SaaS Start-Ups and the 7 Year Rule

Autumn 2015

The Jesuits proclaimed "Give me the boy for the first seven years and I'll give you the man" but what about the start-up? What does the seven year old start-up reveal about the future? Not only for itself but its generation and maybe even future generations?

To find out we have mapped the cost of execution by the cost of invention for a sample of IPO listings over the past 30 years.

In this first visualisation we have illustrated them by type (i.e. SaaS, Social Media/Search/Media/Games, Hardware and eCommerce).

In this second visualisation we have illustrated them by era (i.e. 1980's, 1990's 2000's).

The patterns seem to suggest that media & eCommerce is more efficient than SaaS.

So let's explore the question is SaaS becoming more or less efficient over time?

In this visualisation we have summed up the efficiency of 7 year old SaaS start-ups by comparing operating profits against the % of revenues spent on R&D and Sales and Marketing and then distributed them over 5 year blocks.

The trendline suggests that SaaS startups are becoming less efficient. But is that simply a function of cheap money (i.e. Inefficiency peaks when the market is awash with cheap money to invest in start-ups) or are there other factors in play?

To explore that question we redefined the distribution as Client Server, Dot Com and Social Media Eras

As you can see the trendline suggests a new pattern. Now it would appear the network economy is becoming more efficient.

The cost of discovery (i.e. execution) is still priced at a premium but the cost of invention is falling. So much so that the cost of innovation, although historically very high, has flatlined. Suggesting that SaaS (hereto a money pit for investors seeking to disrupt the Enterprise Client Server Market) may even become a break even proposition over the next decade (i.e. the Mobile Era).

Update: 2023

It's been 8 years since this forecast was posted

How did we do?

If we update the map with the 7 year IPO data since 2015 we discover the identified trend was debunked by the increase in data points

Basically it appears the early IPOs of the Cloud era were more efficient than the later IPOs skewing the data sets for the cloud era

Once we factor the late IPOs into the equation we discover the trendline established between the 80's and the turn of the century continues as the CAC continues to accelerate

The combined cost of Sales and Product Development continues to increase

With the Cost of Acquisition accelerating dramatically and the cost of R&D flatlining... largely because of the explosion in activity fueled by the perception of "zero barrier/cost of entry" across all things digital creating extreme pressure on the skills market for developers