Jonas Christensen 2:49
Kshira Saagar, welcome to Leaders of Analytics, it is so good to have you on the show.
Kshira Saagar 2:57
Thank you so much for having me, Jonas. It's a pleasure to be here.
Jonas Christensen 3:00
It is truly wonderful, because I have followed you career for a little bit because you are a Chief Data Officer and that is the topic of today's conversation. There are a few but increasing number of Chief Data Officers out there or Chief Data and Analytics Officers, they might be called from time to time. But you're really forging a path for the rest of us in the industry in that role. So, we're going to hear more about that. Before we get to that, let's just hear a little bit about you. So, perhaps could you tell us a bit about yourself, your career background and what you do?
Kshira Saagar 3:32
So, ever since I've started working officially, I've always been in the data space, it used to be called different things. When I started work it used to be called Business Analytics and then Data Analytics and then now it's Data Science. Hopefully, we Data Arts in the future. So, it's just that's what I've done over the life. And what I tend to think is I help people make a decision with data. That's what I do. And you've probably heard me say this quite often, I like to think of myself as the ''Lost and Found'' guy. So if you lose something, I find it with data. That's what I do: Data and Tech. And so if there was no data and no tech and then all this happened and we lived 100 years ago, I'd still be in the railway station, being the ''Lost and Found'' guy trying to find something that you've lost. That's what I do for a living. And yeah, so when I try to do this at scale, I tried to do this for an organisation, the elements and aspects to it, but through an algorithm or a dashboard or analysis or any aspect of data. At the end of the day, I believe it's all about people making smarter decisions with it and better decision for them. So, that's what I make happen. I help be the customer, be the employee as an enterprise or be somebody else, a partner that you work with. Help everyone get to make smarter decisions and if we can help make that happen with data, that's what me and the team do for a living.
Jonas Christensen 4:42
That is truly music to my ears. And I do think that for modern business leaders, it is a blessing to have data even though it can seem a bit daunting at times because previous generations would have to only use gut feel. When you can combine gut feel with the actual facts, decisions are much better. Now, you did mention that you have worked in the field of, let's call it, Data Science for now, since you started your career. But back then, I've been through the same journey, there wasn't anything called that and you couldn't do a degree in the subject and so on. How did you end up in this field?
Kshira Saagar 5:17
So, I think growing up in high school in Tennessee is amazing. You like a subject for the teachers, what I always believe, more often than knowledge. It's the teacher that makes the subject interesting. People pick history, because they have a really history teacher or physics. And we had an amazing math teacher in high school. And so, he would try to put everything that we learned. So, he taught us Poisson distribution, right, and it's one of the most void things that you can learn. When he tried to explain how Poisson distribution works, you're standing in a queue because everything in a queue is a Poisson process. So which lane of the queues, if there are two queues and you're standing two windows of the same queue, which queue should you pick, and you can explain that real Poisson process or understanding how infinity works or how calculus works in the real world through real examples. And when you try to do that, you really have a fascination for math and everything math. And so that translates into when I grow up, when I go to get a job, I want to work in the math space. You don't know what that looks like. You can obviously go into theoretical mathematics and do a lot of research. But luckily enough, companies that came advertising when I was in uni the tagline called ''Do the math'' and I'm like, ''That sounds like the company that I want to work for''. That's how it happened. I didn't realise it was a data company, or we were trying to solve with problems with data. It's mostly statistics driven, problem solving, and scattering a lot of math and that's how I got into the field. And then one thing led to another where stats became less and less important and more and more it was about the technology, because people were commoditizing a lot of this. Anybody could write a piece of regression code. Previously, you had to understand how regression works, what are the assumptions, what is normalisation, what was heteroscedasticity. You needed to know by heart what it meant or how it worked. And then as things thought, commoditized more and more, anybody could write one line of LM of something or whatever it is, and then they could get it done. And so, then trying to go on the journey and try to pick it up is what I've done in the last few years.
Jonas Christensen 7:07
Nice. And you're highlighting the fact that teachers can have such a huge impact on our lives, whether they like it or not. And we probably don't often reflect on that, as leaders as well, that we are teachers and we can set people off on a path. I remember having a math teacher who questioned on why are we learning all these equations and so on. And they said, ''I'm not sure what you're going to use it for. But all I know is I've got to teach you what's in this book and you got to be able to pass the exam''. And I found that it's not quite the answer I was looking for. So, I didn't really enjoy that process until I discovered analytics. And all of a sudden, data became this magical tool for me. Similar process but unfortunately I had the opposite teacher to you.
Kshira Saagar 7:47
So, I try to do that. I tried to pass it on in my own small way. I work with high school students here in Australia or uni graduates and try to do my best to make it look attractive, if you want to call it that. Because more often than not, people are like ''What do I do? Why do I learn math? What kind of job am I going to get?''. And I'm like, ''If you learn math, you can do this cool stuff, when you grow up and get into a job''. People don't make that connection and so, trying to do that is my part to pay it back in some small way.
