In this episode, Will Sprunt, Principal Data Scientist and former CIO at Deliveroo, talks about the pandemic’s impact on the food industry and the innovation that has come because of it, the nuances of leadership in tech, and much more.
In this episode, Will Sprunt, Principal Data Scientist and former CIO at Deliveroo, talks about the pandemic’s impact on the food industry and the innovation that has come because of it, the nuances of leadership in tech, and much more.
--------
How you approach data will define what’s possible for your organization. Data engineers, data scientists, application developers, and a host of other data professionals who depend on the Snowflake Data Cloud continue to thrive thanks to a decade of technology breakthroughs. But that journey is only the beginning.
Attend Snowflake Summit 2023 in Las Vegas June 26-29 to learn how to access, build, and monetize data, tools, models, and applications in ways that were previously unimaginable. Enable seamless alignment and collaboration across these crucial functions in the Data Cloud to transform nearly every aspect of your organization.
Learn more and register at www.snowflake.com/summit
Steve Hamm: [00:00:00] So we'll, it's great to have you on the show.
Will Sprunt: It's uh, it's great to be here. Thanks for thanks so much for having me on.
Steve Hamm: Yeah. Hey, a lot of our listeners probably have not heard of deliveroo, so it'd be great. If you would start off by describing the company and its businesses.
Will Sprunt: Absolutely. Um, so our mission is to build the definitive online food company. Uh, we want to be the platform that people turn to whenever they think about food. Uh, we're well known outside the U S um, as a delivery company, like door dash for us listeners or a matriarch, and then China or Rappi and Latin America.
And, and w w we operate in, uh, 12 different markets of over 115,000 food merchants, more than a hundred thousand riders and millions of customers across the, uh, across the globe. Uh, we started out in the UK and we since expanded, um, all over Europe, all over Asia and Pacific region. Um, and despite being a global company, we really think about this as a, uh, you know, neighborhood business delivery [00:01:00] started in the London neighborhood of Chelsea back in 2013.
And, you know, we've. Proven track record of that global expansion through the hyper-local lens, you know, from the very beginning, uh, we recognized that to succeed. We need to get the proposition right. Neighborhood by neighborhood. So that's something that's really, really stuck with us. Uh, but the other thing which we think is that support is, is, is a real focus on food.
Um, a lot of other providers in the space operate primarily as logistics companies. Um, and you can see a lot of companies trying to become delivery systems for lots of sorts of products. But, uh, we think food is special. It's really hard to do. Because people have a real emotional connection to what they're eating and every second counts and getting a good experience with that.
It's like making sure your ice cream gets to you before it's melted or making sure that your burger is still piping hot. Um, does that means to get it right? We've got to be obsessed with that experience and specifically. Uh, you know, we've had really rapid growth, but we think we're only just getting started.
Uh, bring the food category online is, [00:02:00] uh, a huge market opportunity. And the way we think about it is really simple. There's there's 21 times that people are eating most days, breakfast, lunch, and dinner, seven days a week, uh, sorry, Ms. Weeks, breakfast, lunch and dinner, seven days a week. Um, and right now, less than one in 21 of those transactions take place, takes place online.
Uh, and we're working to change.
Steve Hamm: So a huge market opportunity. Locally and globally, that's really great to hear.
Will Sprunt: Absolutely.
Steve Hamm: Yeah. Now over the past year and a half, we've had this terrible thing that the COVID-19 pandemic and it, you know, it's still, it's still affecting businesses. People, you know, individuals, communities, how has it affected your company, its operations and also its strategy.
Will Sprunt: Yeah, that's it. It's such a huge, um, industry defining moment. And I think, you know, before, before we can talk about deliver, actually we have to talk a little bit about, uh, kind of a wider context. Uh, well first a terrible tragedy, um, but [00:03:00] really a category defining moment for a lot of different industries.
Um, but in food, in particular, we've seen it rock the industry from, from top to the bottom, uh, overnight restaurants across the global close to, uh, know, forced to close. Um, and, and, you know, but those traditional restaurants had to think of new ways to adapt and survive. And that's true, whether you're trying to, uh, open restaurant in, in, in New York or operate in London, uh, or you know, it in Hong Kong.
Um, and the, the common thread we saw throughout all of these different places was massive innovation. Uh, you know, obviously we, we really believe in our company. It's something we do an awful lot of, uh, but it's been amazing to see. Actually be industry adapt and, and find new, new ways of operating in in that time.
Uh, so then zoom in, on, on, on us, you know, uh, just going to say innovation is, is, is the best response to that sort of disruption. And, you know, we were w we were really focused on finding new ways to serve, [00:04:00] uh, the three sides of our marketplace. So what can we do what's best for customers? What can we do?
