Sunday, March 20, 2011

Relearning to Startup, Lean

 Lean is a funny word. It has the obvious health and behavioral terminology that people often interpret as "skinny" or "fat"; Connotations of fat are rarely ever good, unless you are rapping about your wallet. Further, people extrapolate Lean to mean "cheap" as well -  thinking Lean means you never spend any money. For entrepreneurs coming out of two different economic downturns in the last decade, this also sounds appealing. In the end, however comfortable "skinny and cheap" may sound, the Lean Startup movement is really not about those things. They are symptoms and output of a Lean way of thinking, but simply re-using your Q-Tips and saving money in a heroic fashion is not lean.

I will not try to retell this history of Lean and Lean Manufacturing here, however, it's a movement that started not with Eric Ries or the internet, but inthe automotive industry, with Toyota. There are more than a few wonderful references on the web were you can understand this history - I'll start you with the wikipedia entry and one specific sentence that resonates for me:

"Lean manufacturing is a variation on the theme of efficiency based on optimizing flow; it is a present-day instance of the recurring theme in human history toward increasing efficiency, decreasing waste, and using empirical methods to decide what matters, rather than uncritically accepting pre-existing ideas."

This post allows me to tell you about some of my experiences, and some deadly pre-existing ideas that were sins of my past I'm "unlearning", to move forward in a more Lean manner. What I hope to expose you to in this post is why, as an experienced entrepreneur, I'm trying to rethink and relearn, why I'm participating in this contest, and why I believe in this movement.

A friend has told me recently Steve Blank has said "There are no 10 year startups - just 2 year startups and 8 year failures". I have no idea if Steve actually said that, as he's not said it any of the talks I've attended, but it's correct on many levels. One thing you have to understand, that will color this blog entry considerably, is that it took us ten years to get my last startup to a successful exit, and we had to pivot our way through 9/11 and the bubble bursting  in Silicon Valley. The timeline simply helps illustrate for you what the world around us was like as we started, and kept alive, my first startup.

The success of a company and entrepreneur is usually judged by the endgame - in this case, we successfully sold the company in a manner that ensured the employees and investors were all very well taken care of, something in which I take great pride. However, I think understanding the little successes and mistakes are were the real lessons are learned.

Commitment to Vendors....or Customers
The spiraling reduction in cost to startup a business today, especially in infrastructure is widely discussed, and fairly well understood. It is a reality I share as well. When we started our last startup, we hosted our infrastructure with EDS, full of  Solaris servers, Oracle database licenses, and million dollar design firm(s). We had a massive amount of money committed and infrastructure in place just to be able to test out basic ideas with customers. In other words, we had to be committed to our vendors before we had any idea how to (or what to) commit to our customers. Today, though, you can do all this and more - with real metrics and feedback analysis - with a Amazon or Joyent cluster and a few dollars per month.

Focusing on the intension of Lean - it's not about the money, but about the efficiency and focus of resources. The capital necessary to run these resources was obviously a huge drain on investment, but what about the human capital?  I am sure you are counting in your head how many people we needed to manage the servers, the relationship with EDS - even just managing the Oracle licenses themselves. Imagine if we could spend that time working with  our customers to understand what they need, and commit to our customers before committing to our vendors.

Voice of the Customer and Analytics
There are two products that really embody the Lean methodology for me, and they are User Voice and Geckboard. I'm really excited to try out both of them in a real world setting.

UserVoice is interesting to me because it most closely matches my belief of building a community of vocal, passionate, champion end users. Further, I believe these champions lead to the lowest cost, highest value acquisition strategy you can have. These champions must not only be acquired, but retained. I believe this is where tools like UserVoice fits in.

Historically, companies are more interested in acquiring your business than keeping it. Of course, they will never admit that, but Sports Illustrated never sent you a free jacket because you were a customer for five years. There are very few industries that celebrate loyalty as aggressively as they do acquisition. It's true that movements like Gary Vaynerchuck's Thank You Economy are challenging this notion, but it still has decades before it's the standard, rather than the exception. I think, implemented correctly, technologies like UserVoice can allow you to build community, feedback, forums, and a connectedness that is necessary for building and retaining passionate user base. I'm also very interested in the social media aggregation aspects of a similar product, Assistly.

