The invite system for the OnePlus One
OnePlus jumped into the smartphone market in 2014 with it’s OnePlus One. Despite being a new smartphone manufacturer, OnePlus gained market traction through several decisions, one of which selling their first phone invite-only selling over one million units by the end of 2014.
In class we are studying game theory and the following model I made represents the two competing desires of the company; maximizing profits and market satisfaction (whether there is enough product out). The reason for the latter desire, is to maximize press and to further diminish attempts to resell the phone at a higher price if the market doesn’t experience great scarcity, especially applicable to the OnePlus One when it was sold only to those with invites. The two variables that can be adjusted are the number of phones produced (controlled by the company) and the number of phones desired in the market (controlled by buyers). Despite modeling the second player with a payoff for the company but completely out of it’s control, this model can give us an idea of the strategies the company will take to respond to the various possibilities. Now back to the model, Let’s say the phone is sold at Y, costs X to produce, and a to a+bn represents the number of phones/phone-buyers where n is some integer (greater than 2). We can represent the payoff matrix as follows:
| OnePlusOne (profits), satisfaction of those seeking oneplus one phones | Phones desired to be bought by market | ||||
| Phones made | a | a + n | a + 2n | … | a + bn |
| a | a(Y-X), 1 | a(Y-X), (a/(a+n) | a(Y-X), (a/(a+n)) | … | a(Y-X), (a/(a+bn)) |
| a + n | a(Y-X) – (nX), 1 | (a+n)(Y-X), 1 | (a+n)(Y-X), ((a+n)/(a+2n)) | … | (a+n)(Y-X), ((a+n)/(a+bn)) |
| a + 2n | a(Y-X) – 2nX, 1 | (a+n)(Y-X) – nX, 1 | (a + 2n)(Y-X), 1 | … | (a+n)(Y-X), ((a+n)/(a+bn)) |
| … | … | … | … | … | … |
| a + bn | a(Y-X) – (bn)X, 1 | (a+ (b-1)n)(Y-X) – (b-1)X, 1 | (a+ (b-2)n)(Y-X) – (b-2)X,1 | … | (a+bn)(Y-X),1 |
Without knowing the price vs. cost of the phone we see that there is a pure Nash equilibrium to go with the minimum number of phones produced (a) to maximize profits in the case there is the minimum number of phone-buyers (a). Although the company could potentially make more profits if they produced more phones, they do not know if they will have enough buyers to make producing that many worthwhile.
Additionally if we say a=1000, Y = 400, X = 100, and n = 100 we could model the matrix as below:
| OnePlusOne (profits), satisfaction of those seeking oneplus one phones | Phones desired to be bought by market | ||||
| Phones made | 1k | 1.1k | 1.2k | … | (1 + b/10)k |
| 1k | 300k, 1 | 300k, (1/1.1) | 300k, (1/1.2) | … | 300k, (1/ (1+b/10)) |
| 1.1k | 290k, 1 | 330k, 1 | 330k, (1.1/1.2) | … | 330k, (1.1/(1+b/10)) |
| 1.2k | 280k, 1 | 320k,1 | 360k,1 | … | 360k, (1.2/(1+b/10)) |
| … | … | … | … | … | … |
| (1 + b/10)k | (300 – 10b)k, 1 | (300*1.1 – 100*(b/10 – 1))k,1 | (300*1.2 – 100*(b/10 -2) ,1 | … | (300*(1+b/10))k,1 |
So without additional estimates providing OnePlus data of desired buyers they would have to start with the smallest inventory of phones, slowing it’s market growth, but fortunately due to the invite system, they can make estimates of interested phone buyers. Although the invite system doesn’t give OnePlus enough time to know the exact numbers of phone buyers as they decide how many to manufacture, requiring some predictions and estimation, the data known allows them some leniency to make choices that appear unstable in the payoff matrix. I would like to warn that the above model does not account for changing price/demand due to surplus/shortage of the phone and other factors, this was just to illustrate the advantage of being able to estimate the units needed to be produced with additional data vs starting off with none.
Lastly the spread of the invites plays a role in graph theory as well as a common concern was whether or not OnePlus lost sales due to customers not being able to buy the phone. With a model network where the network is made up of nodes that represent people who have the phone, those who want it, and those who don’t want it. Edges will connect nodes to represent various relationships between the various people in the network. If the network is well connected and there is a great enough number of people interested in the phone, the phones will approximately reach everyone who is connected to a few individuals who want the phone. As a former owner of the phone, I was worried that I wouldn’t have had the chance to own the phone. Luckily a few months later a family friend had received the phone and sent me the invite I wanted and I was ecstatic to own the “flagship killer” of its time, hearing from friends how the phone was a great balance in performance and price. Even if a node in the network was isolated from other nodes who have/want the phone OnePlus hosted several other opportunities for those interested to earn an invite.
Also below is an old blog post related to the Oneplus One from 2015. Although we have one article in common I made sure to diverge away from his topics by looking at how OnePlus’ invite system can give the company an advantage in determining the number of units to produce.
https://blogs.cornell.edu/info2040/2015/11/14/the-success-of-oneplus-one-marketing/