[JumpCrisscross]

37signals is pricing this deal with the assumption that there will be a significant (>50%) departure of customers upon the deal closing, that cash flow growth thereafter will be meagre (<10%), and/or that the whole thing will collapse within a few (~3-5) years.

If we assume a $480 000 purchase price, annual cash flows of $212 277 (12 times the mean monthly, which is fairly normally distributed save for March 2012) growing at 2% a year, and abandonment after 7 years the investment yields over 10% (IRR) so long as cash flows crash less than 56% from purchase to year 1. In other words, if first year cash flows are at least $93 000 and grow 2% a year from there, the investment will generate a 10% yield after 7 years.

This analysis is stylised, but it remains that even if one takes a 1/4 drop in cash flow from purchase to year 1 and manages to lose 5% of cash flows each year thereafter, the investment will earn a 23% yield assuming abandonment after 7 years.

[moe]

the investment will generate a 10% yield after 7 years.

Except 7 years ago the most popular social network was called MySpace. The iPhone, Twitter and ycombinator did not exist (amongst other things). Rails 1.0 was released that year.

Why would a niche job board, detached from the brand name that was its only asset, prevail through 7 more years of internet time?

[JumpCrisscross]

Fair point, 37signals may also expect the whole thing to organically fizzle out within a few years (updated).

Let's constrict the time frame to 3 years from purchase and continue the abandonment assumption. Assuming 2% annual cash flow growth we cannot sustain more than an 11% initial crash before yield falls below 10%. With -5% annual cash flow growth this number drops to 4%. The investment is probably not reasonably viable with a 3 year time-frame.

Note that a more nuanced analysis would ask for a more rigorously thought out required rate of return (in this latter case 10%).

[jasonfried]

I think a smart businessperson could blow it up much bigger. Open it up to a variety of different industries. Wedding planners, photographers, florists, interior designers, architects, etc. Think big.

[moe]

Wedding planners, photographers, florists, interior designers and architects have no idea who 37signals is. Consequently they won't be fazed in the slightest by a simplistic job board with relatively poor usability entering their respective markets.

Furthermore, even if they knew who 37signals was - that brand is removed from the product after the sale.

Alternatively, if you seriously believe the success of your job board is due to technical excellence and would map to other markets. Why would someone buy the tech for half a million dollars instead of replicating it within a few months for under 100k dollars? Using the very same tech staff that he will need anyway to operate it?

[JumpCrisscross]

To be clear my cursory analysis does not rule on the goodness of this deal for a buyer who believes they can affect the growth variable, it was merely feeling for the edges of plausibility.

Assuming the buyer cannot significantly mitigate any loss of customers resulting from the sale, I looked at the interaction between that crash and expected return. I assumed lower growth rates in my analysis based on the low growth and deviance in the historic cash flow data.

If a buyer got hit with a 1/3 initial crash but believed they could growth cash flows thereafter by at least 10% per year for 5-7 years they could expect an internal return of 21% to 30%, independent of the capital gain on the larger, growing company at the end of the road. The game naturally changes if one ignites such non-linear growth.

[astrofinch]

Sure, but why start by paying half a million dollars for a used codebase?