Sacra Logo Sign In

Former head of strategy at a global on-demand giant on the economics of grocery delivery

Nan Wang

Questions

  1. What have been your key observations in online grocery since COVID?
  2. Dark stores don't have offline footfall. They have an app, but the app is not integrated with any other marketplace to acquire these demands initially, how effective are they with demand generation?
  3. Some on-demand grocery players hire couriers as full-time employees. How does the cost base compare with Uber Eats, Deliveroo type of contractors model?
  4. What trends do you see in basket size?
  5. Given the hyper-local nature of these operations, what determines how capable a company is in scaling the operation?
  6. And in terms of contribution, if we think they can increase AOV to some respectful level, do you have a sense of what kind of steady-state contribution margin we'll be talking about?
  7. Is there anything we haven't talked about that you think would be important to understand for this space?

Interview

What have been your key observations in online grocery since COVID?

Guest: So far what has happened in this space with Uber, Deliveroo, Glovo and all of these players is that they chose to monetize the grocery space by diverting their existing user base the ability to order via participation with like, some of the major grocery chains, like Tesco, Waitrose. They had strategic partnerships whereby they would use stores as trans-shipment points and use that to deliver the last mile. So using their existing fleet and almost hacking the existing system together to deliver things.

That's what happened a couple of years ago and it started moving in that direction. Now that model didn't actually work. Didn't acquire customers in a very effective manner. It didn't retain them for long. The idea is that it didn't form a lot of habits, hence stickiness or AOV wasn't actually predictable in that space.

That model persisted until basically, the entire lockdown happened. So what happened is a lockdown, suddenly forced people to find other alternatives and baskets and order value started really picking up over here. What led to is additional investment in the space to kind of build out that category.

For each of these players, I think people that first started doing dark store in Europe was Glovo. After that Deliveroo started looking at some sort of dark stores alongside their dark kitchens. Essentially, they are using some of that space as inventory. 

But basically, I think starting with Gorillas and Dija and Getir, all of these guys, this wave started towards the end of the pandemic where they essentially said: Hey, let's build this model of dark store delivery. And now there are a good six or seven players, at least in the UK itself that have been funded and are competing in this space.

Largely all of this was happening while the goPuff model was happening in the US unperturbed for all of this, goPuff and Cornershop, there were two businesses that were competing in the US. goPuff basically really blew up and started doing really well. Those businesses were primarily very focused on purely niche baskets. So, alcohol drinks, cigarettes, those kinds of deliveries, but what ended up happening after that was startups started coming in to try to capture that a hundred-pound bucket, order size, which is your weekly groceries.

Grocery basically splits into two things. One is, drinks, cigarettes, desserts or ice creams, impulse buys basically. The other stuff is regular grocery baskets, which in the UK Ocado and Waitrose and all of these guys have been delivering for a while.

What you're trying to do is disrupt that market by making it instantaneous on their own supply chain, in an instantaneous supply network, which basically creates these transshipment points that are much closer to the neighborhoods, that’s allowing you for 10 to 20 minute deliveries. That model has picked up in the last six to eight months. And it's hyper slash almost one could argue overfunded, to the point that there's going to be a decent amount of they're still yet to have anyone prove that demand use case.

And yet to have that consolidation or that understanding of what business model is actually going to survive in which ones are going to die off. So it's a very early stage for that industry.

Dark stores don't have offline footfall. They have an app, but the app is not integrated with any other marketplace to acquire these demands initially, how effective are they with demand generation?

Guest: In terms of localized demand, what these guys are doing right now to solve for hyper-local marketing. There is flyer across neighborhoods and very specific campaigns. There's not a huge amount of marketing that is happening except for maybe Getir and Zapp. But the rest of it is all in the digital space, as well as the hyper-local space.

Essentially, there are decent amounts of, of budget allocated to basically a keyword, the social media marketing, which is very hyper-focused. And then there are flyers and newsletters that are happening in the current form, but broadly the majority of this race so far has been who can set up the biggest footprint in the most number of cities at one point in time using capital.

So it hasn't actually been a rush to acquire customers. It's going to move to a rush to acquire customers. It has been a race to actually deploy supply and deploy logistics, in the last six to eight months.

Essentially the more funded ones, Gorillas are deploying the most. they do have a few dark stores that have basically gone past a thousand orders a day, which is the most that you're seeing in terms of a single dark store.

I think it’s in Germany. I'm not sure exactly what city and it's one store. So it gives you an idea of how nascent the entirety of it is.

Currently like in the UK, for example, I don't think any of these dark stores are doing more than 30 orders a day, 30 to 50, and some even lower. It's lower than one would imagine purely because these companies are essentially right now just fighting that race to be relevant and to exist.

So they need to set up that supply chain, they need to kind of manage the logistics of actually delivering produce because that can become an expensive affair. If you're doing transhipping points to each location and they try to build up apps and basic technology, which currently runs off of Shopify kind of like stack.

Eventually, the idea would be to focus a little bit more on improving unit economics in those warehouses and marketing in that space.

Some on-demand grocery players hire couriers as full-time employees. How does the cost base compare with Uber Eats, Deliveroo type of contractors model?

Guest: Usually, food on-demand fulfillment cost is somewhere between £3.5 -  £4.5 for, I think, the bigger players, which means you will need a higher basket size on a high end of delivery fee to adjust this as a blended fee.

For hyper-local grocery delivery, on a per order basis probably burned a lot of money right now. They also have these people as full-time couriers, a few of the companies have that and a few workers on the gig economy system. The unit economics of the industry is not standardized anywhere at this point.

It's not a gig economy system. They have X number of employees that are doing deliveries. It's slightly different from what happens with Uber eats, where they have contract workers doing delivery. A lot of these dark stores are hiring couriers as full-time employees because they want to have more control over the entire supply chain because they're trying to promise standard 15 minute delivery time.

