1/🧵
Ordering food online is quickly becoming the next big thing.
In this thread (🧵), we'll take a deeper look at
- The market
- The business model
- Basket size trends
- Costs per order
- Economics
This is a long one, best served with a hot cup of chai 🫖
Ordering food online is quickly becoming the next big thing.
In this thread (🧵), we'll take a deeper look at
- The market
- The business model
- Basket size trends
- Costs per order
- Economics
This is a long one, best served with a hot cup of chai 🫖
2/🧵
I’m sharing what I've found during the past few months not to say how things are, but to learn from better-advised contributors. Any challenging or complementary insights very welcome.
With that said, let’s start with the market...
I’m sharing what I've found during the past few months not to say how things are, but to learn from better-advised contributors. Any challenging or complementary insights very welcome.
With that said, let’s start with the market...
3/🧵
Market appeal is obvious; people have to eat.
TAM for food purchases is counted in trillions, delivery businesses have raised billions, and online share growing fast. Some listed companies include DoorDash, Delivery Hero, Deliveroo, Just Eat TakeAway, Uber, and GrubHub.
Market appeal is obvious; people have to eat.
TAM for food purchases is counted in trillions, delivery businesses have raised billions, and online share growing fast. Some listed companies include DoorDash, Delivery Hero, Deliveroo, Just Eat TakeAway, Uber, and GrubHub.
4/🧵
Moreover, food is still largely offline.
Online market penetration levels are still in low double digits, as seen in this market penetration caption for markets Just Eat TakeAway (JET) is active in.
Moreover, food is still largely offline.
Online market penetration levels are still in low double digits, as seen in this market penetration caption for markets Just Eat TakeAway (JET) is active in.
5/🧵
Key metric of food delivery businesses is the total value of the merchandise (food, groceries, other products) they deliver; aka "GMV, Gross Merchandise Value", "GOV, Gross Order Value", or “GB, Gross Bookings”.
s23.q4cdn.com/407969754/file…
Key metric of food delivery businesses is the total value of the merchandise (food, groceries, other products) they deliver; aka "GMV, Gross Merchandise Value", "GOV, Gross Order Value", or “GB, Gross Bookings”.
s23.q4cdn.com/407969754/file…
6/🧵
But total order volumes aren’t the same as revenue. Depending on the chosen business model, revenues account for 10-30% of gross order value.
E.g. DoorDash revenue of GOV, or so called “take rate”, was 11% as described in their S-1.
sec.gov/Archives/edgar…
But total order volumes aren’t the same as revenue. Depending on the chosen business model, revenues account for 10-30% of gross order value.
E.g. DoorDash revenue of GOV, or so called “take rate”, was 11% as described in their S-1.
sec.gov/Archives/edgar…
7/🧵
DoorDash “passes through” delivery so related costs won’t be further deducted from revenues.
On the contrary, Deliveroo and Finnish delivery unicorn #Wolt manage their own fleets and have 30% take rates but they must pay delivery from the revenue.
talouselama.fi/uutiset/wolt-j…
DoorDash “passes through” delivery so related costs won’t be further deducted from revenues.
On the contrary, Deliveroo and Finnish delivery unicorn #Wolt manage their own fleets and have 30% take rates but they must pay delivery from the revenue.
talouselama.fi/uutiset/wolt-j…
8/🧵
Many companies in the industry have created three-sided marketplace; consumers, merchants, and delivery partners.
These delivery platforms monetize both consumers and the restaurants. Here’s illustrative economics provided by DoorDash:
sec.gov/Archives/edgar…
Many companies in the industry have created three-sided marketplace; consumers, merchants, and delivery partners.
These delivery platforms monetize both consumers and the restaurants. Here’s illustrative economics provided by DoorDash:
sec.gov/Archives/edgar…
9/🧵
There is a “religion-like” dispute between 3-sided and 2-sided business models.
