. Cheers.

00:04:07 Tim Gasper

00:04:08 Alexa Westlake

00:04:10 Juan Sequeda
All right, so because the topic today is driving results, so I'm using ChatGPT to help us come up with funny questions, warmup questions. So question today is, driving results is the theme today, if driving results were an actual road trip, what kind of snacks would you pack for the journey to success?

00:04:27 Alexa Westlake
Oh, I love that. So I'd say, number one, I would make sure to ask whoever's in the car with me what their top snacks are, because the journey to success is most important when you listen to the people that you're driving with. I'd also say bring a lot of water, because you're likely going to have to take a lot of detours and you want to make sure that you're very well hydrated and prepared for all of the detours and unexpected trips that you're going to eventually have to take on the way.

00:05:10 Juan Sequeda
That's just an awesome mic drop moment that I was not expecting to come, especially'cause you started off with asking the other people, because I was thinking about this, like, "What do I want?" And I think that's already the problem right there. I was thinking about me and not like, " Hey, by the way, if I'm driving to success, I'm probably not going by myself," so.

00:05:28 Alexa Westlake

00:05:28 Tim Gasper
That is true.

00:05:28 Alexa Westlake
Yeah, the road to success is not something you can often take alone.

00:05:33 Tim Gasper
Yeah. That was a very polished answer, Alexa, to a very sudden and unexpected question, so good job there.

00:05:40 Juan Sequeda
By the way, I told this question to Alexa one minute before it went live, so she's on her toes. Anything you want to add, Tim?

00:05:49 Tim Gasper
I think you stated it very well, Alexa. The only thing I would mention is I think bring some Twizzlers,'cause I think that life is going to send twists, so have some twisty snacks as well.

00:05:59 Alexa Westlake
Love it.

00:06:01 Juan Sequeda
All right, well, so let's kick this off. So, honest, no BS. It seems that most data teams are focused on outputs, when teams should really be focused on outcomes. That's my impression. I've seen things around. Why is that? I mean, first of all, do you agree or is that what you're seeing and why so much focus on outputs instead of outcomes?

00:06:21 Alexa Westlake
Yeah, I would absolutely agree. I'd say that we often, as data professionals, focus on what we can measure and not necessarily what we should measure. So it's easy to understand what the milestones and project deliverables are. So we assume that if we're ticking boxes, then we're doing a good job. And I'd say that it's not that we have to measure outcomes over outputs. I think that the relationship between the two of them is very important. You can't achieve outcomes without those project milestones and often outputs can be leading indicators of outcomes. That being said, if we're not understanding what the outcome of all of the milestones and outputs are, then what's the point? We need to make sure that, throughout the entire life cycle, even when we're done with projects, we understand what the measurable business outcomes that we see from the lifetime value of that project are, whether that is something like an ARR measurement or a case volume measurement, but actually understanding what the business is trying to achieve and how we can set up measurement to understand that those outcomes are tracked over time.

00:07:41 Tim Gasper
Yeah. No, I think that's very well stated. Alexa, you're on a data team and you have to work with a lot of different people to try to drive, not just output, but outcomes, and rally a bunch of other people to do the same. Can you talk a little bit about the kind of work that you're doing and how are you all at Okta and yourself trying to drive this outcome orientation, not this output orientation?

00:08:11 Alexa Westlake
Yeah, so our company vision is to free everyone to safely use any technology, and just like Okta does that, our data teams, North Star, is to empower the enterprise with trusted, timely, and transparent data. So our data strategy focuses on making Okta's products better, developing relevant features, enhancing our security posture, and using it to improve our customer experience. So it's impressed, it's especially crucial to understand our customer. So relevant data gives us the ability to apply those predictive and targeted insights throughout the customer journey, which can help influence our go- to- market strategies, our product roadmap, enable our customer success teams to have knowledgeable and proactive conversations with our customers. So we're seeing tremendous innovation throughout the entire data lifecycle. And what's crucial is to ultimately generate better, deeper insights and making sure that they're deployed to the right part of the business. So we create a closed loop ecosystem, and what I mean by this, to close the loop, is that you can either connect an insight to a person so they can make a better decision or you can enable a system to leverage that insight automatically. And so, really, our data strategy uses a mixture of these two approaches to contribute to data becoming the lifeblood of our organization.

00:09:48 Juan Sequeda
I like how you said this, a closed loop system. And you want to connect the insights either to a person, so you know here's this goal and there's this team, there's actually this person who needs this insight, but at the same time there's actually a system itself, and then that can be automatic. So this is a very interesting, I like how you're presenting this. It can be a person, it can be the entire process, a system out there, and it's that closed loop ecosystem. So you get back from them to make sure that that was the right thing. And if it isn't, then go correct that, and you're always measuring, right? So question here is what are the things that... What should you be measuring? How do you come up with what needs to be measured? Because you can measure so much stuff, right?

00:10:36 Alexa Westlake

00:10:36 Juan Sequeda
You can boil the ocean on measurements and stuff.

00:10:39 Alexa Westlake
And that's exactly the main issue, is data teams are being pulled in so many different directions and it would be impossible for us to measure everything that everyone wants to measure. And so what we do is we create a culture of joint ownership over success metrics and the idea is to look at the results of the most prominent and current projects. So we want to measure what we're aiming to achieve within that quarter. So whether this is an output or an outcome, you're not going to see outcomes every single quarter. If you're in the foundational stage, you want to measure outputs. And so, really, we work really closely with our business partners to ensure that what we're measuring is the most relevant thing right now.