It's too early to determine whether it's a good idea or not. At least two or three of the players that I know have full-time employees. If you look at, Getir, for example, even their bikes to get to your bikes are actually branded. That is the direction that that space seems to be going in. I don't know if that's there to stay or how that's going to change the gig economy system.

In the gig economy system, the number of orders or the highest amount of frequency in an hour or an hour with higher efficiency, we'll keep the cost base low because he's a gig economy worker, which technically means that, if I'm doing two orders per hour, if I had to fulfill an hourly wage as a benchmark, I essentially, need lesser income than if I'm doing one. On the other side, when you're looking at a fulfillment logistics that is completely owned, then you're essentially kind of like you're looking at a fixed wage, or an hourly wage, which is not something that scales up and down. So the more orders you get in, the more efficient you can be. You can start building batching. You can do a bunch of things after that which is very far away for this space.

I think essentially at this point, it's fairly safe to assume that the actual footprint of order fulfillment or the average cost of order fulfillment, is basically extremely higher at this point because your actual number of orders is very low. So you can't get those economies of scale that you get on a per order basis.

What trends do you see in basket size?

Guest: In terms of the actual basket size, these basket sizes are currently fulfilling only impulse orders. So if you look at Zapp, I think they focus a little bit more on the impulse purchase segment itself, like cigarettes and alcohol and ice creams at night and late-night cravings. 

Dija and, I think Getir and all of these guys have groceries as well, and they're broad base focussed. This doesn't mean that one is not going to move in the other direction. I think the idea is everyone's waiting to figure out what works out in terms of margins because with fresh produce there are the storage costs, there are wastage costs, things like that. You have to be very precise with what you want.

There are also selection issues. So if a customer is looking for tomatoes and they don’t have them, that becomes a problem, so essentially that space has more promise to have the higher AOVs. Eventually, if you can be super reliable, a Morrisons or a Tesco's, you can have a £100 average basket size there or at least a weekly frequency as such.

Or you can basically go the other way and you'll have a 20 pound, 15 to 20 pound kind of basket size, which you basically have, but with much lower margins than food delivery. Food deliveries are about, I think an £18 to 25 basket. Essentially what you ended up having is a higher frequency, but much smaller basket size kind of business when you're looking at impulse purchases.

If you have a higher-order frequency, your fixed costs get amortized over more orders. So whatever generates more orders, and on a hyper-local basis, you'll be able to divide your fixed cost divided by more orders. So essentially the greater number of orders you have sometimes, your variable will stay high enough, but your fixed we'll drop that basically becomes important in hyper-local space.

Given the hyper-local nature of these operations, what determines how capable a company is in scaling the operation?

Guest: It's down to the team, that's actually doing it. So these companies have expansion teams. Build these businesses and the playbook, the best practices, and they use that to market and get people on local center. It doesn't cost a lot of money to set up a single it's about a hundred thousand pound investment to set up a single dark store. 

I think it's down to the success of how many people are in that catchment area. What kind of an audience that looks like; What kind of marketing they're doing the variable to acquire customers cheaply and retain them. I'm guessing, in the end, it's going to become a consolidation phase and maybe Getir is strong in the north of London, but it's not in the south of London. So they might sell those fulfillment centers to somebody else. That consolidation phase on a city level would start to come in, maybe they would just shut those centers down. I don't think these guys are going to compete and swap centers at this point. It's more likely that they will shut one down because it's not a huge cost.

And in terms of contribution, if we think they can increase AOV to some respectful level, do you have a sense of what kind of steady-state contribution margin we'll be talking about?

Guest: When you're delivering fruit and veg, as well as, non-perishable items, your margins vary across the board. So it actually tends to come down to what people are ordering. How do you influence that and how do you increase AOV? Those are the two main levers that you might have.

Let's get this customer to buy something that gives me a higher margin and let's get them to buy multiple things to build the baskets. Now, both of these heavily require me to understand how customers work in order to build products around them.

This space and all of the competitors, and I can say this as a general statement, are not anywhere there at least two years further out than they need to be, to actually have technology that can start to do that kind of stuff. These are capabilities that Uber and Deliveroo of the world realized much later. 

So there isn't a lot of control over this space, which is why I don't actually have a number for what a good contribution margin can look like because there's no winner so far.

Glovo has been doing this with the partnerships model. So I don't think they are a good benchmark either. They are all severely in the red right now and they're going to try to get out of it.

Is there anything we haven't talked about that you think would be important to understand for this space?

Guest: Have seen the cycle played out, starting with ride-hailing early in my career, and most of the logistical businesses have been pretty similar, I think each cycle or each iteration of funding for things has been faster and has been more aggressive.

This time around, I mean, I think we're looking at, at least since October, November, some of these businesses were incepted only in like July of last year. That means the space is hyper-funded.

What's going to actually happen is that winners and losers are going to be picked off. Out of the six or seven players in the UK, there's probably going to be two or three left in the next few months.

I think Dija will probably manage to raise. You've got Getir, Zapp... Some of those players are going to consolidate. It seems like in the UK, Zapp will continue since they have a hundred million in the bank.

Disclaimers

This transcript is for information purposes only and does not constitute advice of any type or trade recommendation and should not form the basis of any investment decision. Sacra accepts no liability for the transcript or for any errors, omissions or inaccuracies in respect of it. The views of the experts expressed in the transcript are those of the experts and they are not endorsed by, nor do they represent the opinion of Sacra. Sacra reserves all copyright, intellectual property rights in the transcript. Any modification, copying, displaying, distributing, transmitting, publishing, licensing, creating derivative works from, or selling any transcript is strictly prohibited.

Read more from

Read more from

Read more from