Shutting down my first startup

At the end of December 2014, I shut down my first company, Virtrue. We had one vision – to build a global identity. We are starting to live in a globalized online world – Facebook, Skype, Twitter and just about every other online tool knows and barely recognizes any borders yet our physical, “trusted” identity is still based on country/nationality.

The core idea for Virtrue started when I thought about how I could use my seemingly good “identity” to transact. Transact can mean almost anything – payments, purchases, communications, access, chat and entre into various groups.

Identity is the killer app (or “identity is the platform” per a fantastic post by Chris Messina) and yet it was stuck in the 20th century paper world. I could lie and say that I went to Harvard – most people know the University and implicitly trust someone with that credential. However, it’s hard to verify the veracity of that statement. As President Reagan said during the cold war, “Trust but verify.”  We set out to do that – reduce fraud, reduce friction, reduce fear and enable more positive trusting transactions globally – essentially add verification to every transaction.

With a globally verifiable, trusted and accessible identity, you could do almost anything. Without an identity, transacting becomes far harder – especially globally where no one knows your name (which is sometimes a good thing).

The core part to any transaction is trust. At Virtrue, we initially attempted to quantify trust. The challenge, we found, is that this is impossible. As an example, because I know your friend, you might be willing to meet me for a drink. However, is that friendship a close enough connection to make you trust me to babysit your child? Maybe and it depends – on you, on the relationship, on each individual.

Since you cannot verify trust, you have to verify identity components. Each action we take creates part of our identity. I went to school (one component).  I graduated (another component).  I give off indicators of those components that we will call data exhaust.  Example of data exhaust – my friends who all claim the same component (college and same graduation year).  You can look at the components and combine them with other data that we give off each day we interact online to verify those components.

In the old identity world, to verify my college attendance and graduation is an expensive and slow proposition, which is incompatible with the speed of global transactions. You could verify by proxy such as by searching for someone’s email or name on social services and seeing if, from what you could see publicly, they seemed trustworthy but that inherently had risks (i.e. fake profiles).

Virtrue developed tools, using social networks such as Facebook, LinkedIn, Twitter and other non social tools to verify that I actually attended Vanderbilt. We could complete this verification in under a minute for less than a penny. In theory, with that verification, I could now transact anywhere on the power of my Vanderbilt degree and connection. If people trusted Vanderbilt or had friends from Vanderbilt they trusted, they were more likely to trust me. Or at least that was the idea.

Virtrue began in January 2012. We followed the Lean Startup methodology. Before we built anything, we tested whether trust actually increased the value (or number) of transactions. How? We posted fake tickets on Craiglist and A/B tested our posts. Post A was just a listing for a concert ticket. Post B was the exact same post but with a badge at the bottom saying “verified by Virtrue.” The badge was even clickable and took interested people to a very official looking web page. Our results were bad. There was no statistical difference between A and B.

We ignored our results (lesson learned!). We applied to YCombinator. We did not get in. We raised some money (yay! Startup “success”). We applied to another accelerator – Orange Fab – and did get in (yay and thank you Orange Fab team!). We plowed ahead.

We were sure that the rise of collaborative consumption (aka the “sharing economy”) was our golden ticket. In the summer of 2011, AirBnB had a major incident when crack addicts destroyed a woman’s home in San Francisco. AirBnB did not have any insurance. They did not have any policy to deal with this issue. They could get insurance but, we assumed, how could they deal with the inherent loss in trust and potential fear this would engender in their community? Answer – Virtrue Verified Identity!

Virtrue logo

Not true. Turns out that we should have paid attention to our Craigslist test. People are willing to transact in a million different scary and dangerous ways if they a) want a deal (a big reason why despite no security, Craigslist is still huge), b) think they are part of a safe community, c) feel safe (even when they should not). How many times have you met a new date in a bar because you thought the other strangers around you provided security? Tons. And guess what, for a very large majority of people, they never ever run into a problem. AirBnB and the other sharing economy companies realized this – safety is in the eye of the beholder and most people are inherently trusting.

After exhausting nearly all of our contacts in the sharing economy, we had to pivot. We chose hiring. Many companies spend millions on background checks of their employees. These background checks are slow and not always accurate (they can miss a lot). Example A (and we used this in a lot of pitch meetings) – Scott Thompson. For those who do not remember, Scott was the former CEO of Yahoo (just before Marissa Mayer). He was vetted up and down (including a likely $20k+ background check) and got the job. An activist investor who wanted him fired did a bit more checking and guess what – Scott lied about having a computer science degree. Instead, he took one computer science course. Just like that, our pivot seemed primed for takeoff. Every company would now realize the importance of background checks but not all of them would want to pay a ton to complete them.

Except that we only found a few companies where speed and cost actually mattered. Turns out, a background check that is slow and expensive is also very defensive if an employee later turns out to have major issues. In a CYA (cover your ass) world, using a faster, cheaper solution when a slow solution is safe is a not a winning recipe.

By this point, we were running out of cash and keeping the team motivated was also becoming a bit more of a challenge but we had one more pivot in us – securitized transactions based on identity verification. Put another way, fraud in online transactions is a massive unsolved problem (although a few cool companies are finally doing some things to fix it). We jumped in. Virtrue’s tool could tell you if that buyer from, say Belarus or Nigeria was real or fake. Should you trust their purchase or decline it? We could do that. But…not enough companies have this problem. We learned that fraud generally hovered under 1% for companies that knew what they were doing (and the ones that don’t know what they are doing don’t really have a problem so don’t care). That’s not a big need.

Virtrue was out of time. No more pivots. In the fall of 2013, the team disbanded. Starting, running, developing, pivoting, raising – all an amazing experience. I lost a massive amount of money. It was a great education.

Global verified identity still has not been solved. Facebook sort of does the trick for some things. LinkedIn verifies others. Neither are trustworthy. Peer lenders and fraud preventers are using much of the data exhaust to help make smarter decisions about individuals. I still don’t have an online passport of trust that I can take to any country and use it to transact.

My identity is my data and that is capital. I use it to transact already. I protect it and invest in it. Some day, I’ll be able to transact.

Epilogue: At the advice of our smart lawyers, we kept Virtrue open for 12 more months. We were able to sell some of our assets. Our investors received some money back (not everything but I’m proud of how much given how little we raised and how long we were in business). I’ve received calls from many companies thinking about doing work similar to Virtrue. I happily provide advice. I want to see a verified identity become a reality even if my team and I are not the ones to accomplish it. The team itself has moved onto fantastic new jobs. I helped start Dispatchr. I helped a friend’s company through a really challenging period. I’m now at an amazing personalization company, which I view as almost the next step (but way different application) of data exhaust. Here’s to the future!


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s