What's best for writers? What can we do for the best for our restaurants? And then in addition, what can we do, which is supporting the community. Um, you know, there was a point where. Across Europe, like close to 80% of restaurants, uh, you know, uh, closed back in early 2020, and, and the majority of, uh, of our offering just wasn't there.
You know, we could see our partners really struggling. So first we wanted to take action to them, to, to, to, to help them out. Uh, we brought you new technologies, then you guide them through the new rules where we had, uh, an element of industry expertise. Cause we'd done it a number of times, our partners.
You know, provided, uh, you know, really, really clear guidance of how to stay safe. Um, amped up the amount of additions are delivery only kitchens, um, and made sure that they could get food to the customers. Um, we also want to support, um, our national health service in, in the UK where w where we were based, um, and who were working really, really [00:05:00] hard to keep people safe through the, through the pandemic.
Uh, as we set up an initiative to deliver a million meals, frontline, NHS. Uh, but yeah, coming back to the, to, to the, I guess the business operations, you know, uh, what we saw was that delivery, um, became an, a lot of cases, restaurants only channel for trading, uh, and, and, and as such became a lifeline. So, so it was something that specifically for independence was, was one of the main ways they could kind of get through this really tough period.
And, and it was really fantastic for, uh, to. Uh, you know, helping to grow a business at the same time. Um, you know, it turned into this huge opportunity because as everyone was quarantining, it, it just accelerated the adoption of online food delivery. Probably. You know, a couple of years, uh, because people who, you know, hadn't thought about trying us, we're now in a situation where it's like, wow, I needed a way to get food.
I needed to wait to, you know, uh, um, have a, have [00:06:00] a treat or a reward at the end of like a really hard week's worth of work. And we've seen that through in our results, just, uh, you know, seeing, uh, 88% year on year growth in the second quarter of 2021, that one just finished. Um, and in that same period orders in the UK and Ireland grew 94% year on year.
Um, which is pretty fantastic because, you know, we. Have been in this market since 2013, we've been established, we've got quite a lot of penetration, you know, in, in, in some ways you could kind of look at that business and think this is mature at scale, but it basically doubled in, in, in, in, in the course of just a year.
Steve Hamm: no, no, no, that was great. And I really thought the whole thing of. You know, going from being kind of an addendum to a, of the restaurants, to being a, a real essential strategic partner and using data to do that is really fascinating.
So thank you for that information. Um, I want to just switch back to deliveries kind of business and, and expansion plans, stuff like that for a [00:07:00] minute. I know that the company went public last year. And how has that affected it? Its operations, its ability to expand its strategy, all those kinds of things.
Will Sprunt: Yeah, absolutely. I mean, first off, we're, we're really proud to have this number London stock exchange. It's the place we call home. Uh, and, and we're extremely confident about longterm value of yet. Um, it's it's, it's, it's, it's a pleasure of having, uh, you know, grown up in this market to, to, to, to now be now, now be part of it as a, as a public facing company.
Uh, the IPO itself was, was always about setting up the company for the term, um, which we believe we have done. Uh, we're capitalized really, really well. As a result of the event, we've raised a billion pounds. And we continue to gain market share in the UK and in our other markets, uh, around the world. Um, but at the same time being a public company means we have responsibilities to our shareholders and to report our results in public markets.
So we can't just invest or expand for the sake of pure growth without a [00:08:00] view to the long-term profitability of those actions. We have to be really sure. Um, it's, it's a good use of the cash that we've been, we've been trusted with. Um, Yeah, we've got to, we've got a lot of options on how we could invest.
It's easy to jump to thoughts of like a brand new stuff, brand new country's brand new products. Uh, but there's, there's loads of potential opportunities. We have to weigh that up against, you know, should we go deeper into our existing countries? Should we increase the footprint of delivery additions, which is our, our turnkey real estate solution for restaurants, which.
Some people might call them cloud kitchens or dark kitchens or whatever. Uh, and then how much should we invest in that greaser opportunity? Which, which we talked about a little bit in, in, in, in, in the previous section, uh, we've also got, uh, our plus program or membership, uh, service, which, you know, requires, uh, uh, you know, careful thought and, and, and, uh, use of use of that funding as well.
How much of that cash should we go back and creating additional, uh, you know, value layers for our consumers? The one thing that I [00:09:00] can definitely say that we've we've we've we've announced just is it's really unlocked our ability to invest more in technology and put additional resources into growing our team into a world-leading organization.
Uh, we've recently announced we'll be adding 400 new tech roles, uh, in a team, uh, from everything from engineers to data scientists, designers, uh, data engineers. We're talking about snowflakes today. Um, And adding the capability here is, is really key for us to be able to scale and drive up the growth for platform.
