My Winding Road to Lean Startup

Giff Constable startups

windingroad-romeo66
Last fall, I was recovering from a startup that almost touched the sun, but like Icarus, took a nasty fall. I still had a burning desire to create a great company, but I  knew that it was time that *I* chose what was right and wrong, rather than work for someone else.  If I was going to chart my own path again, I needed to find a better way to do entrepreneurship.

It might sound hokey, but the last 9 months have been empowering. I have taken my startup scars and successes, mapped them against the ideas of people like Steve Blank, Eric Ries, and Sean Ellis, and become a significantly better entrepreneur for it.

Next Friday, I will give a talk on Aprizi’s approach to customer development.  In thinking through that talk, I could not help but think through how I got here. Below are peeks – my personal interpretations — into 4 startups I worked for.

Trilogy: Sell, Build, Design
Trilogy was an enterprise software company run by a smart, aggressive guy named Joe Liemandt. I joined straight out of college back in 1994 around the 75th employee mark.  My colleagues and I used to joke that the company ethos at that point was “sell, build, design”.  In some ways, the approach had commonalities to today’s lean startup advice, where you prove market demand first, get something crude out, and then improve it over time.

The big downside to the “sell, build, design” approach is the painful period during which you have unhappy customers who bought into a big vision, but are stuck using an immature product.  At the time, Trilogy liked the “big sale” with a large ticket price, and that meant making a lot of promises. Unfortunately, overselling caused customer support headaches and brand reputation problems.

There was a lot Trilogy did right, but I was left with the opinion that a little more balance would drive better customer satisfaction and one would come out ahead.

Epilogue: I left the company at the end of 1995 to learn finance at an M&A boutique, and did not get to observe how Trilogy fixed these issues. The company ended up having some high-flying days, growing to one of the biggest privately-held software companies in the world, but I don’t have a good sense of how it is doing today.

Envive: Fake It Till You Make It
I joined Envive in 1997, which made systems management software for SAP installations, around the 20 employee mark.  Envive was a classic VC fumble.  It checked off some VC favorites: the founding team had consulted for SAP (credibility and domain knowledge), had special access to SAP source code (unique technical edge), and a high-vision, reality-distortion-field founder (who unfortunately was not the best operator). It raised a bundle of money from two top-notch Valley VCs with big promises, and started expanding the team to build and launch the product. When the VCs realized they had backed a little too much promise and not enough reality, they retaliated by pulling the founder from the CEO slot.

I joined (originally to lead product marketing) just after this happened, foolishly discovering all this too late. Information flow was less efficient back then, but the real problem was my youthful naivete assuming that I could bank on the due diligence of first class VCs.

Envive talked to lots of prospects and had a good set of beta customers, so they didn’t entirely fail on the customer development side. However, it was  a pretty classic “4 Steps to the Epiphany” example of scaling too early. There were too many employees for its stage, and it wasted a lot of money on expensive direct sales guys who could not translate their success at “big software co” to an early-stage product. I was one of the few people who could close business because I came from the product side, was close to the developers, and showed the depth and flexibility needed to build trust with early customers.

Epilogue: Envive ended up creating some innovations that led to the assets of the business being acquired, although not for a fortune. The company faced serious platform risk from SAP itself, who had their own lightweight product and caused a lot of customers to sit on the fence, but I believe that with a more controlled burn rate, it might have been able to forge a path to better markets. That would have depended on the board’s support for risky pivots and putting product-market fit ahead of growth.

Ithority: Design, Build, Launch
Ithority was the first business I co-founded and ran; Aprizi today being the second. We started it on the side in 1998, with the mission of creating an online marketplace for freelancers and services, and then pivoted downstream to focus on information. Our mistake was swinging *too far* from the Trilogy approach.

We were frustrated with the half-baked products of previous companies and wanted to make something really good. While we “got out of the building” talking to customers, it took us far too long to get a working product in the hands of customers.  We failed to produce what people now call an MVP (minimum viable product) and that was a mistake — it could have made a big difference for us.  Then again, it was a strange, irrational time.

Startup quality in 1999 was largely determined by how much money you could raise and how quickly you could spend it.  I don’t miss those times at all, and I don’t think the bubble makes for very interesting case studies because it was so warped.  My co-founder and I both got our starts at Trilogy, where the CEO respected operational cash flow, and we didn’t want to play that raise-and-spend game.  We bootstrapped the business with a little friends-and-family money until being acquired by a company called Opus360. Opus360 was one of the last companies to go public before the crash, and we watched, not terribly surprised, as much of our stock evaporated (thankfully, we had negotiated some price protection and didn’t walk away completely empty handed).

Epilogue: I worked at Opus for a while to try to right the ship after the bubble burst, which meant firing a lot of people and trying to restructure inherited, badly structured bizdev deals. Opus eventually sold to a larger company, but in a highly dilutive deal. I still wish they had voluntarily de-listed, hunkered down with the not-insignificant cash they had remaining, and fought on with a small but hard core team.

The Electric Sheep Company: Tripling Down on the Wrong Vision
I joined ESC in early 2006 as employee 5 or 6. ESC created virtual worlds and VW marketing experiences, and my job was to build the Second Life practice. 2006 and 2007 were busy years — SL was growing like gangbusters and we built an incredible brand, working with many of the top media, CPG, advertising and technology companies. ESC’s founder/CEO then raised $7M, built around a vision and strategy that Second Life could become a “metaverse” — a 3D layer on top of the Web.

The problem quickly became one of focus. We still had the consulting arm (which had ballooned in size). I was leading a team creating applications on top of Second Life (ecommerce, search and a new client UI). We had another team creating a virtual world ad network, and yet another team investing in AI technology to drive a new form of game play.  That is 4 divisions for a young company in a high-growth but still-unproven space.  Hiring quality slipped. Management was spread too thin. And of course, the underlying platform bet on Second Life turned out to be wrong.

We got caught up in the hectic task of keeping up with our own success, which turned out to be a sleight of hand. We also allowed ourselves to be pulled off of good business fundamentals by the reality-distortion field of a great vision.

This all sounds terrible, but the truth is only about 9 months passed between raising money, pushing into all these areas, and deciding that we needed to fundamentally change course (around November 2007).  That should have been completely recoverable.  If the company had stayed focused, small, and restrained, we would have been left with lots more of two critical assets: money in the bank, and credibility with the investors. However, the vision had been built up too big, the bets too bold, the platform risk too extreme, and so the fall was correspondingly painful.

We sold a lot of customers, but I only wish I had “lean startup” top of mind during those days.

Epilogue: After the early-2008 restructuring, the company focused on the services arm, nurturing a hope of giving another go at a product. Today, the remaining team is working on a new product in the music space, and I hope they can create something amazing.

My lessons from this one are many, but here are a few: choose and validate *one* critical business; tread carefully where platform risk is high; work with early stage investors who understand “pivot” (even if they hate the over-used word); don’t take strategic money early; don’t try to do products and services under the same roof.

Which Brings Us to Today
It can be a little painful to look back on these years, but I am stronger for it. I can take theory from folks like Eric Ries and Sean Ellis, ground it in reality and fit it to my current context. We still have a lot of work to do at Aprizi, but it feels great to approach business with more deliberation and a sounder process than ever before.

I should note that these are my own personal impressions, and my intention is not to point fingers but rather to focus on learning. If it has been useful in helping you think through your own startup, then the post has been worth it to share.

(Awesome photo at top by Romeo66, Creative Commons on Flickr)