00:11:37 Tim Gasper
So that's interesting. Is there a maturity curve around this then where you may actually start a little more output oriented before you get to outputs?

00:11:47 Alexa Westlake

00:11:47 Tim Gasper
And it sounds like outputs are a little more long- term oriented.

00:11:51 Alexa Westlake
Absolutely. And so I'd say, and we talked a little bit about this, is we look at it like a waterfall. You look at your output measurements and then you're trying to translate those to outcomes. The idea is, if you're hitting those crucial non- negotiable outputs in the certain amount of time, then you should see outcomes in the next quarter, in the next two quarters. And this maturity curve, we're still getting good at this, and I think it's something that is a muscle and it's very hard to develop in theory versus in practice.

00:12:38 Juan Sequeda
What I find interesting already diving into this whole outputs and outcomes, first of all, is that I don't think a lot of people realize the distinction about them. So I'd love to go into some examples of let's brainstorm here live about what do you consider an output and what do you consider an outcome? But I think, one, having that distinction that people should be able to say, " Well, is that an output or an outcome?" And then be able to go say, " Well, all these outputs are helping us go drive this outcome." I think that's another interesting takeaway here. And something that I really like here that we're talking about is that the outcome is not something that's going to happen every quarter. I think people were like, " Really?" They want to be successful, they want to do all these things all the time. They're hoping to go see or expecting to go see all the success, we're being successful every quarter and so forth, and so we're expected to go see those outcomes. They're like, " Well, no, it doesn't happen always that way. You have to go do all this work and it's not going to just, every quarter, all the work is perfectly going to fit in three months," right?

00:13:36 Alexa Westlake

00:13:36 Juan Sequeda
It's actually a lot of... I mean, this goes into your yearly planning and how these things happen. So anyways, I'm starting to rant here a little bit, but I'd love to go back and let's go share some examples, come up with about outputs and outcomes and how this would get connected.

00:13:49 Alexa Westlake
Yeah. And I'd say, to build on that, there's this cascading cause and effect and the storytelling around the metrics and how and when they'll have downstream impact is almost more important than the metric itself. So the way that you're telling the story of we did this and we split it up, and I'll add some me to this, we split it up in three different categories. There's the system improvement metrics that are non- negotiable outputs, things like we're going to reduce load time, we want to reduce costs of, say, our snowflake spend, et cetera, outputs. And then we look at the process automation piece of this, so saying, " If we're doing these outputs correctly, now we're going to measure the maturity of our key processes that we're looking to actively improve." And then the third pillar is around people experiences, which those are the hard outcomes that stakeholders, supported by whatever initiative that we're driving, directly feel. So this can be things like a customer satisfaction score. This can be things like increased ARR, decreased churn, those type of hard outcome metrics. And so the overall goal is that eventually the people experience metrics are optimized. However, if you're seeing that our systems are improving, our process automation is maturing, and we're still not seeing that people experience, you have to ask yourself, " Are we measuring the right thing here? Is the work that we're doing actually contributing to increased ARR or is it something different?" And so being able to communicate that story to leadership and to the rest of your company is the important part. I think that you shouldn't hold teams to those outcomes, because ultimately we're all trying to do our best. We're all trying to achieve what we can, but it's more about learning together and it's not supposed to be a got- you. If you're not hitting those outcomes, they shouldn't immediately slash funding. It should be we should, all of us together, because like I said at the beginning, results do not happen on your own, we should all be working to identify that people experience metric that is the most relevant for the work that we're doing.

00:16:36 Tim Gasper
Yeah. I love the way that you're talking about this as a collaboration and also not a zero sum game. And I think that's really important. And one of the frameworks that this reminds me a lot, as you're talking through this, is actually OKRs, objectives and key results. And I know that one of the key tenets around objectives and key results, if you're doing them right, and by the way a lot of companies do them wrong, but if you're doing them right, is that it's not supposed to be the zero sum game, where it's like you hit a hundred percent of your OKRs. It's supposed to be the process of designing the OKRs, the process of measuring and doing the status updates and communicating about OKRs, and then the process of learning when you miss them and thinking about, " Was that even the right goal in the first place? Do we need to evolve this? Are we measuring the right things?" et cetera, et cetera, that ends up making it so valuable, not literally the goals themselves. So just curious about are OKRs something that ties into this? Is that something we unify around? Is that just one example of how outcomes might be measured? How do you see OKRs playing into this?

