Robots deliver dinner and profits

Welcome back to the Barclays Brief, it's Patrick here. Now imagine it's the end of the week. You order a takeaway, and when it arrives, instead of a delivery driver ringing your doorbell, there's a small little robot waiting outside. Or even a drone hovering over your back garden carrying a pizza. Now, I know that might sound like sci-fi, but the rollout of autonomous food delivery is happening fast.

So to help me understand how this tech led future is evolving, I'm joined today in the London studio by our Internet Research Analyst, Guillaume Galland, who has spent the last few months travelling the world to get a closer look at this technology and has just published a fantastic report for clients on the theme. Guillaume, thanks a lot for joining me on the Barclays Brief.

Hey, Patrick. It's great to join you on the show. I know you previously talked to Ross Sandler on robotaxis and AVs. I'm super excited to talk about autonomous delivery. There's a lot to say.

Yeah there is. And we've been talking a lot about physical AI on the Barclays brief in the last 3 or 4 months, because we're kind of entering this strange future where I could finish work, get a robotaxi home, come home to see a humanoid cleaning my house or doing the ironing, and then be able to order a delivery and have it delivered by a drone. What's going on with the autonomous food delivery and what have you seeing in this market?

So, yes, food delivery is going through a pretty fundamental change today. It’s really the last leg, which is changing. So if you think about it, in the past, people associated food delivery with the rider, human delivery. Today, in several parts of the world, you can actually have your order delivered by drone or ground robot.

Yeah, it's very strange. I remember when we had the episode talking to Ross Sandler about robotaxis, it was very concentrated in San Fran. But I know you've travelled the world looking at this technology, it's not the same here, is it?

It's not just the San Francisco story, I think it's global. We've been tracking this tech for quite some time into different markets. We travelled to Helsinki, Dublin, Dubai, Los Angeles. So the penetration rate varies by geography. If you think about LA, it's more around 1%, even less. If you look at Helsinki in Finland, the penetration rate is one of the highest in the world – around 10% of orders being delivered by robots today.

Okay. So I want to talk about why that is. But first of all, set the scene for our listeners. When we talk about autonomous delivery, people often lump it all in together. But as I see it, there are two very clearly distinct types of robots. You've got drones that fly with your delivery, and you've got sidewalk robots that are kind of like big, packed lunches with wheels. Which one is best at delivering food to your door?

I wouldn't say one is better than the others. It's more about the mix itself. So we're seeing two emerging buckets. One is drones and the other is ground robots. Starting with drones, it’s pretty impressive because the delivery time comes down to five minutes. And how it actually works is that the drone hovers over your roof, lowers the food in your back garden, and I think this delivery mode is more suited for suburbs and longer distances. You have like start ups, private start ups in Ireland like Manna Air delivery, doing the work here. “But you also have ground robots which are better suited for very central, dense areas like central London.”

“So obviously you've been around the world seeing these things. I'm yet to see one of these in action. I've never seen a drone flying with, a food delivery. Nor have I bumped into a robot on the street. But I'm always interested in how humans interact with these physical AIs, what normally happens when you're in LA or when you're in Helsinki, how do humans interact with these robots? “

I'm pretty sure it's coming fast to London. You just have to wait a few months or maybe a year.

“It varies a lot by geography, to be honest. So when you think about LA, which is a very nascent early market, I'd say the interactions people, do notice the robot out there. They take pictures, take a selfie. But it's quite impressive how it fades very quickly. So if you look at Helsinki, where as we were saying, 10% of orders are being fulfilled by robots today. People actually barely notice them. They just walk past. It's just another delivery option. “

And I think the reason behind this is essentially the investment that the delivery operators have made to make those robots friendly. They actually have facial expressions. They could smile at you. They also have names like Sarah or John, which makes the whole experience much more personalized.

Ok I don’t know if you chose on purpose, but Sarah and John, the first thing I think about is the Terminator films in the 80s and 90s, which is all about the rise of the robots. But anyway, we need to talk about economics for a moment. Why is capital flowing into this space right now? And when you break it down in practical terms, can you help our listeners understand the value proposition for the big groups here, because you've got consumers that we just talked about the delivery platforms and the restaurants, how do they all benefit from the usage of these robots?

So maybe starting with the consumer side first, the experience is simply better, more reliable, more secure and faster. And it's also pretty cool experience, right?