Jonas Christensen 8:15
Okay, so you go to schools and do educational sessions. Now, Kshira, I want to go to your current role as the Chief Data Officer for Latitude Financial, because it's really an interesting role for me to hear more about and also for the audience, of course. So could you tell us a bit about your role there and the strategic remit that you have as the CDO at Latitude?
Kshira Saagar 8:40
So, I think two things, right: Chief Data Officers, typically, at least right now, my peers, the ones I speak to, more often than not, are the ones that build capability in the form of data. So, they build the data capability in the business. They govern the data. They're responsible for ownership and stuff like that. So, they pocket all of that into the capability. Okay. And then there's the value creation bucket, which is more of the analytics and data science and that's decentralised. That's typically how it's set up. So what I've done and the way I believe in it is: People who build the capability should not be different to people who deliver value, because then what it does is it creates two strata of people. One who's building capability and none of the one who's just generating value without knowing what is possible. And so, we didn't want to do that. We tried to look at everyone together in this version of Chief Data Office, where we have people across the engineering platforms who build the capability to predict interactions. Who not only build capability, but deliver value to analytics and insights, who can add more value to it and then data science and data governance. So, the way we've done it is, I like to think of it as a supply chain of data. Somebody figures out where data is, somebody brings it up, somebody creates value out of it and somebody markets and sells it. So we do the entire supply chain of data for Latitude. And what I've been asked to do is three things. One, make data and insights more accessible for everybody, which is what we tried to do. Nobody should ever feel like they can't see something or they don't know what's happening or it takes forever to understand how things are performing. So, accessibility. Intelligence: making everything intelligent because now, like we spoke about, everything is so commoditized. You can bring data in and you can put it on a dashboard, but how do we make sure it's intelligent and useful? And then make sure it's reliable because if it's not reliable, nobody can use it and it's just complete garbage. So, how do you make sure it's accessible, intelligent and reliable? That's what the CDO teams mandate is right now.
Jonas Christensen 9:01
Yeah and it's such an interesting conversation around how you structured the team around the delivery, because if I reflect on my experience in my career with this, there's often this chasm between the data engineering part that is very busy on making data accurate and stored safely, securely in data warehousing environment versus the analytics functions that are trying to make something out of that. And the disconnect can often mean that we're in one corner building the utopian data warehouse and then the other corner screaming for data that we don't have access to. What are the benefits you've seen of bringing the two together?
Kshira Saagar 11:07
Yeah. So, I think the benefits of bringing the two together is people understand each other's pains. So like, exactly the example you gave, when I took care of data teams in the past, like just a data science team in the past, we work with data engineers who got really excited bringing data in and then that's where their jobs stopped. They're like, ''I brought everything in. I really don't care what you do with it'' and they walk away. And that's a really dangerous business, not because they didn't care. It's just because they didn't know what people did with it. That's when you put the people who use it and the people who bring it together, they can have a really honest conversation about, ''Don't waste your time bringing all of this in. I just need this. Can you make this available for me here and I can use it?''. And the other way around too is because data engineers feel like they do a lot of work and nobody appreciates what they do. Because at the end of the day, they do a lot of work on the back end and somebody comes and presents it all out and they get on the credit. So, if the team work together, they could shine a light on the work that the data engineer - it's amazing what they do too and that's another interesting piece. But the third element to the business, there's data scientists who, say, interpret and make good sense of the data. The state engineers who bring the data. But I also like to believe this third component, translators as I like to call them, these could be the BI people. This could be analysts. This could be product owners or data product owners. These are people who will actually make it look attractive and important, because more often than not, I've also worked with data scientists in the past, who build some of the most amazing segmentation models but won't take the time to explain how it works, what it means or what you should do with it. So, if all those three people work together, then nobody's talking at crossroads and they can actually solve one problem at scale, rather than trying to solve 10 different problems by 10 different people.
Jonas Christensen 11:07
Yeah. So, can we dive into that a little bit? How have you structured your team in terms of skill set and capabilities and remit to have this interaction with the business and make sure that the data products actually become used in the business?
