Everything, Now!
Evolution of on-demand delivery models in retail
Yesterday I ordered groceries and they were delivered to my door within 10 minutes. The service? Gorillas*, a last-mile courier service newly launched in NYC.
In my role at Estee Lauder, I lead our omnichannel fulfillment capabilities, including real-time delivery. While beauty is very different than a late night snack, I was curious about this service and the unprecedented speed it offers—and the potential it holds for categories beyond food.
The warehouse where my order originated is a few blocks away from my apartment. It took over a vacant storefront in prime downtown Williamsburg stacked from floor to ceiling with pallets of miscellaneous grocery items. The common theme in the assortment is snacks—things you’ll pay a bit more for to get immediately when a craving strikes (think: every flavor of Ben and Jerry’s ice cream). It appears to be a common theme: GoPuff, a competing service, was literally built for having the munchies.
Within 2 minutes of placing my order, it was assigned to a courier and appeared to be fully packed. Within 8 minutes, the courier was biking up to my front door.
Gorillas is betting that customers care about speed above all else and building a business model to enable this.
In NYC, there is a turf war in progress among the on-demand courier services: bike lanes swarming with Doordash, UberEats, and Instacart couriers; subway walls plastered with QR codes for 1520 (a new entrant, promising your groceries in 15-20 minutes), Gorillas, JOKR, and Fridge No More; daily push notifications and emails from all of these services offering free meals and perks. It feels a foregone conclusion that grocery shopping and meal delivery are forever changed.
But what about categories beyond food? Do customers care about speed when it comes to clothes, convenience items, homeware, or beauty? And how will a courier entrant be positioned to “win” in this new model?
*Yes this (and the links that follow where applicable) are my affiliate links. Gorillas is offering a $20 credit on first purchase with the link. Let’s get free groceries! Would love to hear about your experiences with these services.
Who is this for?
Consumers today are bombarded by two conflicting messages—on the one hand, there’s the virtuousness of curating a meaningful wardrobe of quality pieces over one’s lifetime; on the other hand, there’s Tiktok haul videos showcasing Shein shopping sprees. When it comes to our food, we’re told to eat local, browse our neighborhood farmer’s market, and perhaps participate in a CSA. But just as we’re shamed for moving too fast, being too wasteful, a service pops up that promises 10 minute grocery delivery for ice cream. What do consumers actually value?
In the past, ordering via a food delivery app was somewhat of a guilty pleasure. There was the feeling that we Doordash food when we’re too lazy to cook ourselves. However, the pandemic may have changed the perception of these services. Instead of connoting laziness, ordering groceries and meals in came to mean caring for one’s neighbors and communities and supporting local businesses. The future of these services will depend on how we perceive the act of choosing on-demand delivery: as a needless perk that potentially exploits gig workers, or as an informed choice with potential for positive environmental impact and options to support the local economy.
It’s interesting to see how upstarts like Fridge No More are differentiating around the treatment of couriers, specifying that they do not employ gig workers. This could represent a move toward a more civic-minded courier business model that will improve the perception of on-demand delivery.
Consumers at large are certainly becoming accustomed to faster delivery speeds and options for immediacy. This likely started when Amazon made 2-day Prime shipping the industry standard (147 million Americans have a Prime membership), and accelerated when Instacart and Amazon Fresh pushed same-day grocery delivery into the mainstream.
Meanwhile, carrier rates for other delivery options (e.g. standard shipping from a warehouse to an address) having increased on average by 5% annually over the past several years, now topping $5.67 for a ground shipping package (typically a few business days). In comparison, my Gorillas courier traveled for ~5 minutes by bike (10 minutes round trip)—assuming an hourly rate of $30 (and this is quite high), this would equate to a ~$5 transit cost—less expensive, not including savings on packaging.
According to Business of Fashion’s Fashion’s Last Mile report from last month, consumers are willing to spend $5.70 above standard shipping to guarantee same day delivery, on average. Additionally, 36% of customers surveyed in February 2021 had paid for Same Day Delivery from a retailer, up from 20% 6 months prior. This indicates a willingness to pay for speed across retail categories, and signals that this customer expectation is only in its early days.
