The Future of Food (Delivery)

In May, I shared a deck made up of insights, observations and predictions for the present and future of the food delivery space, after spending 2.5 years at Uber working on building Uber Eats. Sharing the deck was meant to be a test of how sharing material using slides, rather than articles or blog posts, impacted engagement.

The results were clear — decks result in better engagement. The deck was seen by over 100,000 people (including being shared on Hacker News) in over 100 countries. Best of all, it led to many interesting conversations and even a few new friendships and advisory opportunities.

The problem with decks, though, is that they currently aren’t well integrated into content sharing platforms (LinkedIn, Medium, etc.), so the shared material doesn’t persist.

I’ve reformatted that deck from May and am sharing it here and on LinkedIn in article form to allow for easy future reference. The full deck has more detailed insights around winners and losers as a result of the changing landscape.


What is the current state of food delivery?

  1. Restaurant / takeaway
  2. Grocery + meal kits

Most Delivery Service Partners (DSPs) currently focus on either restaurant or grocery delivery, while others have taken on both challenges.

DSPs are typically available regionally, with no DSP having full global reach.

While the US generates a large portion of the current annual revenue in the space, penetration is behind that of other valuable markets (such as China), creating a significant opportunity for further growth.

Investment has resulted in ruthless competition on end consumer price and for enterprise partnerships. However, restaurants are struggling to adapt as they are aggregated by DSPs , who are operating from a position of considerable leverage.

Traditionally, restaurants operate on only a 3–5% profit margin while DSPs typically take ~20% of each order or more. Restaurants could shift to a delivery-only model, but that carries significant risks such as alienating customers, laying off front-of-house staff and putting trust in DSPs. This is even more true for enterprise chains due to their franchisee model and complex regional or global logistics networks.

Despite this hit to profits, restaurants struggle to say no since as many as 22% of people expect to dine out less frequently in 2019 than they did the previous year. Restaurants that join DSPs have an immediate advantage over competitors that do not in terms of awareness and availability. News publishers have experienced this phenomenon in recent years, with Facebook and Apple commoditizing the accessibility of news.

Why is the category experiencing explosive growth?

Investor optimism has led to huge influxes of capital into a variety of business models over the last 5+ years.

Established brands are excited to expand their reach while not taking on the full logistical complexity of operating a delivery network.

“People are so used to living their lives with one click of a button… Do we want to wait to be disrupted, or do we want to be the disruptor?” — Steve Easterbrook, McDonald’s CEO

What competitive differentiators are emerging?

Travis Kalanick

Dark kitchens, in particular, have generated interest from some big names in tech. Travis Kalanick, Uber’s co-founder, purchased and has been growing City Storage Systems, a dark kitchen pioneer, in Los Angeles for over a year.

Subscriptions have emerged as a differentiator in the US market, with DoorDash’s DashPass fueling massive growth by saving money for frequent users.

Outside the US, food delivery is simply one of many services offered by Super Apps like China’s Meituan-Dianping.

Super Apps have quickly grown in China, India and southeast Asia where companies are (and modern life is) often mobile-first.

DSPs are also pursuing other, less defensible strategies to gain short-term market share. These include:

  • Consumer discounts — Similar to how ride-sharing grew, DSPs are offering considerable subsidies. These will not be sustainable, particularly for smaller DSPs with smaller warchests.
  • Reliability — DSPs are often paying excess incentives to drivers to ensure they can fulfill consumer demand.
  • Tech improvements — With how much venture capital investment the space has received, there is top tech talent at most DSPs. Catching up to competitors who improve their tech or launch new products often does not take long.

Areas for Continued Innovation

  • Continued growth in urban areas will drive down consumer costs
  • New technology & situations for delivery will be normalized
  • The first Super App in the US market will emerge
  • Relationships between DSPs and established brands will evolve
  • New opportunities will emerge for non-tech workers

Pay-to-play will become more common

Food delivery has produced some spectacular failures

No DSP has been able to prove consistent profitability with their existing business model. It follows that they will need to explore additional revenue streams as they scale. DSPs have a few quick-win options available to generate incremental revenue.

  1. Auctions — DSPs can learn from Google Ads and create bid-based systems, where restaurants or brands can pay to be found in search.
  2. Add-on Items — DSPs leverage machine learning to recommend items they predict consumers will like. They could charge partners for inclusion in these recommendations.
  3. Visibility Boosting — DSPs can charge partners to temporary index higher in their machine learning algorithms (and show up more often to consumers)

Continued growth in urban areas will drive down consumer costs

For DSPs, “batching” is the concept of assigning one driver to pick up multiple orders simultaneously. This is typically only done when the orders will be dropped off close to one another. The benefit for DSPs is that they are able to save on costs with minimal impact to food quality.

