What Growth Tactics Look Like In A Two Sided Marketplace: Lessons Learned At Bird (with Declan Bond Schweitzer, VP Growth at Albert)
Matt Bilotti: Hello, and welcome to another episode of the Growth podcast. I'm your host, Matt Bilotti, and I am really excited today to dive into growth tactics in a two- sided marketplace. So Declan Bond Schweitzer, who is a VP of growth at Albert, or the VP of growth at Albert, is joining us today. Declan, thanks so much for being here.
Declan Bond Schweitzer: Hey. Thanks, man. Nice to be here.
Matt Bilotti: Absolutely. So Declan spent about two years at Bird doing some growth stuff, and we were talking through his experience there, which requires some pretty unique approaches to growth that I found really interesting, that I thought would make for a really great podcast where folks can listen. Even if you're not in the industry, it's just really, really interesting to hear how it's approached in that sort of marketplace. So we're going to dive into what growth looked like there, some of the learnings, some of the lessons, and we will go ahead and jump in. So Declan, if you can give the quick intro on yourself, we'll go ahead and start asking some questions.
Declan Bond Schweitzer: Yeah, as you said, VP of growth at Albert now, formerly at Bird, and then, before that, I was at Headspace on the growth team. Before that, in a former life, I was a professor of philosophy and jumped ship from academia to come into the tech world.
Matt Bilotti: Well, welcome. We are excited to have you here. It sounds like you've landed yourself in a pretty good spot. So why don't we start with the framing of the Bird business model, maybe for folks that aren't all that familiar, and how the growth team thought about its approach, and the growth team's focus within the picture of that business?
Declan Bond Schweitzer: Yeah, totally. So Bird is a kind of two- sided marketplace. Almost, I think of it as a two- sided plus marketplace. It's not really a three- sided marketplace per se, but you have this kind of physical object that's also kind of in the mix, the actual vehicle itself, and then you have the riders, demand side, and then the chargers, the supply side, the people who keep the Birds maintained and charged and deployed on the streets. The business of Bird is really connecting riders and vehicles, and doing that by having chargers pick them up and charge them, and put them back out on the street in strategic locations.
Matt Bilotti: I don't think we've said the word scooter yet, but they are scooters.
Declan Bond Schweitzer: They are scooters, yeah, so apologies. I refer to them as vehicles. We always refer to them as vehicles in the office, so yeah, but they are scooters, effectively.
Matt Bilotti: Yeah, makes sense. Was the team split into, this part works on getting more riders and this part of the growth team works on getting more chargers? How did that break down?
Declan Bond Schweitzer: When I started, I joined about nine months into the business launching. And at that point, yeah, we really thought of things in that supply and demand sense. We had a team focused on chargers, a team focused on riders. Over time, we started to coordinate efforts a little bit more. So my original responsibilities were specific to the rider side and incentivizing riders, and then, over time, we started going after charger incentives and thinking about ways we could get chargers to do more tasks, to charge more Birds. In a sense, the reason for that is because, obviously, as a two- sided marketplace, you want to scale up your supply and demand at the same time. But also, just from a tech perspective, the levers that you use in growth are often the same for both the rider and the charger supply and demand side, so CRM, push notifications, email, owned channel communications, and then paid acquisition. These are all the same tools used for different purposes, but it made sense to consolidate. The actual tools and the tech that you use, the levers that you pull for growth, on both the demand and the supply side, are often the same. In our case, at Bird, they're definitely the same. Paid acquisition, performance marketing tools, CRM, push notifications, email, owned channel communications were often the same, so it made sense to consolidate both sides of the marketplace under one growth team.
Matt Bilotti: So one of those core levers that we had talked about, which your team thought about all the time, was around pricing, right? So pricing for the riders, maybe we're looking at discounts in given markets or flash sales, or whatever it might be, and then there's some pricing stuff that I'm sure also comes into play with chargers. So why don't we start with the rider side, and how your team thought about and approached pricing and unit economics?
