Pete Michaud

Talk mathy to me, baby

Fork Pete Michaud on GitHub

Build Meta

November 5th, 2013

TechStars Austin just finished up. I met a lot of super smart and passionate people over the course of it, and I hope they all make a trillion dollars.

I’ve noticed something about companies that get investor attention that I’m going to share.

I was talking to a VC some months ago about a social good startup. She had funded a social good startup before, and I wanted to know what she looked for.


When most people think of a nonprofit or social good enterprise they think about feeding the poor, or educating underprivileged kids. What she told me was that she funded a platform that addressed some inefficiencies in the nonprofit space.

Kiva also comes to mind. It’s a platform allowing people to make microloans to improverished people across the world. Kiva doesn’t make micro loans, they just make such a thing possible.

But this pattern isn’t limited to social good startups and not for profits: companies that get funding tend to be “meta”. They don’t make better Widgets, they solve inefficiencies in the Widget market. They disrupt outmoded Widget middle men, or they seek to become a necessary but missing Widget middle man.

Making Widgets is often a great business, but in the funding world it seems they are mostly considered “lifestyle businesses,” ie. they can make a small living for the person running them, but they will never make the gangbuster profits that VCs are hoping for.

Hidden Meta

Even funded companies that appear to be Widgets often have a meta play. For example, there’s a website you may have heard of that allows you to stay connected to your college social life. Frats and sororities used to have individual facebooks for their members, but a kid realized that that wasn’t as useful as an abstract sort of facebook that could be used by everyone in the college. That was a meta play in itself.

Still, it’s basiclly a widget that let’s you talk to your friends. But what would happen if everyone–literally everyone–were on the site? And then we started keeping track of who was connected to who in a sort of social graph that shows the social topology of our entire civilization? What if the meta data included our tastes in media, our previous significant others, our buying habits?

A widget for talking to your friends is not worth $100bn. The connection data between all the people in the western world very well might be. The widget is not the product, the widget is the machine that makes the product. The “product” of Facebook is the meta data.


Let’s take the recent batch of TechStars companies:

  • Accountable - service that makes HIPAA compliance very easy for any companies in the huge market of those who have to comply.
  • Atlas - a wristband that tracks physical activity with high fidelity.
  • Embrace - platform to track customer interaction with your company, so you can create personal and lasting relationships with them.
  • Filament Labs - address issues with patient/doctor communication across several markets.
  • Fosbury - an abstraction layer on top of virtual currency and payment systems.
  • Gone - easily sell your unwanted stuff to someone who wants it.
  • MarketVibe - link related content across any data source.
  • ProtoExchange - a platform that connects customers to manufacturing facilities.
  • Testlio - a platform for outsourcing software quality assurance.
  • Ube - a new kind of light bulb and lighting control product.

Literally all these companies address meta issues in their respective marketplaces except the 2 hardware companies. The hardware companies address potentially giant markets.

That’s really the crux of the issue: meta companies address markets that are potentially orders of magnitude larger than the companies they serve, plus it’s easy to use the core asset of meta companies, which is usually data, to create new streams of revenue for very little additional capital.

My understanding from talking to hardware startup people is that it’s incredibly difficult to raise the kind of capital required to build hardware at scale. On top of that, it’s hard to develop a new market for something universal like a lightbulb–the space is already crowded, mostly by giants. Meta is both easier and cheaper.

It’s not just TechStars either. Off the top of my head: Google, Twitter, Airbnb, Heroku, Wufoo, Reddit, Amicus, Bufferbox, Coinbase. All meta. If you go through a list of Y-combinator companies you’ll have a hard time finding any that aren’t meta.

So if you want to get funded, consider building meta.