00:17:51 Alexa Westlake
Yeah, totally. So before I worked at Okta, I used to work at a leading strategy execution company. And after that experience, I came to believe that there is no one- size- fits- all solution for measuring success effectively. And while you can use, and you should use frameworks to simplify impacts, it has to be carefully curated and constantly iterated upon, and most importantly, it has to be organizationally specific. The idea of having this goal and reverse engineering from that goal is not rocket science. The impact of this approach comes from tailoring this to your unique business and strategy. And so I think, while OKRs or we call them VMTs, or whatever acronym you use to call this result- setting framework, your data team needs to be your best friend when it comes to the governance around impact metrics. Automation should take out the murkiness of self- reporting. And I think a lot of the times with OKRs and some of these other rigid frameworks, people move fast and end up reporting their own goals, which gets to be an issue because this is not a tooling problem, this is a culture and alignment problem. And when you have everyone reporting their own goals, you get a fox watching the henhouse situation, where everyone's trying to inflate their own achievements, and it becomes almost another thing getting in the way of you actually achieving those goals. So I think that the most important thing is to have a non- biased party within those goals to actually define measurement. And so we see success in this by putting our data team in a three- in- the- box type method with our businesses to make sure that we are defining and measuring things from an automation first standpoint, so that no one is saying, " Ah, we're going to increase that number from six to seven." And it's like, " What does six to seven mean? How are we measuring one? How are we measuring 10?" It's really important to have that collaborative aspect to be able to take that, again, fox out of the henhouse.

00:20:32 Tim Gasper
Yeah. No, I think you're saying that very well and talking about an anti- pattern that often forms here. And I think there's two things. One, is the data team and the role that they're playing in the formation of the goals. And I think a lot of times I see teams set goals or set measurements that ultimately either they could never achieve had they just looked at the goal or maybe it was way too easy if they had looked at the numbers or they don't actually have the ability to measure it. So you end up spending the quarter figuring out how to measure the thing that you had just said that you wanted to measure. So obviously that's more around goal- setting, but then there's the actual execution. How do we hold each other accountable and have third parties, even though you're all working for the same company, obviously, helping when you're actually going to achieve that goal? It seems like maybe that's some of the collaboration that maybe at some companies it's happening better, Data team and the actual line of the business, but maybe in other companies that collaboration is happening much less so, right?

00:21:41 Alexa Westlake
Yeah. Albert Einstein has a famous quote about this, " In theory, theory and practice are the same. In practice, they are not." And we see this, in some projects, it's much easier to do this than others. And the most important thing when you're looking at this is just transparency and being able to communicate that out. And so the impact measurement framework that I'm really trying to instill has four pieces. It's accountability within the planning, so identifying those impact metric sponsors. It's alignment. We learned to... We work to validate and make sure that we've defined the baselines. I'd say that, like you said, getting baselines is the hardest part of this. And you can spend quarters and half a year getting those baselines and it's because this is a really hard muscle to develop. And so after you have a baseline, if you can get that baseline, you want to establish consistency in monitoring, so making sure that there is a performance measurement cadence that's embedded into the actual operating rhythm. So we have a monthly council to do this, to measure how effective our scale is, but if you can get those three pieces, the most important piece is the fourth piece, and that's transparency. And it is debriefing and making sure that there is a broader reporting and communication plan that eyes can actually be on. And we had the CIO of Workday actually come and speak at Okta today, and I asked her a lot of questions about her impact measurement, because she had a lot of really interesting perspectives on it. And she said that the number one thing is transparency, which is also happens to be the hardest thing to develop, because like you said, a lot of the times leadership knows exactly what they want to measure. It's when you get down to the actual actual execution of whether we do measure this and whether it's something that we can even look at historically, that's when this gets hard. We want to introduce radical prioritization. We want to introduce ruthless prioritization, but it takes a long time for the business to get on board. And transparency is how you get on board with the business and everyone needs a formula to get on the same page and that formula is oversharing.

00:24:14 Tim Gasper
Mm- hmm. Yeah. No, I love that. There's so much there. So I think just a little bit of an aside, breaking the fourth wall a bit here, I think one of the reasons that it's exciting to talk to you, Alexa, and learn about how you're doing things at Okta and then sharing that to our audience here is that I feel like you all have a very thoughtful approach and you've got some really great tools and frameworks that I think others can really benefit from that are a lot earlier in their journey or maybe hitting the reset button on the previous approaches that they've been doing things. So I think you just mentioned three things here, and Juan and I are back channeling a little bit, and we're like, " Oh, let's unpack those a little bit more." So those three things are the three box method that you were talking about, the three boxes, the four pieces, curious about, so you named out the four pieces, but how are you applying them? And then the council is another thing that is a really awesome tool. So maybe we rewind a little bit and let's go to actually the three box method. Can you talk a little bit about what does that mean, that the data team is three in a box or there's a three box kind of thing?

00:25:26 Alexa Westlake
Yeah, it means that results rarely happen because of one single team. Often there is a business team, there is a IT, or what we call BT team, and then there is an engineering implementation. So whether it's three, whether it's four, whether it's two, it's just making sure that there is a culture of joint ownership between success metrics. So it's not just the organization that has taken on this, it's all of the other organizations within your company who then help to make that project successful. So I think that alignment is the hardest part. And so making sure that every team, not just the team that's trying to see that outcome, understands that you're working toward that outcome. So whether it is the actual systems team that's working to decrease load time, that they understand we're eventually decreasing load time in order for our salespeople to be able to get to information faster, which means that our salespeople have to tell us what that key information is. And we're not looking to optimize all of Salesforce. We're looking to optimize the key pieces that are the most important pieces to the business, so making sure that the business is not disconnected from the technical teams that are doing that implementation, that we're all moving as one unit.

00:27:02 Juan Sequeda
So I love the three in the box method, right? Business, IT, engineering, implementation. They could be three, it can be four. So these are the groups that should be together to define what we're calling that, it's culture of joint metrics. We'll discuss these four pieces, and if I got them right, accountability, alignment, consistent monitoring, and transparency. Did I get those four right?

