New York State of Mind
Earlier this week, DoorDash announced that it has begun piloting a fifteen-minute delivery service in New York. In fact, it has created a new company, DashCorps, that will support the future expansion of 10 to 15-minute deliveries from a variety of retailers, including grocery and convenience store partners.
DoorDash now joins a long list of firms, such as FridgeNoMore, GoPuff, JOKR, and Gorillas, that are already competing in the exact same location (NYC), and the exact same market (15-minute deliveries). If you do not live in NYC and have never heard about these firms, don’t worry; some may never live long enough to reach your city. One firm,1520, just shut down after burning all of its cash on trying to survive in the cutthroat NYC market.
But I don’t want to focus on whether the time frame that all these firms are targeting is feasible or profitable (short answer: for firms, it might be too expensive, for customers, it’s not fast enough).
In this post, I want to challenge the idea of starting in NYC.
Because, New York, I love you, but you're bringing me down (okay, this is the last musical reference I will include).
I want to argue that New York is absolutely the wrong place to start anything that is operationally-driven.
On the Upside
I think it's important to understand why these firms choose New York City to launch their business.
The reasons for using the Big Apple as a starting point are pretty evident.
First, among entrepreneurs and investors lies the belief that if I can make it there, I'll make it anywhere (I know, I promised, but I just couldn’t resist). If they are good enough to conquer that market and get their metrics right, then they could use that success to launch anywhere else.
But why do founders think they can prove their business model in NYC?
Many of these delivery and mobility models rely heavily on density. The unit economics of these 15-minute deliveries (next-day deliveries, or any last-mile deliveries for that matter) are sensitive to an operational metric that many of these firms track: drops-per-hour (or stops-per-hour in the logistics sector). This metric is particularly important because delivery costs are high (and usually paid in dollars per unit of time). To justify this cost and make the economics viable, a delivery person has to make as many drops-per-hour as possible.
In order to increase the number of drops in an hour, you need population density, and this is the second reason that NYC seems like a good place to start. The denser the city, the fewer customers you need to acquire to have this operational metric look good.
Data from the 2010 Census reveals that New York City is the fifth densest city in the US, but all other more dense cities don’t even break seventy thousand residents. If we consider boroughs as separate cities, Manhattan would be the most populated city in the US. So, the New York City “Metropolis” is the densest metropolitan area in the US and therefore, in terms of density, there is no question—Manhattan is your best bet.
This brings me to my third point. After the San Francisco Bay Area, NYC is the densest in terms of investors. Investors like to look at numbers that they can relate to. They like to invest in businesses they can use and see (and tell their friends at dinner parties), primarily in the early stages of investment. So, from personal conversations with founders I know that many times investors compel them to be in New York.
Finally, New York still remains the media center of the world. Looking at the amount of exposure 1520 received when compared to GoPuff, it’s clear that the former got a lot more exposure given it started in NYC, compared to the latter, which incubated for many years in Philadelphia. This is not an accolade for Philly. It’s just a fact.
On the Downside
But for all the reasons that make NYC suitable, there are several arguments against it.
It is much too dense. A recent study ranked New York as the city with the worst congestion in the United States and 5th worldwide. And while you will argue that many of these deliveries are done on bikes, this is not an all-weather solution (and when it’s snowing in NY, sidewalks and roads are never plowed as fast as you want them to be). And this is before the self-driving cars and delivery vehicles, which will just make things worse.
Because of this density, many delivery firms pay hefty amounts on traffic tickets. Fresh Direct, the first grocery delivery firm in NY, is among the top ten most heavily-fined companies in NY. I don’t think many of the founders account for this in their unit economics.
It is expensive. New York is an expensive place to start. The main expenses for e-grocers are: Customer Acquisition Costs, real estate (if operating a fulfillment center, which many of these firms do), and employees (delivery people). All three are more expensive in NY than in any other city.
Let’s take a more detailed look below:
Customer Acquisition Cost: Since these firms all compete in NY, they must pay to “steal” customers from the competition. They must try to break through the noise, as the photo illustrates:
Real Estate: Commercial real estate was cheaper during COVID-19 but has rebounded almost 100%. When it comes to commercial space, New York City prices are approximately $80/sq. ft per month, while in Chicago it costs about $37. The average prices may vary based on location and terms, but you get the idea.
Delivery People: Given the high cost of living and the fact that there is significant competition over these drivers, wages in New York are higher compared to other places. If we use Uber’s compensation as a proxy, the only cities where the pay is higher than New York are Seattle and the San Francisco/Bay Area.
And From the EastSide to the WestSide
There are many other options: New York has always had a very variable scene of 24/7 delis, small grocery stores, and bodegas, all within walking distance throughout the city. These are also a part of the competition, which increase the overall costs of acquiring and retaining customers. It also means that NYC will never be a winner-takes-all market. As long as you operate in New York, you will always be competing with someone.
Not a template for success: Most importantly, New York is unique. There are 10 cities in the US with more than 1 million residents, and the New York metropolitan area alone has 18.5 million (8.8 million if we only count New York City proper), which means that it’s significantly more populated. So even in the US, it’s unique and cannot act as a model for any other city precisely for all the reasons I have outlined above.
The take-away is that whatever learning you acquire from NY, you probably have to scrap once you move to Boston, or DC, or Cincinnati.
At the early stage of a firm (or a pilot), you want to take the time to experiment and balance cash burn with learning in order to use the same concepts and tools in other cities. Burn cash to build a template you can execute in other cities. But what happens when the learnings are not transferable?
I am sure some of the learnings are transferable, but most are not. From the right size of the distribution centers, to the right marketing policies and modes of transportation, most factors will be quite different when moving from NYC to another city.
And it’s not surprising that many of the current market players have started in Europe. Europe has 34 cities with more than 1 million residents. Of those, Berlin seems to be a much better city template than New York.
And indeed, that’s where Gorillas launched.
But when they decided to try out the US, they began with New York, as if going over the Atlantic results in learning loss.
Over the next few years, I expect the situation to become even more challenging. New York is not getting less congested, and the decline of the major cities in the US may make it even less similar to other cities. More expensive, but also with more coverage and more investments.
If Europe is not an option, then I think the right place to start a fast delivery service in the US are the streets of Philadelphia…because It’s always sunny here (seriously, last ones), and also because all the factors we have analyzed here are at a level and a rate that could provide a good model that could easily be replicated elsewhere.
Thanks for reading Gad’s Newsletter! Subscribe for free to receive new posts and support my work.