It, it means, uh, you know, being able to build better products, um, and, and, and having those benefits really stack on top of each other and next financial way. Um, it means having better data capabilities for us, all of our teams, so we can make higher quality decisions and, and it means opening up new avenues for the business.
Steve Hamm: Yeah. So you just talked about, uh, greatly expanding the technology capabilities, the manpower or human power of the company. Hey, so let's get into your role a little bit. I know. [00:10:00] You were the CIO for awhile, but now you're the principal data scientist. What does that transition mean? What's your day-to-day job like?
Will Sprunt: Yeah. Um, as anyone who's worked in a startup, uh, um, or a scale up, um, I guess will relate to them. I've been at delivery for about four and a half years now. So I've had a lot of different roles in that time. And a lot of that is depending on kind of what the, uh, the most pressing problems of the company are at any given moment and, and also where, where kind of my skills fit best.
In, in some ways, I don't think my role has changed an awful lot. It's it's uh, about using our data and the best possible way. And it's, uh, you know, to be honest about putting out fires and a lot of cases, uh, the CIO role was, was quite interesting because, um, you know, it was, it was, uh, you know, something we hadn't had up until that point, but we realized there was, there was a clear gap in the organization.
Um, you know, we, we didn't really have a lot of central governance over our [00:11:00] third party application stack. Uh, there wasn't a lot of coherent thoughts around, uh, you know, how we use technology and parts of the business, um, to improve productivity outside of via the engineering and tech team. Know, there wasn't, uh, the most thorough governance on things like our devices or our kind of our hardware strategy around the business.
So, so the role itself was actually about spinning up a lot of those teams and getting to a place where we, we, we had maturity, we had a strategy and then finding the right leaders to kind of take those, um, you know, take, take, take, take those departments on and take them from a, not just a zero to one, which I think I'm reasonably good at, but you know, one to two to 10 to a hundred, um, You know, the, the, the switch to principal data scientist has been, has been really interesting as well.
Um, data science has always been something which has been the, I guess, the common thread throughout my career. Uh, you know, having been, uh, in working with data, I guess, for the past, Whatever it is now 15 [00:12:00] years. Um, I'm not even sure if I was called a data scientist to start with, I think people call me an analyst or, or a strategy person.
Um, but the, the common thread is always, how can we use data in the most effective way? Um, the, uh, the transition actually happened, um, last year in, in, in, in quite a. Uh, a stressful and challenging situation for us, the company, um, you know, as part of the response to COVID initially we had a, like, quite a challenging period where, um, you know, as we said before, 80% of restaurants were closed, although it's turned out really well in the end, there's a huge amount of uncertain.
Um, in, in, in, in, in, in, in the world at that point. And one of the actions that we took was actually looking at, um, our organization and saying how, uh, much do we want to invest in, uh, uh, invest in the evidence that we have. And we had this really tough, uh, period we went, but we had to make redundancies. It was, it was honestly [00:13:00] one of the hardest leadership challenges I think I've ever had.
Um, but one thing. Came out of it as an interesting, uh, opportunity was, was I was at the time, uh, you're managing still quite, quite a large team. Um, not quite as large as, as, as kind of a happiness at the peak of the CIO role where I think it was about 120 people. It was kind of beyond my, uh, you know, my Buddhist to remember everyone's, uh, personalities particularly effectively.
Um, and, and, you know, When making all these cuts to make, make, make, make everyone's jobs harder. And some of what we needed was not necessarily more management, although more management is obviously really important, but the way to still get impact and still get value out of our data and have more firepower to.
Uh, the teams. So I talked to that. This is over a lot with, uh, Dan, our CTO. Um, and I decided I could actually move back into a role, um, as an individual contributor and, and that's been great, obviously, it, [00:14:00] it coincided pretty well with our preparations for IPO. And there was a whole lot of work which needed, uh, doing which involve data, but it was.
You know, in a need for confidentiality or, um, you know, sensitivity around it. But, you know, since then, it's, it's, it's meant getting to move on to problems with re really high leverage and, and, and ones which typically cut across different, uh, different groups. Um, I think. Interesting part of, part of it is actually getting a chance to maybe define an alternate model of, uh, of leadership.
Steve Hamm: Yeah. Yeah. So you're you, you put out buyers essentially and big ones. You don't put up a little fire. She put out the big ones. It sounds like very cool. You have had a very interesting and varied career. So far, you talked about doing this for 15 years. You started right out of college or actually in college, you started a number of food and beverage businesses.
You've also worked for a number of tech companies, including Zipcar. What are the most [00:15:00] important leadership lessons that you've learned along the way? And how are you applying them deliberately?
