Myth Versus Data in the Financial Markets

By Ellie Fields 02 Sep, 2009

As we’ve all learned over the last year or so, the financial markets are as full of myth, hyperbole, and occasionally outright distortion as any human institution. The market for IPOs is no different.

The traditional scenario is this: a company builds hockey-stick projections showing amazing growth in a few short years to impress potential investors. Venture capitalists kick in funding, and if the company is successful, the VCs, employees and investors all do well. That growth is realized-- but typically in a much longer timeframe than initially projected.

We analyzed the financial statements of over 100 companies to bust the myth of the hockey stick revenue projection. How long does it really take to build a successful company? Of course, being Tableau, we visualized the analysis:

In fact, for most successful tech companies, it took about 9 years to reach $50 million in sales.

The results are published in a new sister blog, IPO Dashboards. We’ve also just posted on what this and current market conditions mean for the VC business model, which typically assumes much faster IPOs. IPO Dashboards also uses some cutting-edge web technology that we’re calling tableau public for now.

Being data geeks, we love busting myths and answering interesting questions. But we were flattered when the Wall Street Journal decided to run this visualization in one of their own blogs, Venture Dispatch. Take a look at IPO Dashboards and if it sparks your interest, leave a comment.


Submitted by Colin Mason (not verified) on

Can you analyse the performance of companies by the type of financing they had prior to IPO? In other words, it would be interesting to know the proportions of rocket ship, hot companies and slow burners that had raised venture capital or had raised angel finance prior to IPO.

Submitted by Cindi Howson (not verified) on

Really interesting! and of course, I loved that I could filter down to Business Intelligence vendors. Would of course be interesting if and when Tableau IPOs!

Submitted by Dr. Richard (not verified) on

This dataset is skewed because it cherry picks known companies that reached the metric. Hockystick exponential growth from nothing to $50+ million has happened to a few companies but these are only a tiny subset of the set of venture funded companies.

You really need a fuller dataset that has every venture funded company, including the failures and ones that are revenue positive but not $50 million per year. The problem is that data is generally not available for most venture funded companies except those who go public. The majority of venture funded companies never go public -- they get acquired, they go bankrupt, or they operate as private (non-reporting) companies.


Submitted by Ellie F. on

@Colin: Good question. Of course, pre-IPO companies aren't required to provide as much financial information to the markets, so we don't have that data. Would be an interesting analysis though.

Submitted by Ellie F. on

@Dr. Richards: You're absolutely right, this set of companies suffers from selection bias of being the most successful software companies.

We have made an effort to track what happened to all IPOs here:

we can gather information on whether they were acquired, went bankrupt or whatever. But as you said, tracking all venture-funded companies would be nearly impossible because the data is not available for most pre-IPO companies.

Submitted by Christian (not verified) on

One can only determine selection bias by considering the question being asked.

In this case, there is indeed selection bias if the question is, "How many companies reach $50m and how long does it?" However there isn't selection bias if the question is, "How long did it take the world's most successful software companies to reach $50m?"

I wanted to answer #2 because it was easier. But equally importantly because most boards don't argue about how slow you can grow. They argue about how fast you can possibly grow. This sample is a great sample for estimating the upper limit.

Submitted by Chuck M. (not verified) on

The table in the "Company Details" section has a column titled "Avg. Year Founded". Should be just "Year Founded".

Submitted by Brad E. on

Your 1/27/2010 job postings indicate that Tableau is looking to go public. When should I start talking to my broker?