Declan Bond Schweitzer: Yeah, totally. So I think there's two embedded questions here. So one is about base pricing and how to think about the actual everyday pricing, and then there's another one that's specific to incentives and getting people to ride more. So when I joined, we were really focused on top- line growth, both rides, revenue, and pretty quickly, we shifted our focus to unit economics and profitability, which led to a ton of price testing. And one area in particular that I focused a lot on was contra revenue, things that take away from your gross revenue, in e- commerce, for example, things like sales returns and stuff like that. In our case, one of the interesting things that we noticed when we were taking the fine tooth comb and scrutinizing our unit economics was around contra revenue from credit card fees. And obviously, super low margin, low transaction cost business credit card fees can be a big part of the overall margin. One of the things we did, besides just testing different price points, was, we tested two different models of payment, one being pay as you go, each ride has a start fee,$1 to start, and then a per minute cost after that, and you would pay all at once at the end of the ride, to a model where we actually just charged you upfront fixed costs, like five or 10 or$ 15. You would use your balance to take rides. We called it a Bird wallet. And the question there was, are we going to lose out on people converting into riders because they don't want to pay$ 10 for their first ride, versus, do we make that up in the better unit economics that don't have 30 cent transaction fees for every single ride, but rather just one transaction fee, that it's spread across more rides? And I think the exciting and interesting part of it was that, not only did we find that we could make up that cost in the credit card fees and that the conversion difference was relatively low, but also, people who filled up a Bird wallet were way, way more likely to come back and actually take a second ride. This is, obviously, I think, pretty intuitive if you've loaded up your wallet with some credit and you take a$5 ride, you've got$ 5 leftover, you want to take advantage of that value you've committed. So that was a super happy outcome for that experiment, and we ended up scaling the Bird wallet. Most of the players in the space, Lime and other competitors, also ended up going with the wallet. And then, I guess, another thing to touch on there was that... having a reason to communicate with someone, which was the wallet itself and the balance. So we did a lot of work, just making sure that people were aware that they still had a balance. If they hadn't ridden in a while, we would make sure that their balance was clearly highlighted when they opened the app, or we would send them a message, push or an email, at a timely moment to try and keep that top of mind for them, so that we would keep Bird top of mind, in general. So that was, yeah, another way in which that helped us a lot. I think on the other side now, in terms of incentives and actually discounting our product, like I said, when we shifted our focus to unit economics, we really severely curtailed that and tried to be really, really tactical. The idea was that we didn't want to incentivize rides when we would get that ride anyway, so didn't want to have almost any cannibalization, which meant that we had to be pretty tactically thoughtful about when we were willing to give someone a discount. And one of the things that I think was a really successful, first of all, but interesting test, and thing that we implemented in the end, was the idea of getting someone to take a ride when they had shown a demand signal, what we call demand signals. You can imagine, if a rider were to open their app, we could see that on our end with an event and we could track if that rider then went on to take a ride, and if they didn't, why was it that they didn't? They put their hand up and said," Hey, I want to ride a Bird. That's why I opened the app, obviously." If that didn't happen within 10, 15, 20 minutes, we would trigger an incentive, maybe a$ 1 off your next drive or something like that, to try and capture that demand signal. Later, when I was talking about that push notification that we would send, I realized that it mimics the behavior of an abandoned cart message where someone's clearly said," Hey, I'm interested in making this purchase," they don't follow through, and so you follow up with them because you've got a very clear demand signal to work with.
Matt Bilotti: Yeah, I love all that. There's so much to work with there and unpack. I think, going back to your first core point around the credit card transaction fees and the Bird wallet and getting people to pay upfront, I've never operated in that sort of high volume, low cost transaction world before. And now, it makes so much more sense to me as to why all these video games, mobile games, all that, have these upfront, buy your credits type system. I'm sure that, not only is it because they're avoiding those transaction fees, like you were saying, but also, very much, what you're talking about is that you get all this added benefit of, they're locked in, there's commitment bias, they've already put the money towards it, you get to continue marketing to them. It seems like there's generally only upside, and from what you had tested, only a low downside to conversion rate, if any, in the first place.
Declan Bond Schweitzer: Yeah, totally. I think, on the conversion rate stuff, again, it's very context- specific, but you can imagine that, when you're thinking through from the perspective of the user, walking up to a Bird for your first time, it's a likely story that you're either alone, but there aren't many other vehicles or other competitors around, or you're with some friends and you all have them. There's a lot of momentum in that buying decision already. So in that context, I think, yeah, we really didn't see a massive drop- off in conversion because people wanted to ride a Bird. That was what they opened the app for, for the first time, and the barrier to entry. We did a lot of testing on what the right price point is. Is it a$ 5 or a$ 10 or a$ 15 minimum top- up for your Bird wallet? But you have to think through from the user perspective, is that decision one that you're pretty close to making, regardless of what the payment scheme is? And like you said, if it's all upside, because there's not a huge drop- off from the conversion perspective, then that, yeah, makes a lot of sense.
