Hacker News Reader

Comment by relaxatorium | Hacker News Reader

Thinking about New Relic vis a vis a potential startup bubble makes me think about the old saw about the gold rush where most of the people who got rich were the ones selling pickaxes rather than gold miners.

On the other hand, in the current startup economy the pickaxe bubble is almost as big as the gold mining bubble, but I think it still probably applies more than it doesn't.

View on HN

I take it you weren't around for the Dot Com Bubble? I can't tell you how many times I heard this repeated -- and used to justify lofty valuations for infrastructure companies. CSCO, JNPR, ORCL, INTC, EMC, NTAP were all "selling pickaxes" -- none of those valuations ever recovered. And those are the companies that survived; many companies "selling pickaxes" perished through either bankruptcy or (more frequently) fire sale acquisition. There is emphatically a valuation bubble; it will burst; when it bursts, essentially all will suffer -- and life will go on.

Disclaimer: I worked for SUNW "selling pickaxes" before, during and after the Dot Com Bubble, I told myself this lie -- and vowed to never tell it to myself again.

Chart of valuations the curious: https://www.google.com/finance?chdnp=1&chdd=1&chds=1&chdv=1&...

I was not, and for that matter, am not really participating in this one either except as an observer.

I wonder if there's a difference based on how many of those companies were selling mostly actual hardware infrastructure, which has just gotten cheaper and more commoditized over time, versus those which had actual differentiated software.

Both still got hit hard of course, a bubble's a bubble, but I wonder who did substantially worse?

This time around the software is getting commoditized (with open source).

There were some open source software players back then [1].

VA Linux (now Geeknet) was near a $10 billion market cap [2] in trading immediately following its IPO. After several pivots and renames, it's market cap now is just $50 million.

RedHat went public at about a $1 billion valuation (raising $100 million), hit about $3 billion in trading after the IPO open, and has steadily grown to about a $10 billion cap over the last 15 years.

Meanwhile, today Hortonworks ends the day with a cap over $1 billion, and a couple of competitors in the wings that may also be doing IPOs at some point.

It's not that Hadoop and Linux are even comparable, but it's really hard to see from a 2014 vantage point that the Hadoop market will be as disruptive over the next 15 years as it would have been to see Linux disrupting in 1999. So Hortonworks and its ilk fall more in the space of advanced enterprise software--a more comparable relative temporary winner from that era was BEA Systems (which also had a spectacular rise and dramatic fall--amplified because they were a company that was disrupted by free software, and suffered due to unsustainable expectations). [3]

[1] http://www.cnet.com/news/10-years-gone-the-va-linux-systems-...

[2] http://www.theregister.co.uk/1999/12/10/ipo_makes_va_linux_i...

[3] http://online.barrons.com/articles/SB998410322172345858?tesl...

The funny thing about working for a company like this is that you're still placing a bet on it. If you're selling pickaxes, you're fine ... but New Relic is selling a subscription to pickaxes, which will dry up and go away if the gold rush disappears.

So, yeah, pickaxes are great. "Netflix for pickaxes" is more strongly tied to the market's long-term future.

Reply on news.ycombinator.com