Will Sprunt: I mean, uh, I think, I feel like I should start off with, uh, the food businesses. Some sometimes I talk about them as, as kind of my fake MBA, uh, because. You're going into something, um, you know, like that straight out of college. Um, the first thing I did was I saw all of my friends go into like grad programs or, or, or, you know, big industry things that looked, uh, kind of unappealing.
Um, so I, I went and made chocolate for two years and, uh, in retrospect I think it is, it's one of the best career choice I've I've ever made. Um, you know, it got into a situation where. I would need to do everything, uh, uh, with which sounds. Kind of, uh, arrogance, I guess, but at the same time, it's, it's, you know, it's, uh, it's almost a forced position.
Um, you know, if you're the owner of a small business, if you're not doing payroll, no one's saying payroll. Uh, [00:16:00] if, if you're not doing advertising, no one's doing advertising, not, uh, uh, you know, planning out, uh, you know, your, your, your cash flow. No one is doing any of those. And I think what it taught me really really quickly, it was, um, not that, not quite that nothing is hard, uh, but nothing is impossible.
If, if, if, if, if you apply yourself to it and actually get into the stage where you are, um, maybe consciously incompetent, you know, how much, uh, how out of your depth you are, but you can still string your way along, just about, uh, is, is super valuable. And it's something that stuck with me a whole lot as a, uh, Both a leadership lesson, but also as a kind of, um, you know, general work, uh, policy.
Um,
Steve Hamm: really interesting. You kind of, you kind of build confidence and competence at the same time incrementally, but through challenges. So that's really cool.
Will Sprunt: absolutely. Um, I mean, it's, it's, you know, it's, [00:17:00] uh, Yeah, I think having come through a situation where I was forced to do so many of these different things, it's taken the fear a lot out of, out of trying something, trying something new, um, as is a lot of, uh, you know, people who, who, who, who I can see early in their career.
And they've maybe specialized the whole load and think, okay, I'm, uh, I'm an accountant. There's no way I can deal with any of this like data science stuff. And actually it was like, if you learnt a little bit, you'd, you'd be fantastic. Or, um, you know, I'm a data scientist. I don't really have the strength in this kind of user research background.
It's like, actually, you know, you could take a, run, a test and your skills in data science would probably make you, make you pretty applicable and pretty strong at that second thing. Um, I think it's really important to build a kind of a T-shaped set of skills where you have depth in your area of specialization.
But you can, you can, you can, you can cut across a lot of different areas as well.
Steve Hamm: Right, right. What about in the tech industry? What kind of lessons have you learned [00:18:00] there?
Will Sprunt: Yeah, I think. You know, inside tech, uh, it's a little bit hard for me to say which the lessons were attacking, which work, uh, actually working in larger companies. Um, but getting through to the idea of, of actually, uh, uh, perspective and empathy and communication style was, was, was really big. Um, yeah. And everyone comes at problems with very different perspective.
And, uh, obviously from, from my background, I, I naturally gravitate towards data and there was a whole lot of, um, awkward meetings I had earlier in my career where it's like, well, the data is saying this. Why, why wouldn't you believe me? Or I don't know what I'm saying wrong to, to, to, to make you see this and actually understanding there, there are so many different strategic motivations which can come through from, from any of these different angles, uh, where.
Data at the end of the day is only, uh, a thread, a thread of a story. And in tech, and especially in, in, in, in tech in the [00:19:00] last five years, uh, it will only get you, it will only get you so far. Uh, you have to have that flexibility and, and, and constant learning mindset to be able to apply yourself to new problems as they come up.
And sometimes some of those things are. Uh, gainable by, um, you know, I have non quantified beans or by, uh, you know, by, by, by, uh, you know, feeling or, or, um, no, let me say that again. Sometimes some of those things are, are only achievable by, uh, by feeling or by group creation where actually, um, Which decision you make is not even necessarily the most important thing.
The most important thing is that you've all made the decision together and, you know, you've decided on a direction you're striving towards it. After that you can only get incrementally better and remain flexible in your approach to kind of keep on, keep on overcoming challenges. So they come up.
Steve Hamm: Yeah, that's interesting. I think those kinds of skills are [00:20:00] really undervalued within large organizations, but they're absolutely critical and we see it again and again, how, how groups, whether they're five people or 500. Can be dysfunctional. If, if leadership from top to bottom, doesn't have that kind of sense of empathy and a sense of, you know, reading, reading the room, kind of, and responding to the individuals and the group the way they need to be responded to.
So I think that's really cool now. I know that delivery was it's kind of, um, a cloud native company. So it didn't have to go through all the pain and agony of moving the data from on-prem into the cloud. So, so you, you, you, you missed out on all that fun. So tell me when and why did you start using snowflake?
