Delivery Apps Need to Start Treating Suppliers as Partners

by Scott Duke Kominers and Ian Macomber - July 03, 2020

Six months ago, on-demand home delivery was far from mainstream - a luxury for those willing to pay for convenience, skewing towards young professionals in major cities. The Covid-19 crisis changed that, almost overnight. Now everyone is ordering everything from household staples to ice cream on-demand. Existing delivery services like DoorDash and Instacart have stepped up operations, and new entrants have appeared as well; the payments platform Toast, for example, announced its own delivery offering. But there have been growing pains: food delivery companies have struggled with margins, and even marketplace giant Amazon has had to delay Prime Day, its annual sales bonanza.

This upheaval, and the windfall it has brought to certain companies, also presents a novel business challenge: delivery service marketplaces have inverted from demand-constrained to supply-constrained almost overnight. As recently as mid-March, most of those marketplaces were focused on growing market share. Suddenly, customers are showing up in droves - a survey on food delivery run by Brick Meets Click/Mercatus, for example, shows more than threefold customer growth relative to August 2019. The resulting increases in ordering have left platforms scrambling to build more fulfillment capacity.

While some of the challenges of meeting new demand are logistical, there are also reputational dangers to firms that get this moment wrong, as controversies around fees charged by companies such as UberEats and Grubhub illustrate. What’s clear, however, is that the current dynamics are changing the way suppliers are thinking about e-commerce and could be the key to a new, marketplace-driven retail future. As such, now is the time for marketplaces to help their suppliers invest in the infrastructure and business practices that will enable them to compete in a world in which having a delivery sales presence is essential. 

One of us (Kominers) studies the design and operation of marketplace businesses; the other (Macomber) is the director of analytics at Drizly, the world’s largest alcohol delivery marketplace. No one knows how long the Covid-19 crisis will last, but, at least from where we stand, it’s clear that retail won’t be returning to normal anytime soon. And with consumer awareness and utilization of delivery options at an all-time high, delivery demand is likely to persist even after the pandemic is over.

The Shift to Delivery

Drizly is a prime example of a platform being transformed by market inversion. Since March 2020, the company has seen 500% growth rates with hundreds of thousands of new buyers joining the platform. At the same time, there have been inventory stockouts, closed stores, and supply chain disruptions due to Covid-19 - all while the customer service team has been inundated with order volume. That’s consistent with the experience of other mid-size delivery platforms - and even that of Amazon. These platforms have had to switch their focus overnight and develop expertise around the new supply-side constraints.

For Drizly, this has meant a less immediate focus on customer acquisition and retention and more investment in metrics and technology for supporting retailer operations. Pre-crisis, for example, a common complaint on the platform was that drivers weren’t receiving enough orders. Now, those same drivers are overwhelmed with demand, and Drizly has had to figure out how to build product solutions to help stores balance their delivery loads.

Part of the struggle here is that many suppliers didn’t have the infrastructure set up to make delivery work at scale. Until this spring, e-commerce was still a relatively small chunk of the overall alcohol market (2–4%, depending on whom you ask). As a result, platforms like Drizly weren’t a core part of their suppliers’ revenue or in-store operations. Store layouts were optimized for foot traffic, which often made delivery fulfillment unwieldy. (Imagine pulling an order from aisles designed to maximize product discovery, impulse purchases, and time spent in-store!)

But now, it’s worthwhile for suppliers to invest in supporting delivery directly. That’s already happening organically: With in-person retail restricted or outright eliminated, suppliers have been reimagining their store layouts and operations around online fulfillment. More broadly, the crisis has presented an opportunity for “mom-and-pop shops” - including liquor stores - to modernize their technology and pivot their business models towards “omnichannel retail,” simultaneously leveraging marketplace platforms, a self-managed online presence, and traditional in-person sales.

A Bigger Role for Marketplaces

On-demand marketplace businesses are well-positioned to help their suppliers manage this transition. Today, couriers often pack orders from the same shelves as in-store customers and check out using the same cash registers. But once delivery becomes a substantial fraction of a store’s sales, those orders shouldn’t be processed that way; leveraging cross-store learning, marketplaces can help coach suppliers on more efficient fulfillment. They can identify and share best practices around staffing, store layout, and delivery processes. And they can provide technology solutions to help small businesses optimize their operations around delivery, with, for example, cloud-based inventory logs and modern point-of-sale systems.