Kshira Saagar 12:54
So the first thing I'll call out is it depends on the maturity of the organisation and where the data is at. Likely for example, I had really good material even before I came in, thanks to all the smart people who've done that before me. The problem was the last mile. Just people not knowing what to do with it. They really had good data. They had good infrastructure. They just didn't know what to do with it. So, that's a different setup to the iconic that I've worked with. We started from scratch, which is a clean slate. And everywhere I go, we do a two stage process. The first stage is having all the analysts and analytics people and insights people in one cohort. They have the skill sets to go interpret data, analyse it, do whatever it is and they have the business knowledge and SME (Subject Matter Expertise) to understand it. So that's one group of people. The other group of people are the people who will build data assets or build reporting assets so that we can then translate that into something that's more important, interesting and converted into reports and dashboards and stuff like that. The third group are the data engineers and data platform people. So, not only do they bring data in to edit pipelines or to swin pipelines or through infrastructure, they also build a platform that other people need to do their data work, be it a machine learning platform or an ETL platform or whatever it is. And then the last group, the data governance people who make sure that everything's happening and humming and hawing in harmony. So, that's how we typically structure for different groups of people. And so once you do a discipline-based approach, then what we do is we try to pick and choose for a particular project one data scientist, one data engineer, one BI person, one data governance person working together with a data product owner to then solve a problem. So if somebody wants to solve a problem for customer acquisition, you have now four people working on that problem knowing exactly what to do and self-sufficient. And that cross-functional approach, it works really, really well, so that everyone feels successful and everyone knows what everyone else thinks that works for us.
Jonas Christensen 14:24
So you've described an organisation that already had pretty good maturity when it comes to being data driven and metrics driven, data literate. You've been there for about a year now. Could you tell us about the organisation you entered and then what you've done since then to further the use of data across the organisation?
Kshira Saagar 14:59
So when I entered, the challenge was three different teams working on completely different problems. So the data engineering team was part of the technology function or the engineering function. And so their job was to just build a data lake and bring data in, bring the tools. They really did not care about the potential problems that was being solved. The analytics team were trying to solve the actual customer problems or volume problems, but they didn't have access to the right data. They could bring the data in that they needed. Then the BI team was building reports for operational support, other the customer facing events. But they couldn't try to solve other problems that these two teams are trying to do. So everyone was solving a local maxima of problems. And what we tried to do is bring them together and said, ''Let's not solve four or five different problems. Let's solve these top three problems, which is accessibility, intelligence and reliability, and let's arrange ourselves in a way where we can then solve these problems. And that's one thing that we've changed''. And the other thing that has also changed has been anybody who does data comes together, there is also strength in numbers. People feel a little bit more comfortable taking on a challenging problem and saying, ''I think we can do it because I have X, Y, and Z and I know he or she can then get it done for me''. That's the other thing that we've done. And third thing is when the organisation also sees that there is a proper data function, not different data components everywhere, they have to get the confidence that this group can actually solve a problem. So for example, we now have the goal to solve all the personalization problems in-house. We now have the goal to do all the credit decisioning problems in-house. Previously that was going to be outsourced to someone else because we thought we didn't have the capability. But now we've shown that we can do that. We aligned ourselves the right away. We're being asked to solve the more intellectually challenging problems, which is really cool.
Jonas Christensen 14:59
So, have you brought in additional skillsets and additional numbers of people or have you literally just brought together a group that was already there and just made it a super Team, a unicorn team?
Kshira Saagar 15:13
So yeah, mostly 92% of the people were people who were already there. So, we've just brought the mechanism. And 8% of the people are the ones that we've brought from outside. We've seeded some interesting new talent like ML engineers, MLOps people, DataOps people, some talent that was missing. But mostly, just repurposing people's skillsets, that are really good in one space, in another space and trying to bring them all together. And people were really willing to step up and everyone's like, I'' want to learn and do something. It's just the opportunity that was missing''. And we just made sure to make that opportunity available.
Jonas Christensen 17:15
Yeah, I find that people in Data Science are, if I call that whole umbrella Data Science, are often looking for career opportunities and opportunities to really impact the business. But there's just not historically been those career pathways and enough of a structure within analytics functions to allow for, not necessarily managerial progression or responsibility for that, but to grow your skillset or your competency and become a really deep subject matter expert in certain areas, because you're too thinly spread across too many things. Is that something that you feel like you have impacted positively as well through this?
Kshira Saagar 17:51
Yeah, so fair thing, because people care about more than one thing. So they try to do a few things and then when they try to do many things at once, nobody's hearing them out because they're not be able to do even one thing completely end to end. So they're just not succeeding on all those three, four things. So what we do is we identify which one thing is that focus is all about. So we take that person and have honest conversation and say, ''What is that one thing that you want to succeed in this next 6 to 12 months?'' and then make sure that opportunity is available to that person. So, we've had a lot of people move from being, say, in the Analytics team to the Data Governance team. Being in the Data Governance to Data Engineering team. From Data Engineering team to the BI team, so on and so forth. We've had people move around because that was what they really cared for. At the end of the day, when you have an honest conversation, it doesn't mean that you've changed. You can always come back and do something else. So, giving them that option of ''You can go back and do what you did. But you can still try this out'', that flexibility really is something that people appreciate. And given everyone is one ''team'', it doesn't feel like you're moving out of your team. You're just moving within the team, which also makes it much easier.