Will Sprunt: Yeah. So we started using a snowflake back in the end of 2017 start of 2018. Um, yeah, it was, uh, You know, as you were saying before, we luckily didn't have the decision [00:21:00] process where we'd have, we'd had to move from on-prem into the cloud, to be honest, I'm not sure how we would have managed, um, you know, ever managed scaling or managed to have the sort of growth that we had.
If we had to also manage our architecture from a hardware perspective in the background as well. Um, Coming back to, to the, to the snowflake, my migration, uh, we were having issues for our existing, uh, data platforms and specifically the data platforms, which, um, our, our end users would interact with. Uh, the, the things that we were saying were, were, were problems with concurrency problems, runtime, and, and ultimately it was coming down to how accessible could we make data for everyone in the car?
Uh, we used to have a pair that we call the Monday morning madness where, uh, probably about a thousand different people would descend on the data warehouse and start running queries through, uh, our, our, our BI platform. Um, and you know, it would just generate this huge spike in, in, um, you know, the amount of, of, [00:22:00] of, of, uh, uh, queries that we were exposing the data warehouse to.
And, you know, we still wanted to have that data availability. We thought, actually, this, you know, this, this event of people. Really going off the data and being really excited by that and wanting to self-serve and wanting to have this democratizes access is not something that you choke off and you go, okay, well, how can we make this work with a smaller amount, a smaller footprint, or how can we, uh, you know, truncate the amount of data which, which people have access to in order to make it.
You know, pass through our systems correctly. Uh, and, and we select a snowflake just because of the ability, uh, to take that concurrency and to take that shift in demand up and down, depending on what it was people wanted at, uh, you know, right at the minute.
Steve Hamm: Yeah. Yeah. The elastic cloud. Right. And you only pay for what you use. Those are, I think are, are really keys. Yeah. Uh, alright, well, let's get down into the gritty of it. Could you describe a couple of the most important ways you're using the stuff like [00:23:00] technology and what kind of results you've gotten?
Will Sprunt: Yeah, I it's a little bit hard actually to, to, to even pick out, uh, the most important ways because really technology and data underpins almost everything we do in the business. It's, it's so fundamental to our logistics and it's brought down our costs massively over the years. Um, it's fundamentally iterating on our products as we experiment on pretty much everything we can and it means we can prove the value of all the changes we're making.
Um, it's fundamental to our decision-making across the whole business. Uh, to make sure that it's not just tech, but everyone, um, is data first and data driven in their approach. Um, to give a, a really contained example. I think we can, we can take a look at, uh, uh, additions, which we talked a bit about before, uh, which is our, our turnkey real estate for, uh, for restaurants, uh, AKA cloud kitchens or dock kitchens.
Um, so from right from inside, You know, we've got a, a fun data problem. Where, where do we decide to even put the sites? Because we actually have to open these [00:24:00] things up and, uh, uh, you know, make sure that we can serve customers from them and there's loads
Steve Hamm: W w actually we'll take a step back. I don't think people probably, they don't know what Deliveroo additions are. So explain it very basically. Okay.
Will Sprunt: definitely. I, uh, uh, so yeah, to say, um, additions, it really represents what we think of as kind of the operating system for restaurants. We buy the site. We kit out the kitchens. We put in everything that people need from specialized pizza ovens to wok burners. Uh, you know, we, we make it so that if a brand wants, wants to set up a new site, they can just walk in there from there.
All they have to supply are the ingredients and the chefs. Um, and it's been really transformative for some of the, some of the, the relationships of our strongest brands, because we know that there's a huge advantage in getting a brand, which has been performing really, really well in one area. [00:25:00] To drop into one of these, uh, you know, kitchen sites.
They don't have to open up a brand new restaurant, but suddenly, uh, we've got shake shack in a place which didn't never had it before. Or, you know, we've got three uncles. Um, my favorite barbecue meat restaurant place from around the corner. Uh, And you're accessible to a whole group of new people who would never have it before.
Um, from the restaurant side, it makes a whole lot of sense as well because the cost of doing one of these openings is dramatically lower than it would be, uh, um, you know, setting up, uh, uh, you know, a whole brand new site and, and, and working out to bake and even make any money off delivery in the area.
Steve Hamm: Yeah. Yeah. So how do you use data to make these places work?
Will Sprunt: So really it comes from, from, from, from inception all the way through to, uh, operation, um, you know, first, before we can even pick, uh, what it is we're going to do. If we have to find a site and that's informed by data, you know, where do we think we have, um, opportunities either, because there is a kind [00:26:00] of a lack in the consumer offering there we're missing.