Some delivery companies e.g. GrubHub and JET focus on scaling the marketplace (i.e. restaurant menus online), steering away from logistics by preferring restaurants to make their own delivery.
There is a “religion-like” dispute between 3-sided and 2-sided business models.
Some delivery companies e.g. GrubHub and JET focus on scaling the marketplace (i.e. restaurant menus online), steering away from logistics by preferring restaurants to make their own delivery.
10/🧵
Total transactions are the function of order amounts and average basket sizes.
In 2018-2020, orders grew >50% p.a. To generalize, EU customers’ average order value (“AOV”) 15-25 USD, while U.S. 25-35 USD. Increasing AOV would improve economics but it's proven difficult.
Total transactions are the function of order amounts and average basket sizes.
In 2018-2020, orders grew >50% p.a. To generalize, EU customers’ average order value (“AOV”) 15-25 USD, while U.S. 25-35 USD. Increasing AOV would improve economics but it's proven difficult.
11/🧵
Order volumes are not the whole story; orders are not created equal.
Large “quick service restaurant”, or QSR chains (e.g. McDonald’s, Chipotle) negotiate themselves attractive deals that are the least profitable (basically zero margin) for platforms like DoorDash.
Order volumes are not the whole story; orders are not created equal.
Large “quick service restaurant”, or QSR chains (e.g. McDonald’s, Chipotle) negotiate themselves attractive deals that are the least profitable (basically zero margin) for platforms like DoorDash.
12/🧵
Reason to have large QSR chains in the platform is to eventually direct customers to other, more profitable “independent” restaurants where money is made.
As GrubHub shows, one should not wish too big share of QSR.
s2.q4cdn.com/772508021/file…
Reason to have large QSR chains in the platform is to eventually direct customers to other, more profitable “independent” restaurants where money is made.
As GrubHub shows, one should not wish too big share of QSR.
s2.q4cdn.com/772508021/file…
13/🧵
Delivery companies deal with (perishable) products in physical world; food gets cold. Cost could be reduced by delivering more orders per hour.
But scale benefits seem to saturate quickly, avg delivery times:
•Wolt 37 min (2017)
•DoorDash 35 min
•DeliveryHero 28 min
Delivery companies deal with (perishable) products in physical world; food gets cold. Cost could be reduced by delivering more orders per hour.
But scale benefits seem to saturate quickly, avg delivery times:
•Wolt 37 min (2017)
•DoorDash 35 min
•DeliveryHero 28 min
14/🧵
Let’s look at the delivery math in its simplest form.
While delivery times are quite similar with same population densities, delivery costs are very local, e.g. in Finland, #Wolt pays on average 18.7 EUR per hour for delivery.
Source: blog.wolt.com/fin/2019/11/01…
Let’s look at the delivery math in its simplest form.
While delivery times are quite similar with same population densities, delivery costs are very local, e.g. in Finland, #Wolt pays on average 18.7 EUR per hour for delivery.
Source: blog.wolt.com/fin/2019/11/01…
15/🧵
With an hourly salary of 15-20 EUR and 20-30 min avg delivery, costs per order range from 5 EUR to 10 EUR (DoorDash had $7.90).
Considering AOV of ~20 EUR, such delivery cost represents “mark-up” consumers aren’t currently paying for – it's subsidized. But there’s hope.
With an hourly salary of 15-20 EUR and 20-30 min avg delivery, costs per order range from 5 EUR to 10 EUR (DoorDash had $7.90).
Considering AOV of ~20 EUR, such delivery cost represents “mark-up” consumers aren’t currently paying for – it's subsidized. But there’s hope.
16/🧵
Firstly, people may be increasingly willing to pay for “time saved”, meaning they are happy to pay extra to free up time from their short, busy days.
Secondly, courier costs could be reduced by batching orders for more revenue per trip.
Firstly, people may be increasingly willing to pay for “time saved”, meaning they are happy to pay extra to free up time from their short, busy days.