00:27:25 Alexa Westlake
Yep, exactly.

00:27:27 Juan Sequeda
So this is spot on. I really love these four things. Now we're talking about a scaling the council. Please continue.

00:27:39 Alexa Westlake
So we're currently piloting this council to measure scale. Okta is in the phase of immense growth where we're now focused on addressing all of the debt that we have taken on, which is the not so sexy part of business. If you ask your business, " What's your most important priority?" I guarantee you none of them will raise their hands and say" Tech debt is our most important responsibility." So it is the job of our IT org to actually say, " Hey, you want this. Let's back it out into actually what it takes and what that foundational work required to drive that growth and innovation actually is." So it's arguing that tech debt and our data and IT teams need to actually do the foundational work to support that growth.

00:28:42 Juan Sequeda
At what scale in an organization do you start thinking about this? Again, I'm listening to everything you're saying. I'm like, I'm just already putting myself in the mindset of somebody in a smaller startup and larger company. I mean, our audience, if I look at the demographics of our audience, they're literally all over the place. And everything you're saying here is fantastic, because it's something you can apply, but I think one thing that is different is on the scaling side. So you don't want to start over- engineering earlier and stuff. In your experience, when do you start realizing, " Oh, we got a lot of tech debt, and we got the scale that we need to start really focusing, putting some priority on this"?

00:29:22 Alexa Westlake
Yeah, and I've worked at organizations that are much bigger than Okta. I've worked at organizations that are much smaller and more of a startup feel. And I'd say when you are small, you have to take on tech debt. There is an opportunity cost. You can't hire 10 data engineers when you don't have a lot of money coming in. And so I'd say that when you're looking to switch to a more profitable business model, and it's not about capturing your target market, once you have that target market and you're looking to expand, that is when you're looking to scale. And I'd say people throw around the words scale a lot. We talk about often scaling for durable growth. And it's hard,'cause scale and durable growth are at odds with each other. And so it's a very hard balance that I think that most IT teams of a company of immense growth have to face, is at what point do you raise your hand and say, " We can't take on any more of this"? If we look to onboard a thousand users, it's different than onboarding a million users if you have a customer that has a huge login stream. And so it's very unique to your organization, but it is that kind of delicate job of your IT team to raise your hand and say, " Actually, we can't keep operating the way that we move." And I have to say, our CIO, Alvina, does a fantastic job at this. She advocates when it's time to advocate, but she also knows when it's time to let the business dictate what they need. And so I'd say that that job of leaders is something that it's very hard to do and you need to make sure that you have strong leaders. And I will only work for companies with strong leaders who can actually articulate when it's time to flip the switch.

00:31:36 Tim Gasper
Amen to that. And I think that there's a really important role that leadership has to play here. I know we're talking today a lot about data teams, but when you say things like durable scalability, that kind of mission and putting that out there requires leadership that's willing to embrace some of the trade- offs that come with statements like that. And I think that's so critical, to either have that leadership or somebody needs to step into that leadership and provide it. Just to double- click a little bit more into this council, what does it look like? Is it people from those three categories, those three in a box that you mentioned, that represent those three technical layers? Is it different parts of the business? What does it look like in terms of the membership? And is it like a monthly meeting where you bring a certain agenda and things like that?

00:32:34 Alexa Westlake
Yeah, exactly. So it is leaders from across our organization and it is at a monthly cadence. And it kicked off within our IT organization or our BT organization, but it's successful because of the partnership with our business. And I wish that the question at the beginning was not just snacks, but what would you put in the backpack, because I think the most important thing is a notepad. And what that scale council does is, one, it provides a forum for planning and all of the important things that we need to do to scale, but it also provides a notepad for our business to iteratively speak up and provide commentary on the work that we're doing. A lot of the times, councils can be very one- sided. It's something where we give readouts. That's not what this is aimed to do. This is aimed to present information that eventually people react to, and it's a two- way street versus a one- way street. And I think that where a lot of data and IT teams fail is they report out and they don't ask for feedback. They don't ask for the business to engage. It's not a conversation. What the scale council, and not just the meetings, but the fact that we're bringing everyone together, is aimed to do is to create that conversation with our business to make sure that we are iterating on this and we're not going to get to the end of an engagement and actually find it didn't meet the needs of our business. We really need to make sure that we're constantly listening to our business.

00:34:23 Juan Sequeda
And this shouldn't just be only for scale, right?

00:34:28 Alexa Westlake

00:34:29 Juan Sequeda
I think this is for all of the three in the box, right?

00:34:33 Alexa Westlake
Exactly. Exactly. And it's hard, because like you said, and Scott Hirleman has a really great quote around this with Data Mesh, is a lot of the times we think, because we've put time and effort into Data Mesh, that we're in the seventh inning of baseball. And we talk like we're in the seventh inning, but really this is the second inning, because it doesn't matter how much time and money you've put in, the inning could go on forever as long as you don't have three strikes. And so being honest about where we're at, regardless of how much time we've spent on something, is really important. And I'd say that, while we're piloting this, eventually we'd like to move toward councils for all of our huge cross- functional efforts,'cause that's the important thing, is that closed loop ecosystem.