Then if you look at the delivery platform, first is on the unit economics, as you said, you're getting rid of the labour component, which is inflationary through the years through time. And today if you look at best in class markets, autonomous delivery is around $3 to $4 cheaper than human delivery.

Okay. Can you put that into context? How expensive is human delivery in terms of unit economics?

Sure. Yeah. So in the US human delivery is around $9 to $10. We think that ground robots there is around $5 to $7 today. This could go to $1 long term.

Wow that much?

Correct. Yeah. So you're getting essentially rid of the labour component. But it also scales with efficiency, utilisation and we think that a ground robot could actually perform 3 to 4 orders per hour. Drop rate goes up. Whereas for human delivery the drop rate is around 1.5 to depends on the batching.

What about the drones?

So it's a good question. We've been to Dublin. We're sitting at the back of McDonald's looking at the drones take off, and during the peak hours the drop rate, it was around 5 to 7 per hour.

Wow, that's a massive increase for the, the standard delivery drivers. What about the restaurant companies then? How do they benefit? And also help me understand why they don't just disaggregate the platform companies? Why not just buy a fleet of robots and cut the platforms out altogether?

So on the restaurant side, from our discussions with partners on the ground, it feels like it helps the daily flow of operations. It's just much more predictable, reliable. Also, the consumers say it brings some sort of innovation for the brand image. Thinking about the disintermediation risk, a large restaurant group could actually buy a fleet of robots, but it would be practically super hard for them to make it work. Because the key issue is scale and utilisation. If you think about it, food delivery, platforms actually aggregate the demand, but also the supply, which is super fragmented today.

We also get the flip side of that question why don't delivery companies create their own marketplace? And the answer is pretty similar, right. The supply side is hugely fragmented. The platforms also own the traffic and the demand. they can push the right order at the right time. So I think in the medium term, we're not seeing any risk to platforms being disintermediated.

Okay. So we're not seeing that disintermediation risk. What about the other hurdles that stand in the way of the roll out of this technology? There must be quite a few.

The first hurdle which comes to mind is regulation, and that's the biggest swing factor. So the tech is already out there is pretty much off the shelf. What really blocks adoption today I guess, is the regulatory side, especially on drones, which is heavier. So in the US this is done on a state-by-state level. Whereas in Europe this is more city-by-city. The only country which has managed to do it on a national level is Finland, and as you can see, the penetration rate is much higher today.

The second hurdle is more on execution and the orchestration layer. So if you think about it, it's super hard for a platform to actually allocate their right to delivery modes to the specific order type. This could be a rider, this could be a drone, this could be a ground robot.

All this has to happen without weakening the consumer experience nor the merchant experience.

And the last hurdle we're seeing is more around adoption, which is going to be very uneven in the world. This will depend naturally on the layout of a city if there's like wide pavements, but also on the country itself, if it has a high labour cost.

Right, so regulation, execution, adoption are the three biggest hurdles for the technology. Now think about it from the investors perspective and just sort of wrap it up for us. How are they looking to get exposure to this theme? Where in the value chain do you think the best risk reward sits?

So you're not the first person to actually ask me this question since we published the report last week. The competitive edge in this field is more around the operational data. And if you look first at the platform side, I think investors will focus more on leaders with scale, strong data, balance sheet flexibility, and players like DoorDash, Meituan or Uber have started to do the work, and autonomous is part of their strategic agenda.

On the more delivery operators side, what really matters is scale, experience and fleet size. So you have private players like Starship Technologies, who we met a few times. They own a fleet of 3,000 robots they fulfilled over 9 million deliveries to date. Having the longest operating history really stands out and helps to stand out, but you also have public players like Serve Robotics.

In summary, it's too early to pick a single winner. There's a lot of capital flowing in this field today, but I'm super excited to keep you up to speed on what's going on.

Yeah, well, we'll definitely get you back in Guillaume, because it's a fascinating topic. Thanks a lot for joining me today.

Thank you Patrick.

It's been great to have Guillaume in the studio with me today and my three key takeaways, if you excuse the pun are – first penetration in autonomous food delivery is less than 1% today, but it could get to 10% by 2035. That should unlock significant cost savings, generate meaningful operating leverage, and could grow the food delivery market in entirety too.

Secondly, the rollout is likely to be uneven and lots of structural hurdles remain.

And finally, riders will remain a key part of the mix, but autonomous food delivery will reshape last mile fulfilment.

Thanks for listening to the Barclays brief and will be back again at the same time next week.