I would like to return to the issue of the value of LinkedIn. This is prompted by the fact that I have been reading Tim Harford’s excellent book The Undercover Economist. The opening chapters of this book reference the pioneering work of the 19th century economist David Ricardo and his investigations into price (income), costs and the importance of what happens at the margin to determining the relationship between the two and thus profitability.
Ricardo highlighted the importance of marginal costs (i.e. what it costs to produce one more unit of production or add one more customer) in setting prices and thus generating revenue and creating profits. Very basically, what Ricardo showed was that in a relatively competitive and open market, you cannot generate revenue from your product that is significantly more than the marginal cost.
In 2010, LinkedIn revenue was $243 million (according to Mashable) and in March 2011, LinkedIn claimed 100 million users. This means that revenue per user is around $2.43. According to Ricardo this should indicate that marginal costs will be close to that i.e. the cost to LinkedIn for adding an additional user and servicing them for one year should be not be significantly less than $2.43. If they were, LinkedIn is either not operating in an open and competitive market or else something else rather odd, and probably unsustainable, is going on.
What then, might be the marginal costs for LinkedIn? This is hard to calculate, but we can make a guess based on what they have to cover. First there is the server space and any technical and administrative support required to service one additional user. One would guess these would be tiny – much less than $2.43 per additional user, probably measured in cents rather than dollars. LinkedIn currently claims to be adding 1 new user every second. Are its’ costs therefore rising by around $2.43 every second or $86,400 every day? I suspect not. Then there are the costs associated with unlocking the revenue that an additional user represents – basically the cost of people to sell the advertising services. This is quite a strange cost because it is not really a true marginal cost as I see it, it is simply the cost of exploiting the advertising opportunity that a user represents – a cost which is essentially discretionary, i.e. LinkedIn doesn’t have to do it (or much of it) to deliver the service that LinkedIn represents to its users. The revenue LinkedIn generates from its users simply goes to pay for the cost of generating the revenue in the first place, or alternatively straight into profit.
Is it realistic to assume that the majority of that $2.43 goes to pay the costs associated with unlocking the marginal advertising revenue opportunity? Well, LinkedIn is not making huge profits – according to Wikipedia its’ net income in 2010 was $1.85 million. These same set of Wikipedia figures also said that revenue from January to September was $161 million (equivalent to $214 for a full year period) and the number of employees was 1,000. If you guess that the average salary for an employee was around $60,000 and also that LinkedIn is the sort of business where total costs are likely to be in the region of 3 to 4 times salary costs i.e. $180 million to $240 million, these figures all seem to make broad sense, as does the conclusion that most of LinkedIn’s costs are associated with unlocking the advertising revenue opportunity, not actually linked to the true costs of providing a service for additional users.
What can we then conclude – especially about the potential for LinkedIn’s future revenue and its current valuation?
Even if we assume that all of LinkedIn’s marginal costs are ‘true’ marginal costs it is highly unlikely that LinkedIn is going to be able to generate significantly greater marginal revenue from its users. The only way it could generate greater marginal revenue would be by creating a significant level of additional scarcity around the resource that LinkedIn represents – essentially increasing demand for its service (without increasing the marginal cost of providing that service and thus improving the service itself) while restricting levels of competition. If, however, we take the true marginal costs of LinkedIn to be only the server space and technical support necessary to add an extra user – LinkedIn already has levels of revenue which are unsustainably high in an open and competitive market.
The other conclusion is that I have got my estimates and understanding of marginal costs completely wrong!
But if I haven’t got this wrong, Ricardo would predict that a future where LinkedIn lives up to its current valuation of 20 or 30 times earnings is never going to happen – because it is never going to be able to generate the revenue this valuation requires (in an open and competitive market). Alternatively he would predict that LinkedIn will be replaced by a competitor that has a business model based around the true marginal costs of the business – i.e. a competitor who can do everything LinkedIn does but a much lower cost. The only other conclusion he could come to is that LinkedIn operates in a market which is so fundamentally unique that it can sustain businesses where marginal revenues are many, many times greater than marginal costs.
Perhaps those clever chaps in the City and on Wall Street should start to look at a marginal cost model when they come to try and work-out a valuation for properties such as LinkedIn, Facebook and Twitter. And if they don’t want to use a marginal cost model, perhaps they should explain what they know which David Ricardo did not.