Online meal delivery market has been flooded with startups since its inception. We have recently seen a lot of funding and M&A activities in this area, as well as heard news of startups downsizing their teams or shutting down operations. Relatively old players like DoorDash, Grubhub and Postmates have seen stagnant growth recently while new players like Uber Eats and Amazon restaurant have emerged when no one was expecting a new delivery service. These new players have been growing and capturing market share rapidly. As competition is rising, many food-delivery startups are shutting down. I decided to dive deep into meal delivery space to understand business strategy, unit economics, various dynamics and emerging trends.
Quick stats from Morgan Stanley on online-delivery business:
- Currently, $30B restaurant sales comes from online delivery. It could go up to $ 220B (40% of total restaurant sales) by 2020.
- Online orders through third-party delivery apps are increasing compared to phone orders and placing orders directly to the restaurants’ website or app.
- Online food delivery could grow at 16% annual compound rate in next 5 years.
Grubhub has been a market leader in meal-delivery space¹. But, Uber Eats is growing rapidly and is now bigger than Grubhub in 15 major US cities. In recent interview, Dara Khosrowshahi mentioned that Uber Eats’ booking run rate is $6B, which is quite impressive as Uber Eats grew rapidly in such a short span. I believe that the unique differentiator of Uber Eats is their driver network and efficient operations. Uber Eats is available in around 200 markets and has grown by 2400% between March’16 and March’17. In markets like Taipei and Seoul, the profit has eclipsed ridesharing profits.
Apart from that, we have also seen a lot M&A activity in the segment. Caviar got acquired by Square and is integrated with square’s POS system. Last year, Grubhub acquired Eat24 from Yelp, and Postmates and DoorDash are also in discussion for a potential merger.
Meal delivery companies have also ventured into adjacent markets such as catering service e.g. Sqaure recently acquired Zesty, a corporate catering service company. Besides that, Eat Club recently acquired Farm Hill, startup focused on healthy meal delivery. In Meal-Kit delivery space, Blue Apron went public last year. Blue Apron had a challenging time with defending their valuation. Also, startups like Sprig, Kitchit, Door-to-Door Organics couldn’t survive fierce competition and had to shut down their operations.
In order to understand food delivery space and adjacent markets, I have broken it down into four segments: meal-delivery, grocery-delivery, meal-kit and catering service. However, my focus is mainly on meal delivery service in this article.
Please note that the list of companies in the above image is not exhaustive. Also, I haven’t included international players such as Deliveroo of UK, GrabFood of South East Asia and Meituan Dianping and Ele.me of China.
Meal delivery services witness two-sided network effects. As I discussed in network effects article, if there are more choices for cuisines and restaurants on a marketplace, more customers will be attracted to order food, which in turn attract more restaurants to join restaurants and creates a virtuous cycle. As discussed in this article from McKinsey, once customers start using the service, 80 percent never or rarely leave for another platform. That’s why, creating a strong “winner-take-all” effects is very important. In the current state of this space, companies are focusing on growth over profitability to gain market share, so income statement doesn’t provide a real picture of health of business model. So, in order to understand health and viability of a business, it is important to understand unit economics.
Unit economics describe the revenue and cost associated with a business model on a per unit basis². As mentioned in this article, key inputs of the unit economics are churn rate, customer lifetime value (LTV), customer acquisition cost (CAC) and gross margin³. Below is an approach to calculate unit economics of a meal delivery business. In below calculation, unit is an order.
1. Customer Lifetime Value (LTV)
In order to calculate LTV of a recurring customer, first calculate contribution margin and multiply that with number of orders a customer would make in life-time.
Customer LTV = Contribution margin x Average # customer orders
Contribution margin is per unit measure of a product’s gross operating margin, which is revenue per order minus variable costs per order.
Revenue: Food delivery companies generally earn revenue by taking commission from restaurants on their food sales and commission on courier charges. They generally charge 20–30% take rate from restaurants and around 15–25% from couriers. Please note that some deliveries services also provide an option for take-out and may charge different take rate for it.
Revenue per order = (Average Order Value x take rate) + (Courier charges x take rate)
Variable costs: Variable costs are basically costs incurred to complete an additional delivery. In this business model, variable costs are costs of acquiring (referral bonus ~ $ 100), training and retaining couriers, payment processing fees, and customer support costs. Few companies like Uber Eats also run promotion for couriers like offering quest and boost, which should also be counted towards variable costs.
In order to calculate variable costs per order, calculate churn rate⁴ and numbers of deliveries made by a courier in a month and then calculate how many deliveries they would make in life-time.
Churn = (# of churned couriers/total number of couriers) x 100
# months until churn = 1/Churn rate
# of deliveries made by a courier = # months until churn x average # of deliveries in a month
Average Number of Orders
Similar to the calculation for couriers, you can calculate average number of orders placed by a customer.
2. Customer Acquisition Cost
Customer acquisition is generally cost of marketing to acquire new customers. Generally, all meal delivery companies provide double referral — referrer and new customer both receives $ 5–10 credit and/or free deliveries.
CAC= total marketing costs / customers acquired through marketing channels
After calculating all above metrics, if LTV/CAC ratio is more than 3, then business model is considered as sustainable. Thus, success of this business lies in acquiring customers cheaply, retaining existing customers and generating revenue for a long time.
In order to retain customers for a long term, companies have also started rolling out subscription plan and loyalty programs. As companies grow, their focus change from growth at any cost to becoming operationally profitable. Thus, having strong unit economics is an indicator of company’s long-term success.
