Food Delivery Wars and Credit Acceptance's Underwriting Moat

In Practise Weekly Analysis

Why is this article interesting?

In Practise reflects on some of the key lessons and major questions explored in one or more interviews each week


Food (and Convenience) Delivery Wars

We have always kept an eye on Germany and see tremendous opportunities to improve the customer experience. - CEO of Foodpanda, Delivery Hero

It was a big week for European food delivery companies. Delivery Hero announced it’s re-entering Germany to serve customers with restaurant food, groceries, and convenience. This is particularly interesting because Delivery Hero was founded in Germany under the Foodora brand but then sold to the larger, stronger in 2018. Delivery Hero conceded on their home turf only to re-enter 3 years later with a plan to ‘improve the customer experience’. What is the opportunity they’ve seen? I thought marketplaces were so defensible that once someone wins, it’s near-impossible to compete?

Germany is the most competitive market in Europe with arguably the industry’s best operator running the leading company. TKWY saw EBITDA explode from 19m EUR to 128m EUR in the year to December 2020. Liefrando, TKWY’s German brand, has 26,000 restaurants, 12m active customers, and generated 2.5bn EUR GMV in 2020. Here is more of Delivery Hero’s statement:

By August, Delivery Hero will officially enter the German delivery scene with its full platform offering: from restaurant food to quick commerce, including Dmarts (small warehouses in the city center) and local running the logistics via a sizable fleet of couriers, local partners are provided with an advanced infrastructure that enables them to focus on their core capabilities. For local shops, Delivery Hero acts as a true digital partner, making it possible to access both offline and online customers

Delivery Hero believes the customer is underserved by Liefrando. This seems to be based on two core beliefs that drive DHER:

  • customers deeply value convenience
  • In the long run, customers will want one app for restaurant food, on-demand groceries and convenience

DHER believes it’s the duty of the platform to supply the infrastructure to serve customers with Friday night takeaways, groceries throughout the day, and anything from a convenience store whenever they wish. This is Generation 3 of food delivery: Generation 2 (logistics and marketplace) plus convenience delivery.

Our interview with Former Corporate Strategy Director at Glovo proved to be good timing. Glovo is a Spanish logistics company that offers restaurant food, on-demand groceries, and convenience across Europe and LATAM. We wanted to understand the value of adding convenience to a restaurant delivery platform and what risks TKWY would be under if they didn’t follow.

This quote from our interview immediately stood out:

"I think Gen 2 has won over Gen 1 in pretty much every market, maybe excepting Brazil."

It's probably too early to call Gen 2 a winner but it's interesting to explore . Traditional marketplaces are intermediaries that aggregate demand for suppliers. Liefrando generates orders for restaurants who deliver the food to customers. Ebay connects sellers with buyers. Zillow or Rightmove connects agents with homebuyers. In the early days of the internet, these were fantastic businesses. However, if the digital marketplace doesn’t continually move down the customer buying funnel to improve the customer experience, someone else will. Amazon outcompeted eBay by controlling logistics to offer buyers and sellers a better delivery experience. Uber and DoorDash widened the restaurant selection by offering delivery from all restaurants. Zillow has even moved into buying houses to give buyers and sellers a frictionless experience transacting online. Foodpanda is entering Germany to do the same. They are building the logistics infrastructure to offer customers anything they want whenever they want. Marketplaces that don’t continually move down the customer transaction to control and improve the customer experience are at risk.

So does convenience delivery really add to the existing network effects of a restaurant delivery platform? Glovo saw order frequency and customer retention significantly improve when they added convenience in Spain:

‘you get more frequency and more frequency comes with more retention. Today, real estate or your phone is expensive. If you have Uber Eats and Deliveroo and you are comparing everyone for food delivery, then you might go to a competitor. If actually, you say, hey, I’m ordering groceries twice a week or I used Glovo to drop my keys or to deliver packages, you stay with them. You can see those who have more than one category have much better retention, so that’s the first reason. I also think it’s very easy. There is no CAC; you have already acquired the user. You push them towards other orders, to a positive gross margin, just by increasing their activity. Also, the other order curves are very different. You order food delivery on a Friday night, but when you are going to use the courier or grocery service, it’s going to be a different order curve. It helps with the orders during the day, to utilize to the maximum, when food delivery is a bit quieter.

It makes sense. If you’ve acquired the customer, upselling other products like grocery or high margin convenience products can drive customer retention which increases LTV. However, Glovo was already at scale in Spain. Foodpanda is starting from scratch and needs to build a unique convenience proposition that attracts some of the 12m Liefrando customers. Given Jitse is a fast-follower, Foodpanda needs to build a proposition quicker than Liefrando can offer grocery on-demand themselves.

It’s not clear how customers will behave in Europe. Will we have two apps: one for grocery on-demand and one for restaurant delivery?

I guess the question is, how much are customers willing to have multiple apps on their phone to complete the different missions of, I want convenience delivery and I want food delivery, versus having the vision of, I prefer to have everything within one app. That is going to be something that drives how much better your retention is going to be, by offering convenience.

It’s also not clear how Liefrando will offer convenience. DHER DMart’s are vertically integrated dark stores which source, store, pick and deliver 2-3,000 SKU’s to customers. This is retailing. As with all retailing, you need to offer the correct range at the right price to drive sales. Just Eat Takeaway is not a retailer and is more likely to partner with large grocers and convenience stores. The problem with partnering is that you lose control of the very factor that is under risk: customer experience.

The deeper the vertical integration, the more you can control the customer experience. Glovo and DHER rent dark stores, source their own merchandise, and organize the store to maximise picking efficiency. This improves delivery times and gives you better inventory control which drives higher availability and accurate order fulfilment. If TKWY was to partner with a grocer in a traditional marketplace relationship, it would be harder to pick from physical stores and the higher SKU count with a larger mix of branded products would pressure margins. This is likely to provide a poorer customer experience than a vertically integrated DMart but given TKWY’s scale and the density of a large grocer’s network, it may just be ‘good enough’.

Sign up to read the full analysis and hundreds of interviews.


The forum a trusted place for investors to debate and share ideas on quality companies.


Company Channels
Did you like this article ?