00:35:30 Tim Gasper
Mm- hmm, and making sure that you're bringing together all these different ideas across the organization so that way it's not just that top- down approach, it's marrying it with the bottoms up as well.

00:35:44 Juan Sequeda
This reminds me also, the finite and the infinite games. The purpose is winning and we know where the end is, like, " Wow, this thing, the purpose is to continue playing." And the business that you are in today, you want to make sure that business is going to evolve, it's going to change, is going to grow, but the game continues. Right?

00:36:04 Alexa Westlake

00:36:04 Juan Sequeda
And things happen and you just need to know how to adapt around that stuff, but how do you figure out how to adapt is like you are really having that, the three in the box. I love this, the three in the box. It's a great way of putting this together. So for folks that are listening, one of the things I like that you said is the quote here, the strong leaders, " Strong leaders know when to flip the switch." And I'm thinking about more junior folks who are seeing those folks around them, seeing folks who are leaders. What is your advice for the junior folks who are like, " Oh, am I in the right place? Am I surrounding myself by strong positive leaders?" Or maybe you're not, then you got to go figure out what to do. I'm curious, because you've seen, you've been all over the place here. I really love the advice you've been giving, so let's get some honest, no BS advice here, folks, from the junior.

00:37:08 Alexa Westlake
See, I have worked... I'm lucky to work for strong leaders now, but I have worked for not so strong leaders in the past. And the number one quality that I'd say you have to look for is empathy. Do your leaders put the people next to the people at the execution level in the room and give them a voice or do they do this performatively? And I'd say that whether you... Well, actually, it's two things. It's empathy and it's self- awareness, and so if you can use your gut and trust your gut of, " Is this person that is speaking to all of us speaking genuinely and are they listening?" That's the most important thing, because a strong leader can admit that they're wrong. A strong leader wants transparency and they want to understand their business. They don't want to just tell people" We're moving in the right direction." And if you're working for someone that listens to you and has empathy for you, you know that and you'll know that in your gut. And you have to listen to your gut, which is something that I continue to learn every day,'cause sometimes I'll look back and I'll go, " I knew that this was the case, but I didn't say anything." A strong leader will let you ask that question without making you feel small.

00:38:45 Tim Gasper
Yeah, well said. When you said performative empathy, it sent a little shiver through my spine, because I think that is something that you see when leadership is still forming or not there yet. And I think that one of the themes, a couple of the themes that we talk a lot about on the show are curiosity and empathy and how critical they are, not just in leadership, but really in all roles to be successful. And it's easy to say things like" I've got an open door policy" and then shut the door.

00:39:30 Alexa Westlake
Or punish you for what you've done within that door. Both of those things reign true, and it's so important. Like I said, we are human beings. We are not human doings. And if someone treats you like a resource rather than a strategic partner, no matter what level you're at, that should be a red flag for you.

00:39:53 Juan Sequeda
And it goes back to the three in the box. Really defining these themes is you should not be seen pointing fingers or stuff. We're all in here together. We're all figuring out. When you have these councils, it's a two- way street. It happens a lot. It's like, " Yeah, I'm just going to go tell you what I've done," and then goes back to like, " Well, here are the outputs. Here are the results of those outputs," but all of these things get connected and we need to be able to understand how do these outputs get connected to these outcomes? Well, that happens when you are talking to everybody. You have all those three teams in that box and having discussion. So I think this is a fantastic discussion that I think, when I'm talking to folks about where data teams are and how they're providing value, I say this, I write this on LinkedIn, like, " Yeah, you got to focus on making money and saving money and mitigating risk and show me the money," all these things, but really when you unpack that, there's all this people aspect.

00:41:00 Alexa Westlake

00:41:00 Juan Sequeda
I think it's not just, " Oh, just about money." I mean, yes, we live in a capitalist world. It is about money, but the way to get there is really about how we have that people and the empathy. And I think those strong leaders and I think we need to be able to learn from those strong leaders ourselves and just especially for a lot of the junior folks that I'm seeing, that they want to be those leaders. These are the qualities, folks, that you should be looking and having them as your mentors.

00:41:26 Tim Gasper
Here's the recipe.

00:41:27 Juan Sequeda
And before we start wrapping up here, how do we get there? Change management is one of those things here we should, I would love to address a little bit. What are your thoughts about this? How do we get to this very ideal situation that you're talking about, that we actually have everybody talking together, the three in the box? How do we get to that three in the box?

00:41:49 Alexa Westlake
Yeah. I'd say change management and just adoption in general is always and should always be a work in progress. You don't get to tick done on it, ever. That piece is not easy. If you're doing it right, you're constantly iterating. And even if that iteration isn't changing anything, if it's just still doing that pulse check and making sure that you're checking in and looking that you have the right levels of adoption to achieve what you consider is success, it's not done. And I think a lot of the times we assume that, because we provided enablement, that it's done, but communication is ultimately, it's not about the communicator. It is about the person you're communicating with. And if they don't understand, no matter how good of a job and how eloquent you think you've put together this argument, if you are not on the same page as that other person, you have not done a good job communicating. And the only way to do this is to lean in to that uncomfortable aspect of transparency. And I emphasize uncomfortable here because it always will be uncomfortable. It's not fun to hear what you've done wrong all the time. However, it's really necessary to keep a log of that and to keep a log of how can we lean into the uncomfortable nature of change management.