Secondly, courier costs could be reduced by batching orders for more revenue per trip.
17/🧵
However, batching has its problems.
Multiple food orders would travel longer, hence quality drastically decreases with batching. This is well discussed in great podcasts by @TheRideshareGuy about food delivery.
However, batching has its problems.
Multiple food orders would travel longer, hence quality drastically decreases with batching. This is well discussed in great podcasts by @TheRideshareGuy about food delivery.
18/🧵
But batching NON-PERISHABLE goods with FOOD order improves efficiency without loss in quality. Food picked up last and delivered first (still warm), durables later.
This is why non-perishable deliveries added to the ecosystem are in the agenda of all the delivery majors.
But batching NON-PERISHABLE goods with FOOD order improves efficiency without loss in quality. Food picked up last and delivered first (still warm), durables later.
This is why non-perishable deliveries added to the ecosystem are in the agenda of all the delivery majors.
19/🧵
Adding verticals to the ecosystem also increases the total addressable market (TAM) that is a widely used proxy for business potential in the world of venture capital.
But not all TAM is created equal. Logistics business is hard. Very, very hard.
Adding verticals to the ecosystem also increases the total addressable market (TAM) that is a widely used proxy for business potential in the world of venture capital.
But not all TAM is created equal. Logistics business is hard. Very, very hard.
20/🧵
Excellent piece by @NewMooncap argues that grocery deliveries have too much friction compared to the basket sizes customers would likely prefer.
Hence groceries would increase the topline significantly, however with negative margins.
newmooncap.substack.com/p/deliveroo-pa…
Excellent piece by @NewMooncap argues that grocery deliveries have too much friction compared to the basket sizes customers would likely prefer.
Hence groceries would increase the topline significantly, however with negative margins.
newmooncap.substack.com/p/deliveroo-pa…
21/🧵
@JitseGroen, CEO and founder billionaire of JET, clearly prefers the marketplace approach over emphasis on logistics:
“They’re basically giving food away for free”, he said on an investor call. “With logistics, you can’t make any money.”
wsj.com/articles/strat…
@JitseGroen, CEO and founder billionaire of JET, clearly prefers the marketplace approach over emphasis on logistics:
“They’re basically giving food away for free”, he said on an investor call. “With logistics, you can’t make any money.”
wsj.com/articles/strat…
22/🧵
Some of the delivery companies have been around for decades (TakeAway 2000, GrubHub 2004, JustEat 2001), yet in aggregate, the industry is unable to post profits.
GrubHub maintained EBIT of 5-20% in 2011-2018 before intense, VC-funded billion-dollar competition began.
Some of the delivery companies have been around for decades (TakeAway 2000, GrubHub 2004, JustEat 2001), yet in aggregate, the industry is unable to post profits.
GrubHub maintained EBIT of 5-20% in 2011-2018 before intense, VC-funded billion-dollar competition began.
23/🧵
It is true that delivery companies are spending heavily to attract new customers to their platforms;
DoorDash S-1 reveals marketing spend as of % of GOV initially at ~10%, normalizing over time to 2%.
sec.gov/Archives/edgar…
It is true that delivery companies are spending heavily to attract new customers to their platforms;
DoorDash S-1 reveals marketing spend as of % of GOV initially at ~10%, normalizing over time to 2%.
sec.gov/Archives/edgar…
24/🧵
In 2020, global COVID shutdowns presented food delivery companies with “a perfect storm”:
-People forced to their homes
-Many had to order food online
-Greatly reduced marketing costs with major boost in growth
In 2020, global COVID shutdowns presented food delivery companies with “a perfect storm”:
-People forced to their homes
-Many had to order food online
-Greatly reduced marketing costs with major boost in growth
https://twitter.com/CNBCJulianna/status/1377189991693094913?s=20
25/🧵
While growth exploded (tripled for some) and customer acquisition costs declined, the observed group reached adjusted EBITDA profitability of $0.13 per order.