As I said, I'm excited to start using  Geckoboard, a display analytics system. It can pull in data from systems like Google Analytics, CRM, and other packages, and you can configure and display dashboards on mobile, tablet, desktop, and large-format (think, 52" LCD in break room).

Here's the first sign you need Geckoboard, or something like it - you're spending more than 30 minutes a week preparing charts and graphs for executive review meetings, and then figuring out how to updated them for board meetings, all-hands meetings, and driving your business and culture, then you need a tool like Geckoboard. Spend you time analyzing the data, not formatting it. 

Steve Brown, the CEO of my first startup, used to say to me "that which is measured gets done", and he could not be more correct. This was true both for the products we were building (enabling behavioral change in chronically ill patients) and for running the business. Despite our best intentions, the work that went into building a good analytics and display platform was always priority 2, 3, or 4 while priority 0 (customer fires) and priority 1 (unvalidated product features) always took priority. We did what we could in Excel, but today, we'd make totally different choices. One of the last things I did before leaving was try to push through a few big-screen LCD purchases as the way to force the analytics process. Even though my previous company is certainly no longer a startup, analytics and information delivery is the practice of a good company, and not just startups.

QA is not a Person or and Organization. It is a culture. 
The success of companies like Zappos, and their spiritual followers Bonobos and EcoMom have championed a philosophy that everyone in the company is responsible for customer support, that it is a unique selling proposition. However, far to few organizations have applied the same concepts to quality and testing. I include myself in that group - I realized far too late in my last startup that I needed to change that, culturally.

You can see this is conceptually not so distant from our Customer Support and Metrics conversation, although culturally, the idea that the entire organization using the product and feeling (and being accountable for) quality is incredibly important. Do not bury under Customer Support and Metrics Driven aspects of your culture. I've written about this before (lovingly called eating your own dogfood), so I won't repeat that here, but I'll highlight some tools you can use to ensure that you can get this into your culture starting on "day one".

The first such tool is utest, an outsourced testing firm. As an on-demand, pay-as-you go model, they allow you to maximize your resources without making long term employment or contractual obligations. In my world, there is a huge difference between "validator" and "tester." You still need "validator"s that know your market are the voice of the customer, but the idea that you can get specific testers from utest that will execute the hand-ringing aspects of test, including, software installs, browser variations load testing,  and a myriad of other complexities absolutely invaluable. If you're sitting around saying "how have we missed that bug, it was in the last three releases!", then you need the help a service like utest can bring. 

Now, you are probably saying  "Wait, you just said eat your own dogfood" and "make quality cross-company initiative", and then, it sounds like you said outsource testing to utest. Huh? Again, the Lean aspect is not that you spend less money. Outsourcing QA to India or a low-cost provider is not the answer. It's that you did more with your available resources, be it dollars or time. In this case, we're using utest to get us more time to do the most important pieces - validation of our user experience.

Testing the product is not using the product. You need someone, like utest, to execute your test cases, but each of you in the company need to be fanatical about the user experience and the quality of the products. Test cases will pass at utest, but the experience of the user is all wrong. Test cases will pass utest (and match the specification perfectly), but you'll find your specification is wrong. 

Passing test cases is meaningless unless you are sure you built the right thing, and you can only be sure if you are passionate about feedback, iteration, and validation.  Let your vaidators rule - use mockups, use user-feedback tools, and do everything you can to remove checkbox testing out of your organization and replace it with emphatic validation.

This does not mean that your organization does not have quality professionals. Free up those intelligent people you have in test, who joined you company because they are passionate about the solution, to be part of the product, not just checking boxes in your bug tracking system. Just like Bonobos employs customer service professionals (known as Ninjas, even) even though everyone is accountable for excellent customer service , you need to employ "validators".