00:43:23 Juan Sequeda
I think this part of change management is a great way of closing this out, because I think it's never done. It's always work in progress. You can't just check, " Oh, we're done. Just change." It goes back to almost that infinite game. Our goal here is just to keep playing. And so we're never going to be done with this. Alexa, this is a fascinating discussion. I can't wait to actually get to meet you in person. We can keep having these discussions.

00:43:51 Alexa Westlake

00:43:51 Juan Sequeda
I'll be in San Francisco soon, so I'll keep you posted, but let's hit to our AI Minute. We haven't talked about AI, but I'm sure it's something about AI's on your mind. I'm curious. So one minute to rant about AI, whatever you want. Ready, set, go.

00:44:07 Alexa Westlake
Awesome. AI has changed and will continue to change our world. In my opinion, AI's potential is actually under- hyped and we are only seeing the beginning. However, garbage in and garbage out is more of a reality today than it ever has been. And as AI becomes accessible and used more broadly, there will be more and more solutions out there that are better at understanding and marketing your pain than actually providing any solution. And if someone claims to unify all your data, optimize all your data, or protect all your data with some seemingly magic black box AI solution, well, like my southern grandmother used to say, " Something in the milk ain't clean." And if you do not have strong data foundations in order, there is no AI that will truly be able to do this for you. Make sure you understand what foundations are needed for successful implementation before signing onto expensive software. Listen to the people that directly work with your data. Resource of them effectively and understand that advanced analytics capabilities need high data maturity and AI is not, and never will be, a replacement for data modeling, data pre- processing, data governance, and good talent. In order to leverage AI, you need solid data foundations. That's it.

00:45:29 Juan Sequeda
Amen, and especially the data modeling part. It's all about the knowledge and the talent, the people. They're going to be around there. AI will make the people more productive. So I do say that AI is not taking your job away, but the people who are using AI, those folks will take your job.

00:45:47 Alexa Westlake

00:45:49 Juan Sequeda
All right. That was awesome. Tim, let's go. Let's go to our lightning round questions. I think you got them ready. I'm going to kick them off.

00:45:58 Tim Gasper
Yes, go for it.

00:45:59 Juan Sequeda
All right. Number one, if a data team is output- oriented, is that inherently bad?

00:46:06 Alexa Westlake
No. You have to be output- oriented at the beginning. I think that it depends. The answer is no, because it's not every case. Yes, if you've done the outputs, you should be looking at outcomes, but you have to measure outputs at the beginning, you have to. You're not going to be able to reach those outcomes without the outputs.

00:46:28 Tim Gasper
I think that's an important distinction, that it's not outcomes at the expense of outputs, it's the two in combination.

00:46:37 Alexa Westlake

00:46:38 Tim Gasper
Mm- hmm. I think that's important. All right, second question. Should it be mandatory for data people to join the business teams when they set goals such as OKRs?

00:46:48 Alexa Westlake
Yes. Yes. I'll say yes, asterisk. Don't put every single data person in there. I don't think you need 10, 15 people, but you need at least one. You need someone to be able to raise a hand and say, " Hey, that's not feasible at all" or, " Hey, we need this, this, and this before we can get to this." And so I'd also say, use this sparingly, because you are constantly in a relationship with your business teams, and ultimately, and I say this to my team all the time, the business doesn't need us. And if we are making things very hard for us, and say this is a relationship, they will divorce us. And once you get divorced, you can't come back from that. And so you need to be able to put the right representation of a person who can softly and gently ask the correct questions to be able to maintain that relationship so you don't get divorced.

00:47:50 Juan Sequeda
That is an honest, no BS hot take right there. The business doesn't need us, and if we make things more complicated, they're just going to divorce us.

00:47:59 Alexa Westlake

00:48:01 Juan Sequeda
That's another thing that needs to go on a T- shirt tip.

00:48:06 Tim Gasper
I'm just thinking about that right now, because I think the prevailing thought around data teams right now is a bit entitled, honestly. I think it's like, " Damn the business. They can't live without us. How would they even do their jobs?"

00:48:18 Juan Sequeda
Fully about this. This is what gets me so fricking annoyed, is that you have all these data teams, like, "Phil is so fricking entitled." I'm like, " No, get off your... No, stop it. Stop it."

00:48:28 Alexa Westlake
Exactly. They can operate without us. Again, most decisions don't need data. You can make decisions with your gut most often than not. When you're making decisions to fail, you need data, but for the most part at the beginning, they don't need you. You have to prove that they need you.

00:48:43 Juan Sequeda
And even they haven't making it with the data the way... I mean, the business have been doing it already for so long without these sophisticated systems. The goal here is to make it more efficient, but then we get off into this entitlement of... Anyways, this should be another episode.

00:49:00 Tim Gasper
There's another episode in that.

00:49:01 Juan Sequeda
About how entitled the data teams feel. Next question. Well, this next question is a little bit related. Let's see. Can you measure the impact a data team is having or is it always going to be indirect?

00:49:18 Alexa Westlake
I think-

00:49:19 Juan Sequeda
inaudible say it's direct.