POS Integration and Full-Stack Technology Solution
In order to understand customer behavior and drive personalization, restaurants and QSRs are adopting digitization of their processes. However, the issue is that the data resides in different systems. For example, restaurants get a list of their guests from reservation system like OpenTable, however, since point of sales system is different, they don’t know what a particular guest ordered in the restaurant. Also, restaurants get name of diners and their orders via an app. But, they have to manually enter orders in their POS, which may lead to human error and may also slow down operations as until someone enters order, it won’t reflect in kitchen’s screen. Thus, POS integration has become key in restaurant technology.
POS integration could help food-delivery services to get more restaurants partners on their marketplace. For example, Grubhub already integrates with several leading point of sale systems, and CEO Matt Maloney has cited the company’s technology as a main reason it was able to land an exclusive partnership with Yum Brands’ KFC and Taco Bell restaurants. Uber Eats also recently acquired OrderTalk, a SaaS provider, which integrates with more than 10 of the leading POS providers and most major payment processing vendors.
Besides POS integration, as shown in restaurants tech ecosystem map, there are various companies who are technology providers for restaurants to manage guest experience and restaurant operations. Some of the companies have already ventured into adjacent markets e.g. Square, who is POS provider for restaurants and QSRs have also entered into meal-delivery business (via its acquisition of Caviar) and in catering business (through its acquisition of Zesty). I believe that there might be even more consolidation in future and delivery services might start providing full-stack technology solution to the restaurants, which could also help them to get more restaurants on their network.
Use of Data and Uber’s Unique Opportunity
Personalization has become key differentiator for user experience in all industries. Food delivery services collect lots of data on what you order, which restaurant you order from, search activity within app and frequency of your orders, which can help them to create users’ taste profile. They can algorithmically recommend dishes, help restaurants to optimize their menu and attract new customer to try their service. For example, Grubhub has 14.5 million active users ordering from 80,000 restaurants, which generates enough data to provide recommendation. In addition to that, meal-delivery services also leverage data to make their operations efficient by balancing load among couriers. The algorithm optimizes courier delivery process based on current location of courier, restaurant location and destination.
I believe that Uber Eats has a unique opportunity here. Uber offers a credit card and gives 4% cash back on dining including on Uber Eats, which drives customers to purchase food with their credit card. Uber can collect data on spend from restaurants and their competitors. That would allow them to know restaurants you visit, cuisines you prefer and how much money you spend on food. Uber can also get insights on your eating habit, popularity of food and restaurants in your neighborhood. Besides personalization, Uber could also leverage this data and insights to build and operate their own “ghost kitchens” and create menu according to the preference of users, target right customers for marketing and could bundle Uber Eats offers with Uber rides.
Modular Robo Delivery
Human delivery costs account for up to 80% of costs incurred by delivery services. Thus, automation is important for cheaper and faster delivery. For example, Uber plans to deliver food by drone in San Diego as part of a commercial test program approved by the federal government. In 2017, Marble deployed robots on San Francisco’s side walk to deliver food. However, deploying bots on sidewalks isn’t efficient, especially when delivery location is far from restaurant. Also, with high volume of orders, significant number of bots would require and sidewalks could get completely hijacked by the bots.
Early-stage startup, Kiwi has adopted a multi-modal delivery system where there are three types of bots used in a delivery process. One robot is inside restaurants to collect the food from kitchen. Second robot is a semi-autonomous tricycle, which collects food from restaurant with human assistance and it goes in the streets. Essentially, a self-driving car could also be deployed depending on the delivery distance. And the third robot is used to deliver food to the door. The third robot is designed to travel only last 300 meters to the delivery location, so sidewalks won’t get flooded with the robots. I really liked this approach as opposed to having a single bot delivering food. In this approach, machines augment humans as opposed to replacing them completely.
Food delivery services bring extra source of revenue to restaurants and provides them exposure to the market. While it is true that it brings more sales for the restaurants, but delivery apps are accused for hurting restaurant’s margin.
Let’s dig into restaurant’s operations costs to understand this. Restaurants generally operate at a pretty low margin. According to industry standard, thirty percent of revenue counts toward cost of ingredients, another thirty percent towards rent, utilities, insurance, payment processing fees and 20–30 percent of revenue goes to the delivery services. Remainder (~5–10%) is generally a profit. Thus, it is safe to assume that margins are very slim. Sometimes they may just breakeven.
Delivery-only model (ghost kitchen) might solve this issue. David Chang’s Ando (acquired by Uber Eats) and Munchery (offers ready-to-heat meals) are couple of examples of this model. Since they are delivery-only model, they don’t have to pay for wait staff, fancy infrastructure and real-estate, which could significantly reduce cost of the operations. Also, it allows them to change and experiment with their menu easily to understand consumer choices. Consumer will also be benefitted by getting food at a cheaper rate and get cuisines based on their choice. DoorDash has started to experiment with leasing remote kitchen space to restaurants so that they can expand their delivery radii.
Tech has revolutionized food delivery space. But, operational efficiency is the one that truly sets customer experience apart. In order to gain operational efficiency, having last mile delivery network is essential. Uber has more than 2 million drivers in its reliable networks, which really gives them an edge over competitors⁵. However, Uber might still face competition from Amazon and Grubhub as they are investing in last-mile delivery network, but it will be difficult for these companies to create such as huge last-mile delivery network. Besides that, as competition intensifies, I expect more M&A activities leaving only a few major players to compete. As we’re already seeing, these companies might dive into adjacent markets to create additional revenue streams and may become technology provider for restaurants. But, after all, company with reliable logistics network and ability to acquire customers with high life-time value at a cheaper rate will win the market in long-term.
¹Grubhub’s market share consists revenue of Seamless, Eat24, Order Up and DiningIn
²Calculation of unit economics is more like art than science and differs depending on the type of business model
³Discount rate is also generally used to calculate LTV in unit economics
⁴Calculation of churn might be complex depending on the business scenario and complexities
⁵Uber wants all of their orders to be delivered in 35 mins