You know, a great burger restaurant to a missing great Indian restaurant. And we think there's a good place. Uh, but also from an operational perspective, it's like, where do we think, um, is going to be a good site, which is going to deliver efficiently, uh, mean that we can, uh, operate, um, you know, very, very, very.
Good deliveries. Um, and, and, and, you know, choose sites based on that, um, secondarily then you have to pick the, the brands which are going to be in those sites, because actually in almost all of these sites where wherever subscribed. So, um, you know, finding out who is the right, right partner to go into that restaurant is also a data backed decision.
And it's about, you know, finding the right. Between the, uh, you know, the brand and the civic area, uh, then if a decision about how we choose to surface it to customers. Um, so we have these sites, they go into all the rest of them. Uh, you know, our restaurant list and now [00:27:00] they're gonna be part of what the consumer can kind of, uh, look at and choose from it, it on the website, how should that rank for different customers?
How should that be presented for deaf individuals? Um, and then we also give data back to, uh, you know, the partners and we can, we can give them a lot more information, a lot more, um, Grand narrative when you can, for people who are operating in their own condition kitchens, just because we have more oversight on onto the, uh, the operations.
And that means we get, you know, uh, you know, more concise, uh, preparation times of meals, more accurate predictions of how long things are going to take tends to make mean that our, our kind of writers spend less time waiting at those sites tends to mean that customers get a better experience overall.
And we can measure all of those things and output and feed that back into the rest.
Steve Hamm: Yeah. Yeah. It's really interesting about how tightly you're operating with your customers now. And you're almost like interwoven completely in some of these situations. So, so data sharing is obviously very important. [00:28:00] It's a key feature of snowflake. So does deliver a plan on doing a lot of data sharing with business partners, kind of through the snowflake platform, or is that something you do separately?
Will Sprunt: Uh, it's something we use the data from snowflake for, but tend to do it, uh, um, tend to do it separately. Um, so, you know, we, we, we obviously have a whole bunch of, of, uh, you know, products, which, which restaurants and partners we'll interact with in the course of, uh, in the course of their operations, whether that's, you know, a tablet which sends them, uh, information on which, which orders are coming up, or whether that.
Uh, uh, sales reports coming through, um, from account managers on, on, on, on, on, on a regular basis. And, you know, a lot of activators is going to source from snowflake, but we tend to tend to have our own front ends, which we kind of put, uh, put on, on the top. Um, but yeah, as you mentioned before, like this is it's something which, um, You know, is, is a really important feature in these sites, but is, is actually, um, just [00:29:00] generally part of the operating model, which we want to have with our restaurant partners.
Um, and especially in the course of the pandemic, it's meant that we've worked, you know, even even closer with them than, than, than, than before. Uh, you know, we've got regular insights and key stats that we can send them on performance. Uh, you know, In what, like during order stats, uh, we can send them, uh, to let them know about the customer that they're serving.
Um, and we've got a restaurant hub, which is that like saying one of his products we've, uh, we've built that they can, they can log in and have a look at their performance time. Um,
Steve Hamm: that's really cool.
Will Sprunt: and actually one thing I'd love to talk about as well before we, uh, before we move on. Is, uh, we've also been developing what we call our, our, our signature products.
Um, so this is actually providing delivery technology and logistics capabilities directly to as key partners. Uh, so the restaurants themselves, like they have their in front end, but the logistics and the food delivery, um, itself is, is, is being backed up by, [00:30:00] uh, delivery technology. And of course, you know, Uh, you know, that there are partners in, in, in, in, in more ways, more ways than one there, they, they, they become customers of that.
Steve Hamm: Interesting. And you make money that way you charge them for these services.
Will Sprunt: There are two key things from, uh, from we get from the surface. Uh, one is for sure, it's a direct products, you know, we get, we get, um, you know, direct benefits in, in, in cashflow. Uh, but the other is really strengthening upon.
Um, you know, we think it's really important to have the partners here who are on signature on our core platform. We think it's really important that they are able also to continue to serve their customers in the way that they want to and having, you know, signature really strengthens both sides of it.
Steve Hamm: Yeah. Yeah, that's cool. Cool. Hey, no discussion of, uh, of data scientists would be complete without a discussion of algorithms. And, and I understand that delivery has created a collection of core analytics algorithms called Frank. [00:31:00] I'm curious why Frank, but tell us what Frank does and why it's so important to the company.
Will Sprunt: Cool. I'll get to the name at the end. We can talk a bit about that. Um, but yeah, if you think about, uh, ordering a takeaway yourself, um, you know, you think about a restaurant. Tell them, what three do you want? And the mate deliver to you? It's the traditional model? Uh, it it's the model, which, uh, you know, grub hub, uh, for the most part does in the U S just connecting a single, um, a single customer with a single restaurant and letting the restaurant workout how they want to serve so fat.