Marketplace platforms have vision into the whole market, which enables them to predict demand - an advantage they should use to the benefit of their suppliers. Third-party delivery companies can help improve stores’ operational efficiency by structuring a balanced mix between scheduled and on-demand orders and forecasting when to schedule extra drivers on an hour-by-hour basis. They can also increase capital efficiency by helping stores match inventory levels to predicted future demand. In the long run, marketplaces could use their overview of the market to help retailers get closer to an “ideal world,” with exactly the right products on the shelves, and driver teams staffed in direct proportion to order volume. These improvements benefit both parties but require information sharing and communication on the part of platforms.

Shared Benefits

When suppliers become better-optimized around marketplace transactions, the consumer experience on the platform improves, too - often due to increased availability, selection, and price options. Drizly has found that having more stores to choose from and better inventory overall results in much higher conversion rates: a customer with three stores and 1,000 items to choose from spends 67% more per session than a customer facing just one store with 300 items. Plus with suppliers better-optimized for delivery transactions, those orders are processed more smoothly, and with fewer errors, all of which improves the customer experience. If platforms can help their suppliers invest in these sorts of improvements, that will translate into stronger consumer engagement post-crisis.

Meanwhile, for now, delivery marketplaces’ overall consumer value proposition has changed. Before Covid-19, a delivery service’s product-market fit was mostly determined by convenience, speed, and price. Now, by contrast, safety, availability, and breadth of selection are priorities. Customers are placing bigger orders and are comfortable with longer lead times. (Drizly’s “acceptable”  estimated-time-of-arrival windows have widened, and both Instacart and Amazon have delayed order windows by days.) This means many marketplaces can swing economic incentives back towards the supply side. Platforms can lower their take rates to increase supplier participation - especially among chains and other established supplier networks. They can also shift customer service efforts to prioritize responsiveness to retailers and allocate developer resources towards solving suppliers’ operational problems.

On the consumer side, increased order minimums make sense where they wouldn’t have before: a liquor store can’t justify taking an $18 order when there are many $50+ orders in the backlog. Likewise, some fees should be shifted over to the consumers, since they’re now the “long” side of the market. In particular, it might make sense to institute per-order delivery fees, to encourage consumers to place large delivery orders instead of more frequent small ones, as well as to make it more valuable for a retailer to take an additional order. There’s also leeway to make delivery windows wider and more variable to increase the ease of routing and batching orders.

This transition may be especially necessary in the rapidly maturing food delivery space, where the tension between restaurants and delivery companies has increased during the Covid-19 lockdown, and high supply-side fees have come under scrutiny. One restaurant had its Grubhub receipt go viral after showing a take-home of $376.54 on $1,042.63 worth of orders (although some have pointed out that the gap narrows after accounting for order adjustments and promotional discounts). Meanwhile, the San Francisco mayor temporarily capped delivery fees at 15%, and DoorDash and Caviar have reduced commissions, while GrubHub has deferred payments. Food delivery marketplaces’ supplier-weighted cost structure may have worked when restaurants saw having a small stream of delivery orders as a promotional subsidy for in-person traffic - but it isn’t working in the current, supply-constrained marketplace environment.

Bars and restaurants will eventually reopen on a broader scale, and consumers will become comfortable shopping for goods in person again. But the supplier networks that Drizly and other marketplace platforms build will remain. When platforms invest in suppliers as partners, those suppliers will come out of the crisis stronger. And they’ll be better-optimized for online transactions, broadening marketplace offerings, and improving the customer experience in the long run. It’s a three-way win - for suppliers, platforms, and consumers.

Scott Duke Kominers is the MBA Class of 1960 Associate Professor of Business Administration in the Entrepreneurial Management Unit at Harvard Business School, and a Faculty Affiliate of the Harvard Department of Economics. Prior to that, he was a Junior Fellow at the Harvard Society of Fellows and the inaugural Saieh Family Fellow in Economics at the Becker Friedman Institute.

Ian Macomber is the Director of Analytics at Drizly Group, the world’s largest alcohol delivery marketplace. He received an MBA with Distinction from Harvard Business School in 2019. Prior to Drizly, Ian worked on recommendation and pricing algorithms for Wayfair, and marketplace auction models for Zillow.