Matt Bilotti: Yeah, and one of the other things that I would love to hear about is, you're running these incentives to make sure that there's no cannibalization rate, you're not doing them upfront, but you're doing it when you already have some sort of signal, that there's a potential for somebody taking a ride. When you were doing these price relief, discount type experiments, were you generally doing it across all markets, just scaled to the cost of that market, or was there always one base market that you would test in, like the San Francisco market, you run all your tests there, and then you roll them out to other markets, or... How did you think about isolating those tests and rolling them out? Were they across everything or were they in specific places first?
Declan Bond Schweitzer: Yeah, this is a good question. Unfortunately, I would have loved to have lived in a world where we had a representative market that captured what we would see elsewhere, but the reality is that, and this is an interesting part of the Bird business as a whole because it applied to every part of the business, not just rider incentives or charger incentives and things like that, but the topography and geography of the market would change, very drastically, the way that riders behaved and the way that we adapted to those specific market needs. I'll give you an example. In Atlanta, which is a relatively large market, they have this thing called the beltway, which I guess is a long, walking, biking path that goes around a large park. I haven't actually been there, so I don't know, but that specific geographic feature of Atlanta changed the whole composition of the rider behavior there. And with caps on the number of vehicles that we could have, a lot of focus was placed on making sure that we were able to supply the kind of people that were on the beltway. And so, incentives that we ran in that market would be adopted very differently than those in Los Angeles, where a lot of our vehicles are concentrated in certain areas, since it's such a sprawling city. Because of that, we didn't have a single market that was like our Petri dish. And also, we were really methodical about statistical significance. We ended up testing a lot over broad ranges, so nationally, or even internationally. We would break things down and look at them, both in aggregate, and then at the market level. So certain incentives that were profitable at the nation level or the global level would be super problematic and non- profitable at the market level, and we would adapt the next version of the test to try and focus in on those markets and configure things a little differently. That said, the pricing model of$1 to start, or one currency unit, Euro, Pound, et cetera, to start, followed by a per minute cost, meant that we did have a little bit of a uniformity to the way we thought about incentives, the$ 1 start fee or the fixed cost start fee. So for example, in Poland, it was six or seven Polish zlotys, which is close to a$ 1 fee. That was something that we experimented extensively with, not having one discounting it, waiving it, because I think, structurally, it made sense to consumers to think like," Oh, I don't have to pay anything to start upfront where I normally would," and that's a kind of pricing psychology lever that we used a lot.
Matt Bilotti: Yep, so you had one base foundation for pricing, and then there were just so many markets and so... I mean, so many variables and so many different markets. It sounds like a fascinatingly fun and complicated place to operate, but that's awesome. We're just talking about one side of the marketplace here, then there's this whole other side of getting people to show up and charge all of these Birds that are all over the markets. Can you talk to me a little bit about how you incentivize the fleet of chargers to continue to show up and do so, and how you match that with the demand that's in that market?