Jonas Christensen 18:52
Nice. So Kshira, we've heard about your team here, but what does a week in your role look like?
Kshira Saagar 19:00
That's a good question. So, I typically split it into percentage points, so that I can explain to people what I do. So, 40% of my time is in trying to work with the enterprise and the executive stakeholders and the board and working towards that. Trying to help understand why we should invest in the right spaces, because a lot of my job is to try to convince people that these are the right areas to invest in and these are the right problems to solve and this is how we can help your team. So trying to do that. Be it funding, be it executed alignment beats, support that something. That's 40% of my time. Then 20% of my time goes towards managing people. I like to believe people, believe people need to focus more on the people rather than the activity. So we have quite a sizable team. So, we are on 100 people in the business. So trying to help - Everyone has a particular need at a particular point in time, be it job retention or be it trying to find the next career path and stuff like that. So trying to do that is 20% of my time. 20% of my time I spend on actually building stuff because I feel if I don't have my hands on my tools, I'll go crazy. So that 20% is all about either building a dashboard for someone or doing analysis for someone or writing an algorithm or anything that helps. And then I work 10% of times with partners and vendors and stuff like that, so I can really help deliver that, understand what they're trying to bring to the table because we can't just work with our teams. We need to work with partners. And yet the last 10% is trying to learn something new. So that's typically 100% of the week. So 40%, with executive stakeholders and the business, 20% with the team, 20% on the actual delivery and the remaining two 10% on stuff that changes all the time.
Jonas Christensen 20:37
Yeah, so you have a nice mix between typically it's called ''Manage your time and make a time''. I don't know if you've heard that concept before. I personally find it very hard to find time for this, make a time where you're having fun with the technical stuff and not forgetting your old skills that you use to get you to where you are now. How do you manage to do that, when there's so many people pulling your sleeves all the time?
Kshira Saagar 21:02
So, I typically keep an ear to the ground. So when the first IT person I spoke about, somebody always has a challenge that is not being met by the team currently. It's like, ''Well, I want this dashboard. I want this analysis. I want this algorithm''. But the team just doesn't have time. Your team doesn't have time. So, I tried to see which of these can I do really quickly, I've done it before or I know I can successfully pull off, then I tell the team that ''Let me do one for you. Let me pick up something for you''. So I'm picking up something with the team needs to do but it's on the backlog. It's too far away, the business is crying loud about and tried to combine those two and try to pick that problem. And there's always a problem like that. So I budget time in my calendar to make sure that I have time off. So I booked slots on different days of the week to just get some focused time to do this. And when you do that, I then go back to the business and try to explain to them, ''See we can do this only if we had more people who could do this at scale''. And so that helps them understand why they should invest in that area and they can see that we're putting our money where our mouth is and so that's something we tried to do.
Jonas Christensen 21:58
And looking at the ownership structure of Latitude or GE Capital as it used to be called, there is obviously a very strategic focus on data science right from the top because you are owned by a group of private equity investors. And if you look carefully, you can see across their portfolio of businesses, they have also invested heavily in data and analytics across other companies in that portfolio. What does this ownership structure and the executive sponsorship that you must be getting, what does that mean for the data analytics function and remit in your organisation?
Kshira Saagar 22:33
Yeah, so a couple of things, right. So one is the board, which represents the owners, are very supportive of the data programme of what we're trying to do. They're extremely data literate to the point where they can ask a pointed question and try to have a conversation about our algorithmic capability or data governance capability, a data set. So that's really good. So, everyone understands what we're trying to do and ask pointed questions, which makes for interesting conversation. And two, what that also means, therefore, is where you're trying to deliver a capability or deliver value, we don't have to explain why it's being done or what the value is. We just have to worry about how impactful it can be and therefore what can we do to improve it. So the conversations are not about why. The conversation is about what if and so what. So, that completely changes when you have people who understand the concepts, who understand the technology, both in executive leadership level and at the board level. So that's been really good.
Jonas Christensen 23:24
So, I want to just dive into that. So, you're saying that your board members are not just analytics literate, they have a really deep understanding of what data science can do and also how its generated. Is that fair to say?
Kshira Saagar 23:36
That's fair to say and they also come with questions on ''I've seen X happen here. When are we doing that x here and why are we not doing it?''. And so, then you can have a conversation about ''Maybe X works for them. It doesn't work for us?'' or ''Yes, we are always doing that X and you can talk about it in the next quarter''. And so the conversations are never about ''Tell me what data is''. It's about, ''Yep, I understand all of this and I've seen this happen somewhere else'' and challenging us to try to do something better. That's always been really good. So always trying to meet the - Set a higher bar is what they do.