Mockup, down, and sideways
Hint: Consider mockup tools like Balsamiq. There are a number of great wire-framing tools in the market, but there are three things I love about Balasmiq: First, they are insanely well priced for a high quality product, so you feel like you get more than you paid for, which fits the Lean Startup theme fairly well.

Secondly, they have a good community that continues to create excellent templates and examples. I cannot think of anytime I've tried to create something I couldn't find a good example from which I could borrow, so I could keep focused on my product, and not my tools.

Finally, they have really great plugins for most major wiki and bug tracking tools (like Atlassian's Jira Studio) and lots of add on tools to create Flash Demos, HTML walkthroughs, and even an Android application (beta) . It's a great method to get feedback before you write any lines of code.

My Struggle: MVP
Finally, I should  admit some areas of Lean I struggle with - most notably, the Minimal Viable Product, or MVP.  The idea of MVP is to release a version of the product which allows you to collect the most data about your customers with the least effort. Often times, it's used as a discussion catalyst for identifying feature bloat. It's also often attributed to the Reid Hoffman quote: "If you are not embarrassed by the first version of your product, you’ve launched too late."

 I know the old way - year long development cycles, minimal feedback, exhausting timelines - none of those are the "right" way, but I am petrified of a market with minute-length attention spans, and not gaining enough "M" in my MVP to keep people engaged. This probably means I don't understand my target well enough.

The only successful way I've found, so far, of trying to get MVP to work for me is by finding a constraint, such as Luke Wroblewski's Mobile First concept.  I've already written far too much, but I can point you to Luke's presentations you can review yourself. The Mobile First movement is not specific to Lean or to MVP, but one of Luke's key points are using the limits of the mobile device screen size, bandwidth, and other constraints to focus your thinking. So many people start on the desktop, with it's endless bandwidth and large screens, and then try to "cram down" to mobile. It's actually a very empowering methodology to go the reverse direction, and works well with the MVP concept.

You probably have many good ideas for how to really get an MVP out the door and walk the fine line of engaging customers. I've not yet found that method or tool to help me do this really well, and everyday, it's on my mind. I'm looking for best practices, mentorship - hell, maybe a support group. Got ideas? I'd love to hear them in the comments below.

Being Lean
There are people  that refer to The Lean Startup movement as "common sense". I think there is no less damning description of a way of thinking then to describe it as such. However, I think that minimizes the importance of the implementation of the ideas, embodied in many of the products and services in the AppSumo Lean Startup bundle.  The  Lean Startup movement is not only conceptual ideas, but the products and processes that support them.

I don't expect you to read anything here and say "Dear Lord, Clapp as figured out a rounder wheel! Bravo!" However I do hope that putting it in the perspective of the "old way" and the "new way" will help to point out some of the subtleness, and more importantly, some of the areas I still struggle with today. A bit like someone who grew up when smoking was endorsed by doctors, I know some of my behaviors are bad for me, but I'm still having a hard time quitting them.

Let's revisit the end of the quote I introduced at the beginning of this post - not the whole thing but quite literally, the last six words:

 rather than uncritically accepting pre-existing ideas.

Hopefully I've demonstrated to you that simply following what we know from our past, our pre-existing ideas, can be a troublesome, if not deadly path for a startup. We need to mix valuable experience with decreasing waste and constantly measuring what matters, the sum of which will help us adjust priorities. Embracing Lean today is about employing the tools and best practices around you, and changing your way of thinking.

This is a way of thinking that does not stop with a book, or an AppSumo bundle, or a piece of software - the things we consider Lean today might be wasteful tomorrow. We owe it to each other, as a community, to keep pushing the definitions, the products, and the process that make up what it means to be Lean, and share best practices and failures often, openly, and with intellectual vigor.

If you would like to vote for this entry in the Lean Startup Challenge.....

Please post a tweet including the text: #leanvote67

Thank you for your consideration!