I have had this post ready for months, but never sent it off.
I am bummed to have to admit that I am shutting down Hangout – the first 3D experience on Facebook. The original vision of Hangout started in 2006 around the notion of media sharing with your friends and leveraging api’s on Facebook that were not even public yet. I raised capital quickly from world-class VC’s, built a team, became a Techcrunch 50 finalist, pivoted, raised more capital from those same, patient investors, hired a new team, re-launched – and yet nearly 4 years after coming up with the idea — I am shutting it down.
This was not my first company, but it was my first failure. What went wrong?
- Its as expensive to be too early as it is to be too late to a market
Do you know the saying, “Don’t go shopping on an empty stomach?”. Well my vision was much bigger and more optimistic in terms of timing than has occurred. I was convinced that teens in virtual worlds and on Facebook would prefer full 3D with all of its higher quality animations and immersive behavior over flat flash games. Although no one had yet demonstrated success leveraging 3D in a browser, I was convinced we could.We were a very early adopter of Unity — a browser-based 3D game engine, which eliminated the need for a big download, which in turn should have improved our conversions compared to heavyweight downloads like Second Life.
And we could prototype the experience very quickly, which only got our investors even more excited (see below raising too much money too quickly).Little did I fully realize that any sort of download increased the user dropoff dramatically and put even more pressure on building a successful user experience (virtual world or game). The economics of increased dropoff gave me and my teams no room for error. We had to thread a needle. No margin for error. Every user who got through the download process had to be far more engaged than a typical flash game player — and as I describe below — that was not the case.So what do I take away from being too early? Well, the lesson is actually more sublime. In fact, earliness is a symptom of a more fundamental challenge with startups.Add to the dropoff challenge, Facebook changed their virality algorithm, effectively tripling the cost of acquiring new users to the game. Whereas a decent viral game could acquire 1 or more users for each player that entered the game, the new Facebook logic reduced the virality to 1/2 to 1/3 of a user for each new user, dramatically increasing the cost of growing your userbase.
LESSON #1— If you are early, do not differentiate on too many variables at the same time. If you look at some of the great successes, the variation on their first product was on one, not two or 3 axes. Maybe the differentiation is on technology or maybe on function, but rarely both at the same time.For Google, it was better search than Lycos or Yahoo – the base function was the same — searching the internet — it was the technology that was different. They did not come out with Adwords, etc., until they had proven that their core product was better and audiences wanted it. For Facebook, it was a friend-driven social experience than MySpace. There was no technology risk to Facebook — just feature risk. For Mint, it was about making a web 2.0 version of Quicken. The functional concepts were well understood — it was the technology risk of getting users to trust a web service with their financial data that was the risk.
I am sure all of these analogies break down at some point, but I believe that the underlying differentiation of each of these examples was on 1 not multiple axes.If you are going to differentiate on both technology and product, you better have a pile of cash and lots of time because you will need it to isolate the variables that are causing success or failure and focus on those. The more variables in play, the longer it will take to isolate the key variables and find product/market fit.
- In gaming, the location of your team makes a huge difference
Team and Location — Building Facebook games, the team needs to be near the ecosystem – SF/Palo Alto. Be as close as you can to the epicenter of the business, especially if you are leveraging a platform that changes frequently, such as Facebook where exploitation of the rules was an art form and begging for forgiveness was easier than asking for permission. For some platforms, location itself near the platform is irrelevant. For platforms like Apple where the rules are well established and seem to be consistently applied, location is less critical. Everyone is at the same disadvantage regardless of location.
Sure some have thrived outside it, (Wooga comes to mind), but the inside knowledge and the whirl of activity of the last 3 years in this space has meant that if you were not out there, you were at least 1-2 months behind your competition in deploying the latest tactics and learnings.
LESSON #2 Don’t underestimate the power of location when building your next company.