Jonas Christensen 24:03
I think this is just such a fundamentally different setup that you have, because you have that support. When I think about, and I have to be careful here painting everyone with the same brush, but often boards are not that analytics literate, because it's typically in the latter half of your career that you're on a board. And back when you were in senior management and middle management or what have you, as a board member, that person was now a board member, where they were back then the field of Data Science and Analytics wasn't invented yet. So it takes time for those generations to kind of flow through and you're a bit ahead of the curve there, it seems.
Kshira Saagar 24:39
Definitely. To me that one good thing about this role is they've advocated for there to be chief data officer. They've advocated for it to be a function that can deliver change. And so, it's really helpful to have support of both the executive leadership team and the board to deliver this programme.
Jonas Christensen 24:54
Yeah, good for you. I think if you don't have executive support. It's very, very hard to do that things you need to do in data science, because typically it's new and unseen and you kind of have to forge a path in the organisation. So, you just got to have that support. So there obviously is this strategic focus on data science. What are some of these strategic opportunities that you're pursuing and what are tools and techniques that you're using to create data solutions for colleagues and for customers?
Kshira Saagar 25:21
So one is basically improving accessibility to insights. So, we're trying to work on ways where we can simplify all reporting end to end, be it operational or customer or performance, and anybody in the business can come and look at any piece of information that he or she needs to make a decision. So that's something that we're working on. So that's one big piece. It seems like a standard setting that everyone makes. But what I mean by that, it's not just building reports. It's also the building data assets and building a platform, where everyone feels comfortable to come, query for the assets, understand what's where, like a catalogue of metrics, understand how it's defined, who uses them. Everything is available, so people don't feel like they have to ever talk to another person again to understand how this is defined or what it means. So for example, revenue: who defines revenue, what is revenue, how it's defined, what is the right metric, if you want to know more talk to X and without having to speak to anyone through multiple confluence pages, you can get that information. So, we're building a cataloguing tool, centralised reporting platform and a centralised data model. So that's one big area we're working on. The other one that we're working on is a bigger focus on customer. So, that was a key genesis for this role: Customer first experiences. So, be it personalization for customers in terms of their shopping experience, be it personalization in terms of the offers that we can make for them, personalization in terms of the communications we do and servicing we do. So that's another big area we're working on. And the third one that we're working on is building capability, where people can sell some other things like data engineering as a service. Anybody can bring data in without having to get a hired engineer or machine learning as a service. So anybody can build a data science algorithm and deploy it, without having to worry about MLOps and communities and stuff like that. And last, but not the least, a lot of data literacy programmes so anybody can sell some insights. So those are three big areas. That's pretty much our strategy for the next year to try to solve these three big problems.
Jonas Christensen 27:09
And that's very interesting. You talk about the business of personalization, which I think is the next frontier for most organisations that have, call it, an ongoing service relationship with their customers. It's really bringing together technology, data and experience designed to create the future customer experience for the listeners who aren't aware Latitude Financial, and you can correct me if I'm wrong here, but historically was a personal loans and credit card business. So, sort of unsecured personal lending and also now has a ''Buy no. Pay later'' offering, which again you have a lot of transactional knowledge in terms of transactional data history and knowledge of how to use that. But the ''Buy Now. Pay later'' data just gives you an extra edge on the stuff you can see around what the customer's buying behaviour and so on and you can also help curate some of these things. Is that really the holy grail for you and the company in terms of personalization?
Kshira Saagar 28:04
So, definitely personalising somebody's experience with credit and also personalising how they interact with the finances, because right now it's very cut and dry. There's only one way to repay, there's only one way. How do we make sure it's a dynamic way to repay? How do we make sure that everyone benefits from this relationship? You don't have to have high interest rates to pay down stuff. And like, trying to find out creative ways for how people can borrow and also how people can pay back responsibly, so that we cater to the responsible lending obligations is something that we try to work on. So yes to the shopping experience, but also more on the financial health and financial literacy side of things.
Jonas Christensen 28:44
And before you became the CDO at Latitude, you spent three and a half years at the Global Fashion Group, which owns a range of very large online fashion retailers like THE ICONIC, which is the one that's famous here in Australia. What did you learn from your roles there and the fashion industry more generally that you can use in this role?
Kshira Saagar 29:04
So, the one thing I often joke about is in the world of retail people treat customers like their best friends. They understand everything about them. They tried to do everything for them. They go out of the way. But in the financial services industry, they treat customers like accounts. It's an account number and a transaction. Literally and figuratively, it's a transactional relationship. So, my retail upbringing helps me look at the customers in the financial institution as a customer or as a friend and do everything for their benefit and for their experiences. So, that's something that I'm trying to bring to the table.
Jonas Christensen 29:37
So could you give us an example of how you've come in with that approach and how that's changed things up?