Two on-demand delivery business models
In this piece, I refer to on-demand delivery to mean a mode of fulfillment in which the customer’s placement on an order immediately triggers fulfillment and delivery of that order.
While conceptually similar, there are other types of “fast” delivery that use very different business models. For example, Batched Same Day Delivery offered by traditional carriers and companies like Lasership. In this model, the carrier takes all orders from a warehouse or store that are ready by a certain time and delivers them within a geographic area. You will often see on a retailer’s site that orders placed before 10am or 11am are guaranteed Same Day in certain areas. This is a signal they are partnering with a carrier to do Batched delivery. Next Day and Overnight are also flavors of this model, and often supported by more traditional carriers from a warehouse.
With on-demand delivery, there are two types of model that have emerged:
Model 1: Store/restaurant to customer
When Doordash, UberEats, and similar services first launched, the primary business model was taking an order from a Restaurant and delivering to a customer. What helped these platforms grow was the ability to capture sufficient demand in an area, and then to route couriers efficiently so they could steadily accept orders and drop off orders without a lot of downtime. Unlike in traditional warehouse fulfillment, there is no central node that the couriers were returning to after each delivery. Thus, they reduced the amount of unproductive time on their routes.
A possible downside is that customers are restricted to the more limited pool of inventory available at an individual store. Plus, the app has to capture and display that inventory accurately to the customer. In the case of a restaurant, with a limited menu and fungible ingredients, this isn’t too hard. But for a grocery store or retail concept, accurate inventory management can present challenges.
In this model, the courier marketplace hosts the item listing and assigns a delivery driver when a customer places an order. The store/restaurant maintains ownership of the goods and prepares the item for pickup. Thus, the main way for the courier marketplace to make money is from taking a commission from each sale.
Model 2: Microwarehouse to customer
GoPuff and other similar services innovated on the Store to Customer model by introducing microwarehouses—locations in urban areas (unlike traditional, sprawling warehouses that are usually located outside of cities, far from customers) that stock in-demand items. Fridge No More calls these locations “cloud stores,” akin to the increasingly popular concept of cloud kitchens or “ghost kitchens”—restaurants that exist for delivery only, and aren’t open to the public. GoPuff and others in this space acquire the inventory from manufacturers and mark it up to retail, allowing the company to seize a higher margin when compared to the traditional UberEats/Doordash model (see this fantastic deep dive on GoPuff’s strategy).
Doordash and Uber both emerged as leaders delivering from individual stores to individual customers. However, both have been expanding into owned microwarehouses as this model could be more profitable in certain categories, like convenience, grocery, and other consumer categories. On demand delivery companies can predict demand and acquire products directly from manufacturers, then seize the whole margin when they resell. Unlike meals, items in these categories don’t need to be delivered immediately and can be managed more like traditional warehouse inventory. Doordash is testing this model with Dashmart, a convenience store concept. Uber recently partnered with Gopuff to deliver non-food essentials.
I expect that whichever solution tackles non-food categories will take a mixed approach—some fulfillment from existing retail stores, and some from microwarehouses. Fulfilling from existing retail stores allows brands to tap into an existing store footprint and leverage the space as a “free” fulfillment hub. The brand can clear slow-selling inventory and increase productivity of its associates. But using microwarehouses unlocks the ability to position inventory closer to groups of customers that may not be near a store, and to potentially realize more attractive economics by stocking items in bulk.
What type of non-food categories make sense for on-demand delivery?
Grocery is the most popular on-demand delivery category today. But what’s next? To find out, I started by thinking about why grocery made sense as a starting point for so many of these companies:
Predictable demand: The grocery industry has historically had sophisticated demand planning models as they make very frequent, repeated decisions about inventory of perishable items. Predictable demand informs stock levels of specific items and courier availability for certain delivery windows.
Existing fulfillment locations: Grocery stores exist within a short walking or driving distance of the vast majority of customers, thus, they are well-positioned to serve as local fulfillment hubs for a courier model.
High urgency (willingness to pay for speed): Nothing is worse than waking up in the morning and realizing you’re out of milk. Customers have an immediate need for grocery items.