Declan Bond Schweitzer: Yeah, totally. So chargers, just to give the high level, chargers, obviously, what they do is, they pick up Birds when they're out of battery, they use a special app, they can identify and locate the Birds, they go, they pick them up, they bring them back to, usually, our garage or maybe a facility, they basically plug them into the wall for a couple of hours, and then they're ready to be deployed again. And so, the name of the game was really getting people who had large capacity, could charge a lot of Birds at a time, and also wanted to be regular chargers, that were making this part of their routine. A specific area that we found we needed to solve for was these spikes in demand, so you can think like, a long weekend, for example. How do you get a bunch of people who are charging maybe 20, 30 Birds a day or a week to suddenly scale up their charging capacity? And so, obviously, one lever is going to be pricing and incentivizing people with pricing. We would pay out, usually, at a per Bird level. We would pay X dollars to charge this Bird and deploy it again in a strategically designated spot. On a long weekend, when we expected a surge in demand on, let's say, a Monday when we would normally not have that, we started by testing, hey... We would message our charger community and say," Hey, we're going to pay you one extra dollar, let's say, for each Bird that you charge," and what we found was that people responded to the incentive, but not really in the way that we would expect. They basically would charge the same or maybe slightly more than they normally would, and they would basically just get one extra dollar for each Bird, which was tough on our margins, obviously, because$ 1 is meaningful in that business. And so, instead of just doing the$1 incentive, we started making the incentives based on hitting certain thresholds, so if you charge X number of Birds, you'll get an X dollar bonus, and if you charge two X Birds, you'll get 2. 5 X bonus, et cetera, et cetera. I think that method turned out to be way more incentivizing for, not only existing chargers who suddenly would charge way more Birds than their normal capacity, but also is really good at getting people who hadn't charged in a while, maybe who had turned, to come back out and charge Birds for a specified period of time. Again, from a pricing psychology perspective, when you say like," Hey, we'll give you one extra dollar for every task that you do," it's just not a very exciting offer or number.$ 1 just seems like not a big deal. But when you say like," Hey, if you charge a hundred Birds this weekend, we'll guarantee a$ 500 payout," or something like that, that's a big number, and that's really exciting for people. And that's something that they really rallied around, so that was a super important and super interesting learning. The other thing that we really focused on was, like I said, the habitual chargers, people who became regular chargers. Again, going back to this incentive and getting people to go out for one weekend and do a big haul of charging, we had this concept internally of making space in your garage, so using this one big, incentive moment to immediately change your routine and make space for a certain new activity, namely charging Bird scooters. And so, we tracked people who took advantage of these incentives to see, if they took advantage of an incentive on weekend one, did their charging capacity in subsequent weeks increase and stay steadily raised? And that was actually the case, which was super exciting. We would be able to get people to change the way they go about their week to include charging Birds as part of that, so that was super exciting too.
Matt Bilotti: Love that. It's a whole other world of growth approaches for that side of the market. So one other thing that I want to circle back to around pricing is something that you mentioned when we were talking before the podcast, is how commoditization of pricing in a market like where Bird operates has a pretty decent impact on how you think about the growth incentives and levers, both for the chargers, because I presume that these are people that could also just pick up and start doing Uber, or maybe they're also doing Lyft, and so you're competing with their time and the money they can make on that platform there. Plus, as a rider, I look around and say," Well, I could use something like Scoot, if it's in that market, or one of your competitors at a different scooter rental platform." So how'd you think about that, and how did that impact how the growth team approach things?
Declan Bond Schweitzer: Yeah, so I think it's worth noting that, on the charger side, charging Bird scooters is a sufficiently different kind of activity that I don't think we were really ever competing that heavily with the ride sharing business, the key difference for chargers from drivers being that you pick up the Birds and plug them into the wall, and then you don't have to mind them for awhile. It's a kind of set it and forget it, and then go out and deploy them mentality that I think draws people to that, rather than being in your car for every minute that you're on the clock. So there's that. As far as charging for other companies, I think the incentives and the threshold guarantees, if you charge this many Birds this weekend, or if you charge this many Birds this week, that started to help de- commoditize the work. It wasn't as effective to split your time between charging Limes and charging Birds if we were giving incentives for hitting certain thresholds, so I think that side was probably less difficult to solve for. But the rider side is very difficult because, obviously, we took a lot of pride in the vehicles and the scooters that we designed in house at Bird. I'm biased, of course, but they're definitely superior. But from a rider perspective, not only is that not always clear, but whether or not they're so superior as a scooter ride that you're willing to walk an extra 500 feet or what have you to go get a Bird rather than go pick up a Spin, or a Lime, or any of these others, that's a much bigger ask. One of the things that we thought a lot about was the thing that de- commoditizes a scooter when all else is the same, is your proximity to it as a potential user buyer. So obviously, again, going back to the demand signals, we knew, when someone opened their app, how close they would be to a given Bird, or to the nearest Bird, let's say. We didn't know, obviously, how close they were to any competitors, potentially, but we did know that, in general, the app open to conversion- to- ride funnel was weaker when you got further and further away from the nearest Bird. And so, as a result, we tested a lot of different proximity- based incentives. So if you open your app and you're more than a mile away, for example, from the nearest Bird, it's probably a harder conversion funnel than if you're 10 feet away, and so you need some kind of incentive price to go and make the trek to go get that Bird. And so, we did find success with that. The other thing going.... not just capturing demand signals, but actually trying to predict demand and be more aggressive, let's say, in our going after users was leveraging the Bluetooth technology within Bird scooters that would connect to the app and your phone and then allow us to push to you, a message about riding Birds. And this was an interesting use of the actual hardware technology, where people would... We were able to connect to Bluetooth on your phone, and as you passed a Bird, we could send a push notification to you, saying like, hey, you're within a hundred feet of a Bird. Why not take a ride? This was a controversial experiment. It definitely was very taxing from the tech side, and it was hard to really nail down the perfect timing for when a user was near a Bird, but also in the frame of mind to ride a Bird. So a lot of people in growth talk about right message, right time, right user, and it was just a really good, valiant attempt, I think, at getting that right. But it's really hard. You can imagine driving your car past a Bird that's on the corner of a sidewalk and getting a push notification saying, hey, you should ride a Bird right now, and it's completely the wrong context. So that was an interesting foray into trying to be more on the kind of pushing to people rather than pulling people in once they gave you the demand signal. And yeah, we didn't scale that program a ton.
