The (dis)economics of Digg scaling

From Business week article about Digg:

Last year the company lost $2.8 million on $4.8 million in revenue, according to Digg financial statements reviewed by BusinessWeek. In the first three quarters of 2008, Digg lost $4 million on $6.4 million in revenue. Adelson declined to comment on the figures.

Thats an increase in loss from 58% to 63% of revenues as it grew them by 71% - ie the loss increased by about 1 % point for every 15% growth. Clearly there do not appear to be economies of scale operating here, so the bigger Digg gets, the more it loses.

Assuming Digg grows by the same % again next year to c $11m it will thus lose between $7m (keeps ratio) and $7.5m (increasing rate of loss).

Out hypothesis is that this is because the cost of supporting each transaction is higher than its revenue benefit, and that the cost is growing at a faster rate than the benefit as network complexity sets in (Metcalfe's Law has a dark side - transaction costs also increase with scale, so beware if revenue starts to reach a diminishing return point) without being able to raise revenue in line with this

So the issue for Digg is how to push that diminishing rate of return further up the growth scale. One approach that may work to good on the revenue and cost side is filtering out the cr*p early. Because those transaction costs are non-zero, the Web 2.0 dogma of "publish, then filter" breaks down after major scale is reached. Less cr*p would probably also bring in more high end readers, who avoid Digg due to - well, the cr*p content.

Update - looks like SAI has picked up the same observation this morning.