Kshira Saagar 29:42
A really good example is personalising the experience with us, right. So in the retail world, if you buy the T-shirt on THE ICONIC, somebody will tell you, ''Well, this T-shirt is great, but why don't you buy this pant that goes well'', not because other people are buying it because that's the style that goes well. So, we make a recommendation that's driven by style, driven by fashion, driven by context, not just driven by, like in Amazon, people who buy this also buys that. Not to make you buy more but help with your outfit. So we complete the outfit for you. So, you can buy a shirt, pants, shorts, a shoe, everything else associated with it. And you can decide what to buy and what not to buy or we can personalise. When you finish any offers, for example, we won't send you any products that are not in your size, not in your case, not in your colour. So you don't have to waste your time looking through hundreds of things that you never were going to buy. But in the banking world, there's no such thing. Everyone gets the same offer. Everyone gets a 10% off or wherever it is. Everyone gets the same interest rate. That's not how it is. If retailers can do it, I'm sure financial institutions can do it and go out of their way because we understand customer needs better. So, how do we do it a needs-driven marketing and moments-driven marketing rather than just a one-size-cut-all marketing is probably one thing that I've seen massively different in the two industries.
Jonas Christensen 30:59
Yeah, I often reflected on banking, when I worked in that industry of being very product driven versus industries like retail actually starting with the customers or the customer walks into the store. And as you said, you can dress them in different ways. It's all about it matching and fulfilling the customer's need to give them the full outfit, whereas a lot of financial institutions are structured functionally, operationally and the way that they also interact with the customers around products and this sort of very siloed product lens of afterworld. But in actual fact, financial services are a means to an outcome, not a goal in itself. And so, I'm really interested in seeing how you will succeed. I have no doubt you will succeed, how you will succeed with this Latitude. Now, Kshira, let's talk about creating a truly data driven organisation. Because that obviously takes a lot more than just producing technically sound data products. So in your eyes, what capabilities must the data analytics functions have to deliver into the business ad what capabilities must the business have to be able to consume the output that you create correctly?
Kshira Saagar 32:10
So I like to think that that's what I call the datasets skillset mindset approach. So one that you mentioned is definitely the dataset problem, right. When I say dataset, not just the data. The data and the tools associated with it. Everyone says, ''Oh, we have Tableau. Why don't you just go and learn it?'' and I'm like, ''That's not how it works'' or we have two X or two Y or platform x, we have everything, people just have to come and write a piece of SQL and learn it. And people don't realise that people who are not in the data space, the moment they see a graph, it just is so off putting for them. They don't want to do anything with it and I've worked a lot more with them. Yes, getting the dataset and the tools is the first step. Once you do that, then you need a massive journey of getting the people on the tools and trying to make it in a way where everyone feels like they can come and use it. So skill set is all about people feeling like they can come and use it and training them to use it in the context of your business. So for example, if you have a dashboarding platform where we can see dashboards, it's not just about, ''Go use the dashboards''. First of all, was about what happens if you use this? If not, that number says 103. What is 103? Is it good? Is it bad? How do you interpret it for your particular division, for your particular business? And that training of understanding it and connecting to daily work, if we don't try to do that - Everyone tries to do skill set programmes like LinkedIn learning or Udemy or Data camp. And it's very generic. So, people can go learn SQL or Python or interpreting numbers. But in the context of your business, if you don't do it, if you're not personalising it for your business, then it doesn't work. So that's something that is needed. And the last one is mindset. And when I say mindset, it's basically people believing that there is credibility in this team. That they can trust them if things are good and bad. So let me give you a good example. You and I, we work in the data space. If we deliver good news, everyone's like really happy with us. They're like, ''Their the best team that we have'' and the moment you start saying, ''Oh, you can't do this'' or ''this is bad'' or ''it's not working'', they're like, ''Oh, we don't have data. We don't have the right people, we don't have the right problems''. And so, how do we build that credibility that people trust us and use us in good and bad, like they do trust doctors, right. So yes, you can get a second opinion, but you still trust a doctor for what he or she deliver. So, that credibility is the last part and that's really hard to build. And so, once you build that is when the organisation becomes truly data driven. So, if you have the tools, it will not be like the people who know how to use the tools and you will also have people who trust that ''This can actually make my life better, even if it's something that I don't believe in right now''. Because over a long period of time, it will definitely add value to my role and my decision making. So those three things are definitely needed to make data.
Jonas Christensen 34:36
So in terms of credibility, what have you seen that is really critical to make that work?