High cost to ship (substitution cost): Before Instacart and Amazon Fresh started offering grocery delivery, most people went to the grocery store in-person to shop. It was very costly (if not impossible) to ship grocery items via traditional carriers, and eCommerce penetration was very low. Without another online option, there was no valid substitute. The ability to get groceries delivered represents a significant savings of time and effort for customers.
Large basket size: Compared to other types of goods, people tend to buy groceries in bulk. Economics of the model require high basket sizes, as couriers get a lot of compensation in tips (typically defaulted to a share of the sale) and on-demand companies get a commission on the sale (%).
Margin profile: Interestingly, grocery isn’t known for high margins (vs. other consumer categories), but I think this will be an area of focus as on-demand concepts expand
To be successful with on demand delivery, I don’t think a category needs to fit all of these criteria. But I think the ones that align most closely to grocery will be where Doordash, GoPuff, UberEats, and the larger players will focus in the near future.
Based on this high level analysis, it’s not surprising to see that alcohol, convenience, and pet are emerging as battlegrounds for on-demand delivery of late (while convenience is a bit weaker strictly on these fit criteria, it’s likely a very large market). Uber acquired alcohol delivery startup Drizly last year, and GoPuff acquired Liquor Barn in June. Doordash has recently expanded into Pet and Floral.
Where things get interesting is the next frontier of gifting, fashion, and beauty. I also suspect there are micro-categories within these larger segments that are particularly attractive.
A few predictions on where we are headed:
On-demand delivery companies will invest in and incubate DTC brands
For emerging DTC brands, in the food, beverage, and other consumer goods categories, securing distribution at scale is one of the biggest challenges in the company’s early years. Listing on these platforms early provides a chance at exposure to a large audience, and the opportunity to gain detailed insights about local consumer demand. I would not be surprised to see DTC brands making this a cornerstone of their distribution strategy, for on-demand delivery companies to invest or incubate their own brands, based on their learnings about the market (similar to how grocery stores have robust private label strategies).
Gifting opportunity will be a source of differentiation this holiday
Gifting is an opportunity for retailers to experiment with on-demand delivery. From sending flowers to a coworker to ordering a bottle of champagne to congratulate a friend, sending a gift is a predictable occasion that is well-suited to the courier model. I expect on-demand delivery models to offer more gifting experiences in the app—for example, adding gift messages for the recipient and allowing the recipient to select the gift themselves within a gift allowance (similar to the fun gift selection experience that startup Goody offers today). Beyond just flowers, brands that play in giftable categories (home, alcohol, candy, snacks, paper goods) will partner and experiment with gift sets and kits on these platforms as it is an opportunity to drive awareness.
Fashion will test on demand delivery with a focus on high-touch experiences
We will see the emergence of high-touch business models that leverage on-demand delivery—for example, the ability to select multiple dresses for a special occasion, and have them couriered to your home to try on. I am in the midst of a Fashion Emergency at the moment—I have a big night out planned, and the shoes I ordered don’t fit. A service that would bring a few options to my door and allow me to try them on would be extremely valuable—and my chance of conversion would be very high. This would be a tough business model for a fast-fashion player, but contemporary or luxury brands could test this out in limited geographies as an interesting differentiator. Bundling additional services, like tailoring or personal styling, with this model could further expand the opportunity.
Currently a limiting factor for fashion on platforms like Doordash and Uber is that customers can only shop from one store at a time (1 order=1 store=1 pick up for the courier). To seamlessly address my Fashion Emergency, I would need the ability to browse a catalog across stores/brands. This would either require brands to be housed in the same fulfillment location, require different courier routing logic than what exists today, or imply that the retailer I’m shopping from is a multibrand player like Nordstrom that carries a wide assortment.
Final thoughts
In my day job, we see a huge amount of opportunity in the real-time delivery sphere—enabling gifting, providing an option when the customer needs something now, leveraging our stores, delighting loyal customers, and enabling other services and events. I’m very curious to see how the on-demand space will grow and evolve. If you have a prediction or experience, or want to chat further about this topic, please reach out.
✨ Stay Curious,
Melina