Um, but for us, uh, there's an extra side to the market where we have writers who are separate from the restaurant and they're going to deliver the food. Um, and it's the relationships between those three sides, which makes it, uh, Complicators interesting, fun. All of those things. Um, if you have one writer it's really easy.
Steve Hamm: too, right?
Will Sprunt: Exactly. Exactly. Uh, if you have one rider it's, it's, it's really easy. Um, you know, the food's ready, you just send them, uh, [00:32:00] we've got loads of writers and loads of customers. So, so you've gotta make that decision on who should pick what, and you don't just have to decide the right pick for that one. Uh, you have to make sure that every order gets the right rider, uh, when you've got thousands or tens of thousands or hundreds of thousands of possible interrelated combinations in, you know, every single minute, uh, you can just send the closest rider, uh, when it's, uh, when, when, when the food is ready.
Uh, but that might actually slow down other orders that you have more. So you have to look at the whole picture to make your picks. Um, and as you quite rightly mentioned, if you make your pickup. Uh, the network of writers as a whole becomes way, way more efficient than any single writer or any single restaurant can never be on.
And, you know, the algorithm which makes those sessions is, is, is what we call Frank.
Steve Hamm: Interesting. It's really a massive systems and systems modeling problem that you're solving. Really cool.
Will Sprunt: Exactly.
[00:33:00] Steve Hamm: Okay. So Frank, why the name.
Will Sprunt: Oh, yeah. Uh, so if you know, uh, uh, an old, old sitcom in the U S called, um, taxi with, um, uh, Danny DeVito and Ted Danson, I think it is. Uh, yeah, Dan to be, so it plays a character called Louie. Who's always on the phone, dispatching out taxes and it's like, Hey, we've got a new, we've got an order here, go here.
We've got an order here, go here. And the very first version of a ride re right dispatch album algorithm is called Louie because it would just be the thing which would send out, uh, um, send out riders all the time. Uh, when we had a new version. Which was, um, trying to do this in a much more sophisticated way and looking at the holistic network all together, we had to come up with a good name for that.
And so naturally we looked at Dan Debussy and you know, who he's been more recently in a set, calm and went to always sunny in Philadelphia. And there's Frank.
Steve Hamm: Yeah. Interesting. So your system ultimately is a very sophisticated display.
[00:34:00] Will Sprunt: Exactly, exactly.
Steve Hamm: Right, right. Cool. Hey, I want to ask you to put on your visionary cap here, look out five years or more one year. That's not enough. How do you see the cloud-based data analytics? You know, all these technologies affecting business and even society.
Will Sprunt: um, I think the, the first trend to think about is, is kind of that democratization of data. So, you know, we, we talked about our Monday morning madness where you have a thousand people trying to run queries on a database since like, okay. That's pretty, that's pretty different. Um, you know, you only have to think, uh, 10 years back and anyone who could code in SQL was like a rarity in the organization.
You think, you know, 30 years back and it's like not even 20 years back. And it's like, it's pretty rare, but someone knows how to use macros you 30 years back. It's pretty rare, but someone knows how to use XL. You four years back. It's rare that someone knows how to use a computer. So you do the same trend, but extrapolate it forward.
And I think the amount of. Familiarity and, [00:35:00] and the expectation that people will have for familiarity in business, um, is going to change dramatically. Um, what, what, what I think is really interesting about that is, is it then changes the sorts of problems which, uh, people think about as data problems. You know, there's a, there's a, uh, you know, a consistent model where it's like various hard data.
Therefore we can apply, you know, data science skills to it. Uh, but actually you're seeing more and more, um, how do we, how do we extract the, uh, the kind of soft ideas, things like sentiment analysis or things like brand position or things like, uh, you know, social, um, you know, social, social information and turn that into stuff you can use for.
And I think the, the, the most concrete example of this is, is I hope I'm going to come down in, in, in, in actually hardening up some of the stuff we've talked about or some of the stuff which the welders talked about around ESG, uh, [00:36:00] you know, you see everyone from investors to, uh, uh, you know, people looking for their first job saying, I really don't know.
What like the environmental and the social and, and, and kind of the governance, uh, situation of an organization or a, you know, a country looks like. Um, but then when you try and extract information from it, you either get these slightly weird, like checkbox exercises or these really, uh, abuse things saying like, this organization has been rated two stars out of five.
And actually there's a real opportunity with the awareness around data and the willingness to apply it to hard social, proper problem. To specify what we mean by social impact, whether that's, uh, you know, equality of opportunity, whether that's, um, uh, uh, you know, amount of suffering generated per, per dollar of capital, you know, whatever the, whatever the objective happens to be.