While growth exploded (tripled for some) and customer acquisition costs declined, the observed group reached adjusted EBITDA profitability of $0.13 per order.
26/🧵
Adjusted EBITDA is also what the delivery companies seem to be focusing on.
JET has long-term leadership incentives tied around the metric and in 2019, GrubHub believed to achieve $2.00 adj. EBITDA per order (if they wanted to).
s2.q4cdn.com/772508021/file…
Adjusted EBITDA is also what the delivery companies seem to be focusing on.
JET has long-term leadership incentives tied around the metric and in 2019, GrubHub believed to achieve $2.00 adj. EBITDA per order (if they wanted to).
s2.q4cdn.com/772508021/file…
27/🧵
But there’s an elephant in the room.
Almost without exception, when delivery companies refer publicly to “profitability”, they point to “Adjusted EBITDA”. In 2019, our group getting from adj. EBITDA to breakeven would’ve required another $1.2 billion or $0.74 per order.
But there’s an elephant in the room.
Almost without exception, when delivery companies refer publicly to “profitability”, they point to “Adjusted EBITDA”. In 2019, our group getting from adj. EBITDA to breakeven would’ve required another $1.2 billion or $0.74 per order.
28/🧵
Surely, for example #DoorDash has shown great progress in unit economics. But they are already dealing with nearly BILLION orders annually worth $25 billion.
If scale is not even at these levels enough, when it will?
Surely, for example #DoorDash has shown great progress in unit economics. But they are already dealing with nearly BILLION orders annually worth $25 billion.
If scale is not even at these levels enough, when it will?
29/🧵
Existing customer cohorts seem to be spending more as shown by #Deliveroo prospectus, which is great progress.
But if average order economics are not greatly improving from where they are, such aggregate growth matters little.
corporate.deliveroo.co.uk/protected_file…
Existing customer cohorts seem to be spending more as shown by #Deliveroo prospectus, which is great progress.
But if average order economics are not greatly improving from where they are, such aggregate growth matters little.
corporate.deliveroo.co.uk/protected_file…
30/🧵
Moreover, 2020 was extraordinarily good year with reduced marketing needs.
Despite the tailwinds and companies nearly doubling in size, marketing is still more than a dollar per order. For some, new customers seem to be even more expensive to acquire than earlier cohorts.
Moreover, 2020 was extraordinarily good year with reduced marketing needs.
Despite the tailwinds and companies nearly doubling in size, marketing is still more than a dollar per order. For some, new customers seem to be even more expensive to acquire than earlier cohorts.
31/🧵
Regarding topline, growth is normalizing from strong 2020.
Take #DoorDash, the clear leader in US. In 2020 they posted GOV of $24.6 bln while projecting $30-33 bln for 2021. This represents an annual growth of 22-34% compared to growth rates beyond 200% the previous year.
Regarding topline, growth is normalizing from strong 2020.
Take #DoorDash, the clear leader in US. In 2020 they posted GOV of $24.6 bln while projecting $30-33 bln for 2021. This represents an annual growth of 22-34% compared to growth rates beyond 200% the previous year.
32/🧵
The biggest counterargument for the lack of profits is growth marketing.
Assume delivery companies would need NO marketing (totally unrealistic outcome, even Booking.com is still spending 30% of revenue in marketing even though it has 70% EU market share).
The biggest counterargument for the lack of profits is growth marketing.
Assume delivery companies would need NO marketing (totally unrealistic outcome, even Booking.com is still spending 30% of revenue in marketing even though it has 70% EU market share).
33/🧵
In 2019, our aggregate delivery company would have had net profit of $0.04 per order without marketing.
In 2020 (missing Delivery Hero FY results), net profit would have been $0.47 per order, or 880 MUSD for ~1900 mln orders.
In 2019, our aggregate delivery company would have had net profit of $0.04 per order without marketing.
In 2020 (missing Delivery Hero FY results), net profit would have been $0.47 per order, or 880 MUSD for ~1900 mln orders.
