The math behind what is a startup
-by Sarthak Jain
“A startup is a company designed to grow fast. Being newly founded does not in itself make a company a startup. Nor is it necessary for a startup to work on technology, or take venture funding, or have some sort of “exit.” The only essential thing is growth. Everything else we associate with startups follows from growth.” — Paul Graham
It took me about 1.5 years of running a startup to embody this concept. I’ll take a crack at building some math to better explain what I learnt.
First Businesses (10,000BC)
The basics of any business are, building the product and distributing the product. The first businesses were built around building something and trading (or later selling) it one by one. Let’s say I built a wheel, it would take me $50 to make it and $10 to find somebody who wanted to buy it and get it to them. If I built 10 wheels, it would take me $500 to make and $100 to find 10 people who wanted to buy it. The constrains to the amount of money you can make is how many products you could build and how many customers you could find.
People soon realized it was inefficient to find individual people to sell to. They set up shops and spent a large amount to attract customers, but whether you got 1, 10, 100 or 1000 customers your DistributionCost was going to be the same. This required business owners to spend money before they sold anything. This made doing business risky.
Innovators further realized that it was inefficient to build a single product. They built the production line which made businesses scale where building 1 product or a 100 cost the same. If you sold too few products, your fixed costs would take you to bankruptcy, which made businesses more and more risky. However the upside of success made success very lucrative.
These business now are built to scale, the more people you get the more money you make. Without having to put in any extra time or effort. This was utopia, till people realized that when things get bigger and bigger, everything isn’t that straightforward.
Inefficiencies — New Dependencies
That equation is correct for when NumPeople is reasonable. The problem with this is that inefficiencies become violently obvious at scale. If you have to move goods from your manufacturing plant to your shops, it requires trucks. 1 truck is enough for maybe 100 products, but what happens when you have 10,000. A single shop cannot attract 10,000 people and things start to go south. You need more trucks, more shops and factories.
The dependent factor then changes. It is no longer is dependent on the number of people, but dependent on some other factor which is in some way related to NumPeople. Maybe you need to add 5 stores for your next 5,000 customers and 10 trucks for the next 1,00,000 customer. At scale the equation starts to look something like this.
Please understand that this is overly simplified as well as overly generalized. The fundamental idea is that the costs incurred are dependent on the NumPeople but how much it is affected by the number of people decreases very very rapidly. Which is what enabled the advent of the first very very large companies. Examples of companies who this model applies to are Apple, Chevron, Bank Of America etc.
The key to success for this model is scale, scale and scale. However to achieve this kind of model the ProductionCost and DistributionCost are exorbitantly high. You can’t run Intel unless you manufacture millions of chips, it just doesn’t work. The unfortunate thing is that most businesses can’t begin at scale from day one. This leads to very few large businesses with this model.
When we entered the 90s all hell broke loose. People realized that none of the conventional rules applied anymore. The world had changed, and all of this happened because because of the internet. You no longer needed shops, or trucks. You have already paid the cost to acquire customers and build the product, irrespective of number of customers. Business were built that where it cost the same if you were selling to an individual or to an entire country.
This is where things get really interesting and this is where we get to what makes a startup a startup. The true definition of a startup built for scale is something like this:
This is glorious, mind blowing and for some, spiritual. The idea that a business can be built where you only spend a certain amount of money to build the Product, and no extra costs are incurred irrespective of number of customers. The 2000s brought something even better, your production costs plummeted and you were literally paying nothing to get started because of cheap hosting. You can write a blog post that a billion people read and all you spend is the cost of a laptop and cost of internet in producing the post. You can charge people a single dollar a month a make a revenue of $1 Billion by reaching a billion people. If 7 Billion read it you make $7 Billion a month without a change in your production cost!!!!
These businesses are brilliant they are the most efficient form of business the world has ever seen. This applies to things like Apps, Websites and Video channels. It also applies to non web platforms like radio and television.
Unfortunately like all good things this is a slightly utopian view of the world. At ultra large scale things break down once again. So let’s look at some popular companies and where things break down for them.
Google, Facebook, Yahoo
For these companies the ProductionCost and DistributionCost are very different. Their product is not a web portal, the product in fact is the consumer. Their real customer is the Advertiser paying to buy the consumers attention. The ProductionCost is the cost required for the web infrastructure.
Ecommerce — India vs USA
Though they seem to be identical businesses they really are not. Amazons business is real simple, host a website, list sellers, register buyers, stock products (a little complicated), give the product to UPS to deliver. I take the real simple thing back but just humor me. Whereas in the case of Flipkart it is not that straightforward, they were forced to build their own delivery arm which takes up a lot of resources. However when the economy gets to scale, Amazon still has to pay money to UPS per deliver, even though it is a small number. Whereas Flipkart at scale saves a lot of money by having it’s own delivery arm. The crux is that Flipkart is forced to build two separate businesses, one an e-commerce website which is a super scalable startup and another that needs but-loads of money to get to scale where it is sustainable. Sometimes even $1 Billion of funding just isn’t enough. Webvananybody?
Uber is not a taxi service, that is what enables them to provide it’s customers a great taxi service. If Uber owned every cab that its customers used, it would bleed to death. Though ebay is not as profitable as amazon, it is a lot more efficient as a business model. Both Uber and Ebay have no skin in the game. They aggregate people onto their website and allow those people to have a transaction. However, like all inefficiencies that creep into a system the both of them have to spend time in ensuring the quality of one of the participants in the marketplace. For Uber it is ensuring the drivers quality and some equipment and for ebay acquiring the sellers.
No discussion on startups is complete without talking about venture capital. Everyone seems to understand that venture capital is important but nobody seems to understand why. Never in the history has venture capital probably been as important or talked about as it is today.
The role of Venture Capital is very simple. Pay for the inefficiencies in the business model. All of the businesses we have seen have some ineffecency. In Flipkart’s case it is the distribution, for Google it server space and advertiser acquisition for Uber it is quality drivers and cars. What venture capital pays for is this inefficiency that takes away from a perfect system. The ideal business model would be to spend no money in building a business for which you get paid per person who buys your product:
Any inefficiency you add to your business model has no way to be paid for apart from putting in your own money, waiting for accumulating wealth from profits of the business or taking venture capital. The problem is that entrepreneurs don’t have any money and waiting for the business to generate profits is too long and or the inefficiency is too large for the business to solve through it’s own profits.
The more efficient the business model, the less the need for venture capital. The faster the growth. The closer you are to
and the more you can count yourself a startup. All of this assumes that a billion people want your product to begin with ;)