00:49:22 Alexa Westlake
I'd say, no, it won't always be direct. It's hard, because again, a lot of work the data team does is so foundational that you can't always look at the way it translates. However, it depends what role your data team is taking. So if your data team is like an extension of your product team or an extension of your marketing team, then yes, but if your data team is... Yeah, it really depends on the role that your data team is. And the role that your data team plays in your organization will also have to do with how mature your organization is. And so it depends on what the tenants and core strategy of your data team is. That wasn't a yes or a no, but that's my answer.

00:50:16 Tim Gasper
That's an important nuance. It's kind of both, depending, right?

00:50:19 Juan Sequeda
I think everybody who listens to this realized that we have these lightning round questions that are yes and no, but they're never end up being yes or no.

00:50:26 Tim Gasper
Yeah. We try not to pigeonhole our guests. We're nice. All right, so last question. Is it getting harder to be a data professional?

00:50:38 Alexa Westlake
Yeah, and it's getting harder because everyone thinks that AI is going to do our job. And like I said during that AI rant, most executives don't understand data, and that's a good thing, because an executive's job is not to be able to do analysis. It's to be able to be as decisive as they possibly can. And you need to do everything that you can to enable them to be decisive. And now that they have a thousand solutions out there with these people that say, " We're going to protect all your data," you'd say, "Hmm, I don't know if I need a security team because this guy is going to do it for 2, 000 a month and I have to pay you guys you know what." And so it's getting harder, because now these people who are inherently technical people, also have to be very good at arguing and debating. And again, that goes back to having a leader that is good at putting themselves in the room and knowing when it's time to push and when it's time to step back. And so I'd say, yes, it's getting a lot harder, because data professionals have to almost be everything. And when you're everything, it's like how can you hone in on one thing? And so you have to build data teams with people that are both technical and inherently non- technical, because you need to be able to wear a thousand different hats and no one can do everything. And so making sure that you have a team, and I'm very lucky to work on a data team that I can say is extremely well- rounded, to be able to both do the analysis and do the arguing is really important and it's just going to get harder.

00:52:26 Juan Sequeda
I'm already imagining a T- shirt that says, " Most execs don't understand data and that's a good thing."

00:52:32 Alexa Westlake
That's a good thing.

00:52:33 Juan Sequeda
And then go walk around a conference, like big data lending will be, and see what people will say about that.

00:52:39 Tim Gasper
It's a little bit counterintuitive, but it makes sense. It's perhaps a hotter take than it should be. And I love your statement about everyone thinks AI is going to do our job. And I see the hype building, where people are like, " Yeah, AI. BI will finally be solved." Let's go rewind about 20 minutes. Alexa, you said garbage in, garbage out, and people have 10 BI tools and only one data quality tool. Maybe they should have one BI tool and 10 data quality tools. Just throwing that out there.

00:53:10 Alexa Westlake

00:53:12 Juan Sequeda
All right, so we've gone through a lot. Tim, take us away with takeaways. Kick us off.

00:53:19 Tim Gasper
All right, takeaway time. So we started off with honest, no BS outputs versus outcomes. How do we get there? And, Alexa, you really started us off by saying, " Hey, a lot of folks may be focused on ticking boxes and just the output aspects," and later you did add a lot of clarity that it's not that you shouldn't focus on output, but it's that the outcomes are ultimately where you're trying to drive towards and that's where you're really going to get scale and impact from your data teams. Outputs are leading indicators of outcomes. If we aren't understanding outcomes, then what's the point? And you said that over at Okta, you've got certain key drivers of impact for your business, whether it's the development of key features or enhancing the security or improving the customer experience. And so driving the metrics that drive those particular outcomes, that's where a lot of the data team focus needs to be. And in collaborating with the different parts of the business, whether it's marketing, customer success, et cetera, to drive those results, which I think was really, really important. You talked about understanding the customer journey and really collaborating around that and really trying to generate better and deeper insights that can be deployed in collaboration with the business, this idea of a closed loop ecosystem, where the insights connect to the right person or a system to leverage it as automatically as possible. We talked about measurements, right? You can't measure everything. You have to create a culture of joint ownership. You talked about how outcomes aren't necessarily a quarter or less. A lot of these outcomes are a lot longer range. And you talked about these three categories of things, system improvements, which are non- negotiable, like reducing load time, reducing costs, process automation. If we're doing these outputs correctly, we should see impact around process, and then people experiences, which are really the hard outcomes that people really feel, things like customer satisfaction, churn, et cetera. And all three of these categories are important, but you can't put the burden of the outcome on a team. It has to be more than one team. It has to be a shared burden. We talked a little bit about OKRs and how that may be related, but that's one framework, and don't become over- indexed on the frameworks. And through talking about OKRs, you mentioned about these three in the box method where results don't happen, back to that comment, just because of one team. It's the business, it's IT, and it's engineering, slash, implementation working together. And whether that's three teams for you or four teams for you, the key is that you're working together and you have joint metrics. So thought that was really, really, really critical there. There's so much more, but Juan, I'm going to pass it over to you to give your takeaways.