34/🧵
These delivery companies have raised more than $7 billion in capital since inception and hypothetical "marketing-light" profits are hovering slightly above 10% of this amount.
Billions raised:
#JET $1.9
#GrubHub $0.284
#DoorDash $2.5
#Deliveroo $2.2
#DeliveryHero $6.6
These delivery companies have raised more than $7 billion in capital since inception and hypothetical "marketing-light" profits are hovering slightly above 10% of this amount.
Billions raised:
#JET $1.9
#GrubHub $0.284
#DoorDash $2.5
#Deliveroo $2.2
#DeliveryHero $6.6
35/🧵
The industry and related investors seem to be relying on “rationalization” via consolidation, or a premature exit via buyout by a rival.
To me, this is not healthy long-term setup for any business.
The industry and related investors seem to be relying on “rationalization” via consolidation, or a premature exit via buyout by a rival.
To me, this is not healthy long-term setup for any business.
36/🧵
To make things more difficult, in 3Q20 conf call #DeliveryHero ruled out improving profitability by
(i) Increasing commissions (restaurants won’t allow it)
(ii) Reducing courier salaries (already a major issue as is)
(iii) Increasing prices (customers won’t allow it)
To make things more difficult, in 3Q20 conf call #DeliveryHero ruled out improving profitability by
(i) Increasing commissions (restaurants won’t allow it)
(ii) Reducing courier salaries (already a major issue as is)
(iii) Increasing prices (customers won’t allow it)
37/🧵
Hence according to #DeliveryHero, investors should focus on “operational efficiency lever”.
In other words, delivery companies have little other options but to keep growing at any cost to capture such efficiency gains.
Hence according to #DeliveryHero, investors should focus on “operational efficiency lever”.
In other words, delivery companies have little other options but to keep growing at any cost to capture such efficiency gains.
38/🧵
For example, #Wolt reportedly tried to raise prices for customers: “Impact was clearly negative”.
The price elasticity of restaurant meals is high; prices go up, demand goes down. Delivery business have little, if any pricing power.
talouselama.fi/uutiset/wolt-j…
For example, #Wolt reportedly tried to raise prices for customers: “Impact was clearly negative”.
The price elasticity of restaurant meals is high; prices go up, demand goes down. Delivery business have little, if any pricing power.
talouselama.fi/uutiset/wolt-j…
39/🧵
Regulatory risks for the business model include labor agreement disputes and 10-20% delivery fee caps introduced by various cities, including Washington, Seattle, LA, Boston, Chicago, and San Francisco.
3-sided delivery model will be most hit.
dlapiper.com/en/us/insights…
Regulatory risks for the business model include labor agreement disputes and 10-20% delivery fee caps introduced by various cities, including Washington, Seattle, LA, Boston, Chicago, and San Francisco.
3-sided delivery model will be most hit.
dlapiper.com/en/us/insights…
40/🧵
For the same exact reasons, delivery company #Deliveroo had to lower its price at the wake of its IPO.
Major investors announced to pass on the IPO due to labor disputes and other regulatory uncertainties. Naturally, valuation also played a role.
cnbc.com/2021/03/29/del…
For the same exact reasons, delivery company #Deliveroo had to lower its price at the wake of its IPO.
Major investors announced to pass on the IPO due to labor disputes and other regulatory uncertainties. Naturally, valuation also played a role.
cnbc.com/2021/03/29/del…
41/🧵
It is a reasonable question to ask, why is so much smart VC money being poured into foodtech?
1) The opportunity is obviously huge
2) Smart money can be wrong, it has been before
3) VCs may be part of a game following the logic below
It is a reasonable question to ask, why is so much smart VC money being poured into foodtech?
1) The opportunity is obviously huge
2) Smart money can be wrong, it has been before
3) VCs may be part of a game following the logic below
https://twitter.com/hkeskiva/status/1371532007054962690
42/🧵
It’s reasonable to expect that few major companies will prevail, eventually being able to add $1-2 dollars to every order without growth loss, after having changed consumer behavior for the good.