Uh, and we can start transitioning these things from, I believe, or I think about this to [00:37:00] actually, we really want to target this. And I think that that could be, that'd be pretty cool.
Steve Hamm:
So how exactly will, does data help people appreciate wine?
Will Sprunt: So I think this comes back to something we were talking about in the previous section where, uh, the application of data to soft problems actually makes them more defined and more targetable.
And at least for me, more enjoyable. Uh, so, uh, um, drinking wine is, is, is, you know, purely, uh, uh, Yeah, qualitative experience, you have a sip and it's delicious, or it's not delicious, and that's kind of what you get from it as a first impression. Uh, but if you talk to a wine expert, um, you know, they'll often say things like, oh, I can detect, uh, you know, suffer flavors of this and aromas of, of, of that.
And as a, a kind of lay person, you may initially not, not get any of that. [00:38:00] Um, but what's interesting. The actual act of description and the education of people on kind of what memories it is there Woking or what tastes it as they are kind of guessing actually changes your experience and your perception of the thing, which Shaw your, uh, uh, your experience today.
Um, another one of my personal hobbies is actually perfume, um, falls very much into the same space. I think it's really interesting because, uh, As a novice or you can get, is it smells nice or not nice? Uh, or it smells maybe like, uh, you know, uh, Woody or smells, uh, fresh. But when, when you are able to break those components down into more, uh, defined notes, Actually, it gives you more appreciation of what it is that you like.
And don't like, um, and I th I think that that same transformations happen in a bunch of places. Like people can talk, uh, you know, for hours on end [00:39:00] about like different cuts in clothes or different shapes in bags. I don't seem to get any this, uh, but like the, the availability of a language actually changes the, uh, the perception of the products.
And in turn, I think. Um, you know, increases the, the appreciation and the levels of nuance that you can get in it. So I think wine is especially interesting because you have applications like, uh, Vino, um, which, which are doing a fantastic job, educating people about these, about these nuances and these, these flavors and doing it in a really unpretentious and straightforward.
Uh, um, and, and it comes back to like accessibility of this data. It's taking something that it's very, uh, you know, very exclusive knowledge or perceived as very, uh, like, you know, Pumpsie actually, that's the wrong word, but it's something that's perceived as, you know, very grandiose or selective or, [00:40:00] um, exclusive knowledge and putting it into hands of people and actually making it way more accessible and accessibility.
In my view, almost always leads to good things. We'll have better wine producers, better wine. Appreciators
Steve Hamm: well, I'm not quite getting the connect, the dots here. So in this case, the precise descriptor words are the data in question. So how does analyzing that data? Lead an individual to a more, a better wine experience or to buy the right wine.
Will Sprunt: Right. That's a great question. It like, to me, it comes, it comes down to, uh, um, you know, you've, you've actually extracted data points from it, from this and are able to make a decision. Uh, if you look at descriptions of wine bottles, you might see something described as, as. Uh, you know, fresh and fruity. And if that's all you get, that's not a very detailed description of that wine.
If you have. Fresh and fruity layer, one of data, you've looked it up on an app. And actually it will [00:41:00] tell you later, too, is, uh, you know, this is a, um, you know, this is a semi-warm from this region in the wild valley, uh, um, this altitude and you know that by connecting the dots, that means certain things about the characteristics of that one.
Okay, cool. That's extra information. And if you also have the awareness that, uh, you know, in this case, fresh means, uh, uh, Really really dry and, uh, citrus notes. And you've had the awareness to all that. You know, think about wine that way previously you can then connect the dots. So it sits creasing, additional accessibility for data.
Uh, not everyone's gonna want to use that data of course, but it gives, gives anyone a chance to, uh, to interact with questions and to, you know, to build it, their, their knowledge of, of, of wine or perfume or whatever they want.
Steve Hamm: Right. That's very good. That's. All right. Well, we'll, it's been great talking to you today.
It's really been fascinating and I mean, [00:42:00] really a lot of information. Thank you so much. I mean, to me, Interesting part was when you talked about additions, I had no idea you guys were doing it. And then, then to, to hear how, I mean, these are essentially data-driven kitchens. You know, we talk about data-driven business.
So really cool. I mean, using data, everything from selecting the place to the menus, to the who, who you put in one. The roots and all that kind of stuff, optimizing it. It's really fascinating stuff. And I think it's also very accessible stuff for, for the podcast listeners. So thank you so much for being with us today.
Will Sprunt: No, thank you very much for, uh, uh, for having me and, and, uh, new, glad, glad we got to chat a fair bit about wine as well.
Steve Hamm: I'm gonna go drink some, you know,
Will Sprunt: , that's an excellent plan. Okay.