Network Effects #2 | Much Halla about the Mohalla
I thought I will write about something today. Nothing came to mind. It was deeply distressing to be experiencing writer’s block when one isn’t a writer.
The only productive outcome of today was that I can recommend this wonderful show called Justifed to you. Timothy Olyphant is masterful and sublime as Elmore Leonard’s Raylan Givens.
Then, I received this email from Grofers:
In this moment, I was very happy. I finally believed in the existence of God. Yay?
You see, this blog for me is like Bachelor Cooking. It is a ton of fun if you think of it as generally setting fire to things and making a mess. It isn’t really that much fun if you think of it as dinner. Anyway, I digress.
This post is basically stream of consciousness. I saw this email only about 20 mins ago so if some idea is half baked, pls let me know. I’ll try and flesh it out / correct errors later. Before you go further, it may be a good idea to relook at Metcalfe’s Law and types of network effects.
Here we go.
1. Context
A week or so ago, I wrote a post on Restaurants. In that, I made a statement:
Food is a very local business. Unless you are a big national food manufacturer with a FMCG-like footprint (e.g. a Nestle, Britannia or ITC), food businesses will likely have catchments that can be measured in tens of kilometres at best.
I thought this may be a good jumping off point for today. I have spent some time looking at QSRs. My ideal company was one which had a templatized model (or store-in-a-box manual) for a 100 SqFt or smaller store. By doing this, three things happen: 1) SKUs are rationalized; 2) Sales Per Sq. Feet can be maximized; 3) Costs of Failing are small, so if it’s not working out, it won’t destroy your quarter (or year, or your entire business model) The same pool of capital can be used to take many more stabs at the same geography.
Now, let’s consider for a moment, that maybe I am onto something. Micro Markets are perhaps important to examine? Let’s take Grofers as an example and look at this subject.
2. Width vs. Depth
Geography, or specifically, Urban Design is a fascinating subject. Not too long ago, I was discussing a game of life style approach to predicting urban design and how graph theory and a “cost of traversal” model may allow for better predictions of city evolution. Anyway, Arindam Jana knows a lot more about cities than I do. I am just a comp science nerd who loves graphs. If you have questions about how cities evolve and behave, write to him.
Now, Let’s talk about width and depth. It has been, and will remain a subject of deep interest to me. Some businesses allow width to trump depth (Amazon) for at least some time. Some only operate on depth (last mile logistics). Some need both (courier companies).
Geography will affect any business model in two key ways:
It will inform your demand and supply distribution. Think of pre-pandemic days. I ate lunch at work, but I ate dinner at home. The same user is looking for very different things at different times of the day in different places. So residential areas tend to have more fine dining or ghar ka khana style outlets, commercial areas have coffee shops, buffets and generally offerings that service folks pressed for time, or for business meetings.
It creates a challenge of bridging the demand-supply mismatch.
Let’s expand this.
How do we enter a market? First Lens - Demand - We see if the customer base will generate enough orders that the effort of serving the market will be profitable for us. OK, hold on, perhaps we have enough VC funding so we can put short term profitability on the back-burner. The market still needs to generate enough orders that the merchants remain interested, and the daily volume must be enough to create a living wage for the delivery partners.
Second Lens - Supply - the question of servicing this demand. We need quality stores, spread out (relatively) in clusters and enough delivery personnel. We will over-provision the delivery supply to make sure orders are not lost.
This is where the urban development aspects come in. Some areas will see stores generally well spread out (e.g. in Mumbai and Gurgaon). Lots of people live stacked on top of each other, stores are present every 2-3 kms. In such an area, it is easy to execute a Grofers like service. Eventually, there will be pockets (pin codes) where the stores and demand are far apart as to make servicing demand in a meaningful (read headline grabbing) time difficult. These are the areas we eschew. Or, we try to Morpheus the hell out of these (more later).
3. Lego Analogy
Let’s talk building blocks - Lego Bricks. Think of micro-markets as each being a Lego brick. Lego bricks are replicable, fungible. Perhaps the holy grail for Grofers is their Lego Brick. A set of characteristics that make a locality viable for them - a Minimum Viable Pocket, if you will.
It is not really a city. It’s 2-3 pin codes1.
Now, let’s examine this Lego Brick a little more. We need 3 things from this unit at the lowest possible cost:
Maximum Demand
Maximum Supply (Stores)
Optimum Number of Delivery Staff (to match demand to supply).
These 3 drivers will be informed by a number of factors. What is the level of income in the area, how much of that is disposable, what is the propensity to pay, what is the cost of acquiring marginal supply, what are the sources of friction and so on. Each of these answers crystallises and shapes our brick a little more.
This is also where business model is quite critical. Grofers is primarily aimed at Residential areas - there isn’t much of a use case for Grofers in BKC. However, there are a ton of use cases for a Pidge in Commercial areas. So while Grofers will do well to steer clear of BKC - Pidge will likely have 2 hubs.
4. Dark Matter?
Grofers is in some ways, the ultimate Cockroach Company2. Every time I thought it was dead, it has come back leaner, meaner and hungrier. It has proven itself incredibly hard to kill and will likely survive all the inevitable carnage.
Now, if you look at the Grofers email, they talk about Dark Stores. What they are doing is creating their own captive supply. They have found enough demand in areas but the distance equation is not working out. Solution - build pick-up only, small warehouses, limit SKUs, deal directly with the FMCG companies or a limited set of distributors, buy in bulk to get savings and the retailer margins to fund your discounts. AKA, this.
Effectively, Grofers has operated long enough that it has figured out what the ideal Lego Brick should look like. Areas where the brick needs more work, it is putting its finger on the scale make it look more and more like its Ideal market. Sure, there is a cost, but the cost of failure is low and success metrics are known well in advance.
And if you think we have examined this set of characteristics before, we have.
Remember Dominos?
Of course, this will differ for different businesses. For some businesses, their MVP radius may be 50 kms, for some, it may be 2 kms.
It is actually the highest compliment I can offer.