The question is, how much more money has to be invested before that happens.
It’s reasonable to expect that few major companies will prevail, eventually being able to add $1-2 dollars to every order without growth loss, after having changed consumer behavior for the good.
The question is, how much more money has to be invested before that happens.
43/🧵
Some companies will prove the presented points wrong. But the companies themselves know that delivery isn’t the most lucrative business in the world; public long-term targets are a far cry from “tech economics”:
Uber 25% EBITDA margin
Delivery Hero 4-8% EBIT margin
Some companies will prove the presented points wrong. But the companies themselves know that delivery isn’t the most lucrative business in the world; public long-term targets are a far cry from “tech economics”:
Uber 25% EBITDA margin
Delivery Hero 4-8% EBIT margin
44/🧵
For further reading on delivery businesses, I recommend @SecretCapital_'s excellent write up: "DoorDash: Re-Inventing Last-Mile Logistics" secretcapital.substack.com/p/doordash-re-…
and @NewMooncap ‘s substack newmooncap.substack.com
For further reading on delivery businesses, I recommend @SecretCapital_'s excellent write up: "DoorDash: Re-Inventing Last-Mile Logistics" secretcapital.substack.com/p/doordash-re-…
and @NewMooncap ‘s substack newmooncap.substack.com
45/🧵
For more similar business threads, make sure to FOLLOW @hkeskiva.
Regarding (food) delivery business updates and thinking, an ex-Uber @ProtagorasTO is surely worth a follow, just like @jitsegroen, @SecretCapital_, @paulfagan98, and @NewmoonCap.
For more similar business threads, make sure to FOLLOW @hkeskiva.
Regarding (food) delivery business updates and thinking, an ex-Uber @ProtagorasTO is surely worth a follow, just like @jitsegroen, @SecretCapital_, @paulfagan98, and @NewmoonCap.
46/🧵
To summarize, as a happy consumer at the current prices, I applaud the delivery businesses.
As an investor, I steer clear away from them unless other data points emerge or my thinking evolves with your help.
If you are still with me, huge thanks and happy Easter! 🙏
To summarize, as a happy consumer at the current prices, I applaud the delivery businesses.
As an investor, I steer clear away from them unless other data points emerge or my thinking evolves with your help.
If you are still with me, huge thanks and happy Easter! 🙏
47/🧵
BONUS: Got good comment that country economics are more revealing. However...
i) cash generating companies don't dilute their ownership by the billions
ii) hundreds of millions of overhead still have to be carried by the orders
Let's look closer
BONUS: Got good comment that country economics are more revealing. However...
i) cash generating companies don't dilute their ownership by the billions
ii) hundreds of millions of overhead still have to be carried by the orders
Let's look closer
https://twitter.com/PLLepine/status/1378302548621279243?s=20
48/🧵
BONUS: Here we have #DeliveryHero broken down to country-level EBITDA.
Then all the remaining expenses all the way to net loss are allocated evenly to all the orders. This way we get approximations for country-level profitability.
Let's look at #JET next.
BONUS: Here we have #DeliveryHero broken down to country-level EBITDA.
Then all the remaining expenses all the way to net loss are allocated evenly to all the orders. This way we get approximations for country-level profitability.
Let's look at #JET next.
49/🧵
BONUS: #JET is a great case of high penetration economics. They've dominated NL market for two decades and on average, they serve every NL citizen (+15yrs) 3.3 orders annually.
With chosen approach, this best case yields 2.3% net profit of GMV and 16% net profit margin.
BONUS: #JET is a great case of high penetration economics. They've dominated NL market for two decades and on average, they serve every NL citizen (+15yrs) 3.3 orders annually.
With chosen approach, this best case yields 2.3% net profit of GMV and 16% net profit margin.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