VCs like Andrew Chen know that in Chinese cities, where order density and batch rates are both very high, low costs are already the reality for consumers. Ordering in can be cheaper than dining out.

Uber Eats batching flow

In the US, DSPs have begun creating ways to artificially increase batching opportunities. Uber Eats, as shown, looks to concentrate demand / orders on certain restaurants at certain times through time-gated discounts. They set the expectation for batching upfront in these cases. Other DSPs have since followed suit, both inside the US and internationally.

New technology & situations for delivery will be normalized

Removing labor costs from the equation will likely be dramatically cheaper for DSPs than maintaining a fleet of autonomous vehicles will be. Temperature-controlled autonomous vehicles, such as those being built by Nuro (pictured), will allow for a more consistent experience for consumers. Autonomous food-trucks that cook food to order as they drive to the consumer already exist, with the Bay Area’s Zume leading the way. Between that and more clear-cut opportunities like drone delivery, DSPs will be able to go from receiving an order to dropping it off significantly more quickly.

There is also an opportunity for DSPs to add incremental business through logistics and integrations. DSPs can geo-encode GPS points to act as pick-up zones. This enables consumers to order from, essentially, anywhere.

DSPs have begun to integrate with traditional POS systems but, in many cases, are better suited for the job themselves. In what’s the first of what’s sure to be many startups that pop up to support DSP integrations, LA-based Ordermark aggregates inbound orders from all DSPs to one central device. Many enterprise chains have custom configurations and metrics that they use to track productivity, so this space is likely to continue to evolve.

There are logical opportunities to improve the experience for events like stadium sports and movie theaters via in-seat ordering. Some DSPs, like Caviar, have already begun offering this at certain stadiums.

The first Super App in the US market will emerge

The mobile-first Gen-Z will create opportunities for multiple approaches in the US market, which already has many established DSPs.

Established DSPs will look for adjacent verticals (retail, pharmaceuticals, etc.) or complementary pieces of the value-chain (payments, distribution, etc.) that require minimal integrations.

DSPs looking to gain market share quickly in new verticals or regions, or fend off competitive threats, will continue to look to acquire other DSPs or pieces of the value-chain.

The most likely DSPs to attempt to become a US Super App are already major names in the US market.

  • DoorDash — Their Drive product already focuses on verticals other than restaurant delivery. It will only become easier to acquire users for their subscription, DashPass, as they offer more services with it. Similar to Amazon Prime.
  • Uber — Uber-as-a-platform is their current strategy; grocery-centric roles are currently posted on their careers site. Despite losing market share to DoorDash, Uber has already proven it is more than a one-trick pony. Its Rides product gives it a huge advantage here.
  • Amazon .- Their acquisition of Whole Foods put them in prime position (no pun intended…). However, Amazon Restaurants did not made a dent in the US market over the last few years and has been shut down.
  • Foreign super apps — It would be surprising to see none make an attempt, but none have expressed an interest in doing so.

Relationships between DSPs and established brands will evolve

For DSPs: Having greater restaurant selection is a more flexible path towards creating stickiness and increasing the retention of consumers.

Established brands are already looking to avoid exclusive relationships with DSPs

For established brands: Being available on a single DSP greatly restricts established brands from maximizing their reach, relevancy and revenue.

Ride-sharing has a similar dynamic. Uber and Lyft don’t demand exclusivity from drivers since it would be worse for both platforms if they did, and the most successful drivers drive on both platforms to ensure they get the most jobs.

New opportunities will emerge for non-tech workers

New working models and types of work will grow in popularity as companies expand their control over the value-chain. Instacart employs in-store shoppers at high volume grocery stores. Jyve offers ways for gig-economy workers to graduate to higher paying tasks based on performance.

“Having the confidence that you can find experienced help quickly from a trusted source like Instawork is invaluable. It ensures our restaurants are running smoothly and it’s a seamless process for our chefs and managers.” — Joel Dixon, Instawork User


  • Competition will likely continue to be a primary driving force in the industry for some time. The market could end up with large, regional players like ride-sharing
  • The rise of DSPs is an indicator of larger, ongoing changes in consumer behavior, the future of work and the views of established, global brands


































Product Manager @ Airbnb. Startup advisor. Previously helped build Uber Eats.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store