Matt Bilotti: Yeah. The Holy Grail of it is, can you figure out the moments in which you can flag to somebody that a Bird is available before they open that app? And yeah, it sounds very tricky. I can't quite imagine how you would nail that down and know that they are actually in a state of mind, or going towards a location that makes sense for them to grab a Bird. It sounds like an ongoing challenge for that team.
Declan Bond Schweitzer: I mean, to be fair, there are definitely moments that it's more straightforward, right? So if you have a big game at a stadium nearby and you have placed a bunch of Birds, knowing that people are going to get out of the game at some time and want to ride Birds back to their car, or avoid the traffic of the parking lot, or whatever, those moments are relatively straightforward, but you don't really need the Bluetooth connectivity acrobatics to really just send a push notification when the game is over and say, hey, you should grab a Bird, or something like that. There are moments where we were able to leverage that push mentality rather than just pull, but that specific use case with the Bluetooth connectivity was a tough one to land. Also, worth noting that different geographies, again, have a very different profile. So European cities behaved very differently because people tended not to be in cars, so when they were near Birds, they tended to be on foot often. Just varied results from those experiments.
Matt Bilotti: This is awesome stuff. I really love learning about all of this. It's not something I've operated in. It's fascinating. There's so many variables to it. I work in B2B software and it's pretty straightforward what your variables are, but Bird and that sort of marketplace is a whole other level, so thanks for talking us through it all.
Declan Bond Schweitzer: Yeah, absolutely. Thanks for having me.
Matt Bilotti: Of course. Before we wrap up, anything else that you feel like you wanted to touch on, that we didn't get a chance to cover?
Declan Bond Schweitzer: No, I think that's it. Yeah, like you said, Bird and the scooter marketplace is very complex. We didn't even talk about the entire chain of logistics related to owning and operating a fleet of physical vehicles and how that plays into all this as well. We talked a bit about unit economics at the beginning of the show, and I think in most places, at least, that... Well, maybe that's not true, but I think it's definitely true that unit economics are relatively stable often in SAS companies. And we had different unit economics, market by market, and so what a profitable incentive or a good tactic for incentivizing chargers was in one place would be completely different, because just the logistics and the cost structure of a certain market might be very different. And so, yeah, it's just immensely complex and nuanced.
Matt Bilotti: Yeah, that sounds like it could be a whole other podcast, for sure.
Declan Bond Schweitzer: Definitely.
Matt Bilotti: Cool. Well, thank you again. I really appreciate you joining here. For all the folks listening, there are plenty more episodes with other amazing people. Hit that subscribe button, go check them out. If you've got any feedback at all, my email is matt @ drift. com. As always, I really appreciate you listening in. I know there are thousands, or an unlimited amount of things you could do with your time, and spending it here, listening to this is greatly appreciated. So thank you again, and Declan, thank you for joining. And I will catch you on the next episode, if you're listening. Thanks.
Declan Bond Schweitzer: Cheers.
Bird is a classic example of a company with a complicated two-sided marketplace that achieved explosive growth. While spending two years on the growth team at Bird, a rentable scooter company, Declan Bond Schweitzer helped execute growth strategies and tactics in different markets on both the supply & demand side of the business. In this episode, we talk about some of the core lessons learned with tons of great examples of what they did on the team there.
Like this episode? Be sure to leave a ⭐️⭐️⭐️⭐️⭐️⭐️ review and share the pod with your friends! You can connect with Matt Bilotti on Twitter at @MattBilotti & @DriftPodcasts