Description

Food delivery apps are proliferating.  I can’t discern any strong first mover advantages, or network effects as the industry is structured today.  Switching costs are zero for both users and restaurants, and most restaurants are set up with multiple platforms.  I believe GRUB’s strategy is weak, and two delta force competitors (Uber and Amazon) have recently entered the fray.  I also believe that traffic to Seamless began to decline in Q1, even in their core NYC market, and daily average users for the Seamless app began to decline in Q1 for the first time.  I believe Uber has a superior product and the lowest cost network.  The sell side believes GRUB will grow >20% organically in 2016, but I think that’s unlikely.  Longer-term I think there’s better than even odds that the industry devolves into a gory death match.  I think GRUB squandered their first mover advantage and is unlikely to replicate its NYC success anywhere else.

GRUB’s legacy model is to route orders to restaurants and charge a +/- 15% commission to restaurants, who then either fulfill orders via the restaurant’s delivery drivers or via in-store pickup.  This part of the business could potentially have network effects, but it doesn’t because GRUB hasn’t integrated with the restaurant POS systems so there’s no lock-in, and they’ve chosen to over-monetize rather than create a virtuous circle.  That the company believes their model is a virtuous circle illustrates how poor their strategy is and how unprepared they are for competitive onslaught.  15% is a lot to charge for taking an order, let’s be real, and restaurants don’t like GRUB and actively try new platforms.

Under GRUB’s new model, in addition to routing orders, fulfillment is handled by GRUB’s proprietary delivery drivers.  In this model GRUB takes about 40% of the economics of the order.  I think there are a few missteps in this strategy.  First, the take rate is higher than Amazon’s 20% take rate and Uber’s 30% take rate for order and delivery.  Second, GRUB is at a stark disadvantage to Uber’s driver network which numbers about 30,000 drivers just in NYC (growing 100% YOY), a density which I don’t think GRUB can ever match.  Three, the cost of Uber’s driver network is structurally lower.  A food-only delivery network will always suffer from uneven volumes which surge in relatively small segments of the day, whereas a network which serves riders and food delivery can subsidize the food delivery network during off-hours by providing rides.  Since GRUB is unlikely to make inroads into ride sharing, I think GRUB is asymmetrically disadvantaged relative to Uber.  Similarly, Amazon can subsidize their food delivery network by delivering Prime Now orders at off-peak times of the day. 

UberEats launched in NYC and several other markets about a month ago.  I think UberEats is already a superior service and is likely to improve.  First, the cost to the consumer today is zero during the launch period but after the launch period will be $3-4 and will not require the user to tip.  $4 is less than the $5 average delivery fee+tip that GRUB users pay.  And if UberEats follows the pattern of ride sharing, Uber has shown that they are willing to pass scale benefits through to customers, which makes them a powerful virtuous circle.  Second, UberEats restaurants are curated for quality, whereas GRUB is a collection of low quality and phantom restaurants.  One restaurant owner told me that he refused to be on GRUB because he didn’t want to be affiliated with GRUB’s roster.  GRUB allows inferior restaurants to pay to receive priority placement in searches.  I believe GRUB should prioritize relevance and quality over money, much like Google AdWords which doesn’t guarantee priority placement to the highest bidder.  UberEats also carefully merchandises the restaurant’s products, sending a professional photographer to arrange and photograph the items.  If there are photos on GRUB they are poor quality and sad-looking.  Merchandising may sound like a trivial issue, but it matters to restaurants.  It also matters to consumers, if only subconsciously.  Third, UberEats average delivery times are already faster than GRUB.  I believe this is due to Uber’s existing dense driver network.  GRUB cannot replicate Uber’s driver network even in Manhattan, and is an even more improbable winner in smaller markets.  Uber brings with it millions of very engaged and supportive customers who can learn about UberEats via very cheap email marketing.  I think most people, even restaurants, heretofore have underestimated the significance of UberEats.

With a $1.8 billion enterprise value, GRUB trades for 28x 2015 free cash flow (adding back stock-based compensation and working capital outflows).  If 2016 free cash flow grows organically, I think it will grow at a rate that disappoints the sell side.  And at any rate GRUB is priced as if it will grow at 20% forever, and that seems unlikely to me without first having a shakeout of excess profits.

Catalyst

Q1 and 2016 earnings disappointment

Declining web traffic and daily average users

UberEats rollout in March