- In gaming, your team’s domain experience matters alot —
Despite warnings by one VC to the contrary, I was supported by many who said that I could build the team here in Boston to build the first version of Hangout — a media sharing platform and virtual world. Boy was I wrong. I don’t mean to beat on Boston, but the virtual world and gaming market, and especially Facebook gaming, is not an obvious strength.I hired a smart web team because I believed that the experience was about web first and gaming second. But I was wrong. To attract and retain customers, we needed increasingly good game designs and game designers. Beyond poking or throwing sheep, Facebook games have increasingly required good game designers. I just never got that memo.And even when I pivoted and moved the company to LA to work under a well-known general manager out of Disney and his team, I still did not understand the difference between a top game designer and a GM or a Chief Creative Officer or whatever other title you want to make up.
LESSON #3 — Make sure you have team strength in the domain you are going after. If its gaming, make sure you have a top quality, experienced game designer. If its a consumer app, make sure you have good UX strength. If you insist on learning on the fly, expect everything to take much longer and consumer more capital than you planned. And that is on top of already doubling the time and capital of your original forecast because you will always be wrong in your first forecast.
- The more money you raise, the higher the expectations and the shorter the time frame to hit those expectations.
VC’s need to make a return on their capital. They have Limited Partners that they have to answer to. The more capital they give you, the more they expect you to put it to work. You raising it and banking it is not what they want. The capital should be put to work.There is an old saying raising capital that says you basically can never raise too much money. That is not always the case. I raised too much in my Series A on the expectation from my investors that I would use it. Short of an economic downturn (which also happened), VC’s expect you to execute on the plan they invested in.If you change direction, they are usually willing to go along with change (see Pivot lesson below), but you need to know how to pivot.
But I had a more basic problem — I was not ready to spend the money because the product was not ready to be scaled. It had not achieved product/market fit when I got the capital. So instead of using the money to begin scaling, I was using it to build the product. That is not what a Series A investment is for and I think most investors would agree after the 2007 investing bubble burst in 2008.There still remains the question of whether you should raise seed money from VC’s. Again I would come back to the expectation cycle of the VC’s. Unlike most Angels, most VC’s must answer to their LP’s. Thus, I have a hard time believing that VC seed investing is a pattern for the future. The pressure to grow a seed is as great as growing a Series A company with a lot more product risk and insufficient amounts of capital to make their returns to their LP’s (Fred Wilson has a great blog post on the outsized returns needed on large funds).LESSON #4: — Raise money from VC’s when you have Product/Market Fit.
- If and when you pivot, pivot hard and fast.
Back in early 2008, I had the idea to build a Facebook game that would showcase our differentiation — 3D, physics, etc. I built a prototype basketball game and showed it to a bunch of WC entrepreneurs who were in the Facebook ecosystem. They were blown away and said I should do everything I could to put this game on Facebook. I had also showed it to the NBA and they wanted to sponsor it — they offered to spend $250K to drive traffic to it. They offered all of their licensable content –uniforms, logos, etc. — for free!!I thought I had hit a goldmine. So I brought it up to the board and caught a bunch of questions about moving into the gaming market when my plan was to build a media-sharing platform.Now at this point, entrepreneurs who read this post might shake their fist and say, “See another company micromanaged by a VC.”THAT is not the case here. I made the critical mistake of pivoting too softly.I built the game and released it, but I did not have the cojones to completely pivot in this direction. I allowed perfectly reasonable questions to feel like commands, which was not the intent of the VC’s.
Their job was and is to ask questions and to test my certaintude around the pivot.In fact, I was not yet convinced that Facebook gaming was where we should be. Was it a shiny new object, or a real, up and coming market? So when I explained to our team that we were going to build a Facebook game for the next 4 weeks instead of focus on our media sharing platform, I couched the work as an exercise, or prototype instead of as a full pivot.As a consequence, the game was built, but not with the complete devotion and energy that a startup needs to have when building for a market. It was a half-assed attempt to experiment. And after the game was built around March Madness, we abandoned the game as soon as the NCAA playoffs were over and went back to building the virtual world/media sharing platform. A big opportunity was missed because I did not act out my instinct. I allowed a few questions to build uncertainty into my head and avoid the hard decision of changing direction.
LESSON #5 — When you pivot, pivot hard and quickly and don’t look back.
I had excellent VC partners and great team members — all of whom I would work with again, in the right circumstances, building the right product at the right time. Nevertheless, there are reasons that so few startups make it. Even though you have ridden the horse successfully a few times, there is no guarantee that you will not fall off on the next ride.