Kshira Saagar 34:41
I think for credibility, it's basically, again, if people don't want to trust it because they feel you're just giving them a solution and walking away, if you're willing to stick with them and say, ''if it works or doesn't work, I will be there and I will be taking accountability for that final decision''. So I'll give you a really good example. So, we wrote this amazing picking algorithm in THE ICONIC. And so, the picking algorithm would tell somebody, a picker at the warehouse, how to go pick items when she gets a batch or he gets a batch. And so, they go pick stuff in the warehouse and so the algorithm was obviously a simple operational research or tool space algorithm. So what it would do was people, it would say, ''Go to A2 now and go to Z25 next'' and so it doesn't make any sense for a human. They're like, ''I don't understand that. It seems too complicated. You're just giving me a random path. I won't use it''. And so what we said was, ''No, trust us. This is the algorithm mapping, simulating like hundreds of thousands of paths for every single combination and coming up with the best combination. We can't explain the math to you because it's complicated. But what we can do is we can walk with you and prove that it's right''. So we have the data scientists literally walk with them for the first couple of days and show them, ''See if you did it this way, it costs X. If you did it the other way, it costs Y and therefore there is value in it''. And then over a period of time, as they rolled it out, we were there at every step of the way. And so once that happens, they buy that but this team is not just there to deliver it and walk away. They're there to implement it with us and take the blame and improve it. That is what I mean by credibility. So it's not just about delivering a piece of work. It's also about them living with it and making sure people get the right support they need through it and that builds credibility. And then the second time you walk in and try to deliver something, they probably won't be as against it. They probably be like, ''I think I can trust them a little bit more''. So, this time I'll ask them the right questions or not have to build, start from scratch. So, that's really good.
Jonas Christensen 36:31
So, I'm hearing in that a lot of empathy for your stakeholder and a very strong focus on getting the outcome as opposed to implementing a technical solution. So, that is definitely something that we can all learn from. Now short history lesson: If we talk about the first corporate executive committees and we're probably going back to before the second world war here, they typically consisted of a financially astute CEO and they would have a small number of general managers who would oversee whole business divisions. So, let's say someone oversaw North America. Someone oversaw Europe and so on. And then it was for the whole of the business of finance and payroll and operations and production and what have you. Then as businesses became more complex over time, we've started seeing all these functional experts become chief executives. So CFO, CRO, CMO, CIO, and so on. And in my opinion, the next 5 to 10 years, we'll see a shift in who sits on that executive committee. And it has to be more and more data and technology led customer experience, we've talked about and I get the sense that your board agrees, which necessarily means that there needs to be space for Chief Experience Officers and Chief Data and Analytics Officers, or whatever we call this sort of work that you do. There has to be space for them on these executive committees. And these two roles combined with the Chief technology officer or Chief Information officer, wherever we call them, will have to deliver the future of data driven customer experience. That's my sort of vision for how things will play out at a very high level. How do you see the evolution of the Chief Data Analytics role in the next 5 to 10 years?
Kshira Saagar 38:16
One thing I see is quite like every other role, every other functional leadership role, as I want to call it, data currently is fully in the limelight for the next five, seven years. There'll definitely be a lot more CDO's coming up or CDAOs coming up and trying to do this. But I think after that, given how commoditize it's getting and how it's all about people making smarter decisions, I think once organisations go past that, like, I keep telling this capability building and assess value delivery from data. Right now, there is the capability which is everyone using the same data sets or everyone using a machine learning platform or everyone using reporting platform and stuff like that is very fragmented, but because we have already seen a lot of consolidation and commoditization of that. So once that happens, capability building will become less and less important and value creation will become more important. And the moment that inflection happens in that, it's all about creating value and making use of the data. It will not be a separate function, but it will be part of every single function. That's how I think about it. So if a business is driving, say, commercial businesses driving change, they need to have data quickly. If a technology part is driving change, they need data to be part of it. And that's how I like to think about it. So in the next five, seven years, the CDO role just like the other older CDO, which is Chief Digital Officer role, will not just be a rule that is separate, but will be part of every single function. And the people who are running data organisations now would probably step up to do more focused work that demand a lot of data. For example, marketing is one area I believe, needs a lot of data experience. Yes, creativity and branding, I'm not denying that. But given performance marketing is all the rage, you need a lot of data and data insight. So somebody who's run a data function will start running marketing, for example, or the customer area for example. That's another area in your data. So that's how I think it will transform from being a data function to being a function that does a lot of data and also delivers a customer or enterprise outcome and then the other parts of it will go into the aspects of business, how I like to think about it.
Jonas Christensen 40:14
And what do you see as the biggest challenges and opportunities for Chief Data and Analytics Officers today?
Kshira Saagar 40:20
So the biggest challenge definitely is buy-in, right? I think a lot of people pay lip service to, "Yes, it's the right thing to do. We should do it" but the moment you say, "Yes, if you want it to work, you need money". And that's where you're stop short. They're like, "We don't have the investment for it right now" or "We can't invest in people or time or technology now". That's the challenge, right? Because you're handicapped, there's only so much you can do with good will and with a handful of people. You need investment. You need time and resources and the industry and the enterprise buying to do it. So that's the biggest challenge I see my fellow data leaders face. It's they have really good ideas. They know what needs to be done. It's just that they're limited because there's always a competing list of priorities and data never makes it to the top. So that's one big challenge. The other challenge obviously in the coming years is going to be about talent. So, you have funding, you have bind, but you don't find the people and that's going to be the next challenge. So hiring the right people, making sure you're building the right talent and retaining them. That's going to become harder and harder in the coming years.