00:56:15 Juan Sequeda
So in addition to the three in the box, you talk about the four pieces, accountability, alignment, and getting those baselines is the hardest part because it's a real hard muscle to develop, consistent monitoring, and then transparency. The transparency overshare, this is how we get on board with the business, because we're being transparent. And then we talked about this whole council to measure scale. It's really, when you're at that moment of scale, addressing all that tech debt that we have, that's not sexy, but hey, if you want to drive growth and innovation, that's the foundational work that needs to be done. You need to ask yourself, " What is that foundation that we need to be building on so we can go scale?" Now, note that if you're coming from the small organization, small company, you have to take that tech debt, right? There's opportunity costs, so you have to take that tech debt, but at scale is when you're actually switching to a more profitable business model. You have your target market, that's when you start switching. And I think the strong leaders, they know. They understand when you have to flip that switch. When do you raise your hand and say, " We can't do this anymore this way. We can't operate this way anymore"? Those are the moments that you start realizing you're switching into scale. And part of this council is leaders from across the org, that three in the box, you talk about even monthly cadence, and in your case it started from IT, but it's really successful because it works with the business. And I think you have this great commentary about bring your notepad, like, " Hey, we're going on that road trip. Also bring your notepad, because there's a lot of comments around here that we want to keep track of." It's not just a reporting, like, " Here's what we did." We want to also get your feedback and it's a two- way street conversation. And then talk about strong leaders. Remember, they're the ones who know how to flip the switch. Look for leaders. You know they're strong leaders if they're empathetic. Look for empathy and self- awareness. Is that person who's speaking to us, are they being genuine? Are they listening to us? You want that transparency. Don't just say, " Hey, we're moving in the right direction." And strong leaders will really make you feel comfortable asking those hard questions. And then we wrapped up with change management. Adoption and change is always a work in progress. You can't just tick the box right there. Keep doing that pulse check to make sure that you have that, adoption and changes for the stage you are in is key, and let's lean into the uncomfortable nature of change management. How did we do? Anything we missed?

00:58:37 Alexa Westlake
I love it. You guys did a really great job and I'm so grateful to be here today talking to you. This has been a great experience.

00:58:46 Juan Sequeda
Well, we just repeated what you... This was all you.

00:58:49 Tim Gasper
Yeah, we didn't do anything. You did it all.

00:58:51 Juan Sequeda
We just listened to you.

00:58:52 Tim Gasper

00:58:52 Juan Sequeda
So to wrap up, throwing it back to you, what's your advice? Who should we invite next and what resources do you follow?

00:59:00 Alexa Westlake
Yeah. Well, I'd say my advice is, for data teams, the closer you are to your business, the better your data is. Take the time to listen and listen iteratively, and also wear sunscreen. I got fried on my vacation and I'm so scared of skin cancer, so wear sunscreen. It's good advice for everyone.

00:59:28 Juan Sequeda
This is a better one. I love how everybody, well, everybody gives really profound advice like this, right? Either about data, about life, but then once in a while we have those little comments which are really important advice too, like wear sunscreen. And there's one advice that still sticks with me. A long time ago, c couple years ago, I remember it, Pat Berry, he was on our... Yeah, he said, " Have a long USB cable." I have a long USB cable I have all over the house with me. It was a game- changing advice. Anyway, back to you. Sorry.

01:00:05 Alexa Westlake
I love it. Who should we invite next? And I have to say that I went through and everyone that I thought we should invite has been on here for the most part. The one person, and he has no idea I exist, but I love following him, is Robert Yi. And the work that Hyperquery is doing to work to revolutionize shareable analysis, it's so cool. And Robert, I'm following you. I'm rooting for you. I think you should be on this podcast.

01:00:37 Juan Sequeda
I'm also curious, who are you thinking about that has-

01:00:40 Alexa Westlake
I was thinking about the girls from Prequel and-

01:00:45 Juan Sequeda

01:00:45 Alexa Westlake
...I was thinking about Chad Sanderson, which is some, is a resource I follow, that I love Chad. I love his simple truths about data quality. He tells it like it is. Yeah, and I guess to move into, now that I've used Chad to transition into resources I follow, I am a firm believer and I always have been in Zhamak Dehghani's Data Mesh and a big fan and friend of Scott Hirleman. I love the Data Mesh paradigm shift. I started my career looking at Data Mesh and I have never looked at something and said, " This makes sense," but basically to enable capabilities to get out of your expert's way is the only way that data works at scale. You have to enable self- service. You have to look at data as a product. And I think that the Data Mesh captures that so perfectly in the way that this is not a tooling problem. This is a people, a processes, and an organizational problem, and we should treat it as such.

01:01:57 Juan Sequeda
I think that is a beautiful way of wrapping up today. Thank you, Alexa, so much. This was a fantastic conversation. Just a reminder, next week Tim and I will actually be in... Next week, we're in Amsterdam, Tim. And we'll be doing the-

01:02:12 Tim Gasper
Yeah, we'll be coming to you from Amsterdam.

01:02:16 Juan Sequeda
We'll have a rant session next week, and then after that, we'll actually be at Big Data London and we'll be live for Big Data London with Chris Tabb. And with that, Alexa, thank you so much. We really appreciate this conversation. It was phenomenal.

01:02:30 Tim Gasper
So many good insights.

01:02:31 Alexa Westlake
Thank you. Thank you for having me. I appreciate it so much and I hope everybody that's listening had a good time. I sure did. Thanks.

01:02:41 Juan Sequeda
Cheers, everybody.

01:02:42 Tim Gasper
Cheers, Alexa.

01:02:42 Alexa Westlake

chat with archie icon