Jonas Christensen 41:16
Now, most organisations are still without a formal Chief Data Analytics Officer and there'll be a lot of listeners to this podcast that have the ambitions to step up into a role like that. But typically, they'll have to forge a path for themselves. What do you think aspiring CDAOs could do to prepare themselves for that top job today?
Kshira Saagar 41:40
The most important thing is to be able to connect what data people do for a living to what the business needs and being that person and trying to explain why they need to do it and the what if and so what of it is probably more important. So I mentor a lot of people who are like Heads of Data Science or Head of Data Engineering and stuff like that and the challenge with them is they're really good at what they do. But the moment somebody comes and says, "Okay, I want to give you money. What will I get there?", they're like, "You will get this technology, this technology, this outcome, this...". They're like, "No, what will I get?" and so that answer of what will you get, why should you do invest money in it, what is the value to the business or the customer, I think being able to explain that, that extra level of why somebody should care about your functional area, if they can do that, that can really helped them really get a seat at the table and also start the conversation on, "Okay, this person understands the commercial reality of it and also the technology side of it and therefore they should be the right people to lead this function". And that's how I think about that particular role.
Jonas Christensen 42:38
The way I think about it is we talk a lot about tools and techniques and one year it's Psse, the next year it's Python or R or Power BI versus Tableau or whatever it is. But the thing that never changes is, the master code is in our case is the English language and you gotta master that and be able to really communicate what it is that you want to achieve and to deliver value to the organisation. So I couldn't agree more with you from what you just said. Now, Kshira, we're coming towards the end of the conversation. Is there anything that you would like to get across or mention that we haven't discussed so far.
Kshira Saagar 42:57
I think it was really good to chat and I've heard your other podcasts and when you speak to the other leaders of analytics. I think the one aspect, everyone talks about having empathy, everyone talks about trying to explain this to people, I think the only part that I would ask all the other leaders to do like you keep saying is pay it forward, and pay it forward in their own ways and not just career progression and stuff like that. But pay it forward into the community and help foster the idea that data does not have to be, like you said, nobody has to do a degree to get into the space. They can literally walk up, be willing to learn and for that to happen, they will have to go out into the community and talk a lot more about it and get people excited. So that's something I'll definitely ask leaders in the area to do more.
Jonas Christensen 43:53
Wonderful call-out. So everyone on the podcast listening in, please follow good advice here and Kshira, you did mention pay it forward and it is something that we mention here on Leaders of Analytics. And typically I ask I will asked, you know as well, who would you like to see as the next guest on leaders of analytics and why?
Kshira Saagar 44:13
So there's this person called Dr. Alex Antic. He is the adjunct professor at RMIT and is also the professor at ANU. And he's also recognised as a top analytics leader in Australia. The reason I really like what he does is he is the bridge between academia and industry. So he runs a lot of these data innovation hubs, where he makes it possible for students to get a pathway into data science and analytics, the one that I spoke about in the past. So he runs all these programmes across universities. He finds the right employers, he finds the right students, connects them. And so, more people can come into this space without having to go elsewhere, find out and eventually come here. Directly, they can finish their uni or as they're finishing up their uni, they can get a paid job and come in. So he does a lot of that great work. And he's also a big proponent of how people can be more data literate. So that's something, he'd be a really good person to talk to on this space.
Jonas Christensen 45:01
Wonderful recommendation and I'm sure we can convince Alex to come on the show. I will definitely be reaching out to him. So, thank you for that. Now, lastly, where can people find out more about you and get a hold of your content?
Kshira Saagar 45:14
So, I typically try to write on my blog kshirasaagar.com So it's my full name.com, because my name is unique. It's so easy to get that domain site, nobody hogs it. So if you just put my name.com, you'll go to a page. I do this weekly newsletter where I put together a top five articles that everyone needs to know and can share with their friends and peers. So that's another thing that I write a lot. So, somebody if they're interested, they can hear more, and I do a lot of speaking. So everything's said on that site. So that's one place you can find more about me.
Jonas Christensen 45:42
Yeah, I did have a look at your website and I was very impressed with the number of speaking engagements you've had over the last three or four years, not withstanding pandemic and other challenges in the world. So well done for continuing to add to our wonderful data community. Kshira Saagar, thank you so much for being on Leaders of Analytics. Really enjoyed the conversation and I have no doubt the listeners did too. And I hope you have a wonderful day and we'll see you soon.
Kshira Saagar 46:09
Thank you so much, Jonas. I appreciate that.
