If Lean Works, Prove It
edwardaten.posterous.com
edwardaten.posterous.com
Lean shouldn't be a way of doing business where you pick the method first, then magic happens. Nope. You look for the magic, then adapt the method. The results drive the choice of activities, not the other way around. That's the entire idea. That's why it's different.
So if your reasoning is "show me other people who use this method and do well" then you've already started on the wrong foot. A better question would be "what do startups that are successful do that we can copy?" Some of that might fit under the marketing term "lean". Some might not. Don't sweat it.
Lean is about copying what works and stripping out what isn't helpful. That's it. Yes, there are some general things that generally work better than others, but those things are just templates for you to take and adapt as you see fit. Lean is about nothing if not adapting.
I would never look at a lean practice and say "This practice will make me successful" because that's whacked. In a startup situation even if you were "perfect" you will probably fail. Instead, I would look at a successful team and say something like "Cool. There's a lean practice I can test in my situation"
We don't know all the answers when we talk about how lean applies to startups. Hell, we don't even know the questions. But it looks like there are some useful concepts we can copy and experiment with. That's what so cool about it.
From what I understand, "lean" as Eric Ries & Steve Blank describes it has a very specific meaning and it's mostly a collection of activities that is suppose to lead to fast, customer-driven product development and better ROI on marketing dollars, while reducing wasteful activities that suck up time and money.
Coming from this perspective, I find Edward's query worth exploring.
I'm no "lean" expert so I might be wrong. I've been looking forward to Eric's upcoming book to read and get a better grasp of what he's talking about. :D
Update: This presentation provides some real insight into how Steve Blank and Eric Ries define lean startup: http://www.slideshare.net/sblank/customer-development-past-p...
Practitioners will take that a step farther and claim that there is some irreducible set of activities that somehow guarantee results.
It's been my experience that such practitioners commonly overstate their case, thereby creating an opportunity for dissapointment with folks like the article author.
This is a never-ending treadmill: some great guy jumps into agile or lean, has a few good projects, and suddenly is an expert on it. So he writes a book with his "prescriptions" Some folks buy, some folks balk. There are videos, seminars, folks who "get it", folks who do not, etc. It just goes on and on.
Please don't take that as cynicism on my part, because it's not meant that way. I think it's up to the people who are serious about this to decide if they want to work at a meta level -- understanding the concepts that are being applied and feeling free to include or toss things -- or if they simply want a recipe book.
If you want a recipe book, good luck to you. There are a lot of them out there, and more will be coming in the decades ahead.
I don't mean that in a snarky way. I am eagerly looking forward to where the agile startup movement is going because I think it's going to have an immensely positive impact on all kinds of things. But if you don't understand which questions to ask, as this author did not, you will never be happy with the answers. This author had a poor model of the problem he was trying to solve. This model will continue to prevent him from being satisfied until he changes it.
Who could possibly disagree with doing that? We're all lean now!
How do people fail? There are many ways; here are two:
They use the wrong definition for "what works". "What works for me", "what works for the management team", "what works in the lab", "what works in theory", "what works for the people who don't pay", and "what works for my highly-recommended and talented designer" are not necessarily the same thing as "what works".
They don't "strip out". They talk a good game about focus, but when push comes to shove they don't really throw away anything. It's hard to throw away a feature that you've worked three whole hours on. It'll come in handy someday. Somebody will use it. If we deprecate it, it will annoy all three people in its installed base and they will Tweet nasty words. It's already built; how much can it cost to maintain it in our spare time? It's profitable -- one penny per user per year, but we'll make it up in volume. We'll look stupid if we don't have it. We're expanding, and the new hires will help us maintain it. It's a loss leader. The VCs really loved its demo. Techcrunch is wild about it. Your mom smiles when she sees it. It helps kittens and puppies; think of the puppies.
"It works! If it doesn't work for you, it's because you're doing it wrong!"
Now... where have I heard this before?
Oh right, Agile.
This morning I found myself wondering if the job of the talking head, is to inspire people to take action and feel empowered? Or to execute and prove the methods they preach?
How different is Lean from Tim Ferriss, or "The Game"? Or any other self-help religions? And when does it become dangerous to believe what they say?
For those not there, somebody was pitching a book on "The Lean Way" pretty much every 45 minutes or so. It had a very Tim-Ferriss like feel.
I think lean startup advocates would argue that lots of household names (google, facebook, youtube, etc) utilized lean methods, even if they weren't consciously doing so. Not sure if I'd agree with them, but they probably would say that.
Consciously promoting lean startup ideals != doing things the lean way. The lean startup movement is more about a label and a tribe than it is a set of behaviors (not that this is necessarily a bad thing).
Facebook, Groupon, Google, Dropbox, Mint, Heroku...dare I go on?
The methodology goes beyond just web tech companies and I'm afraid us in the tech industry fail to see that. Xerox once was a "lean startup".[0]
[0] - http://tbmdb.blogspot.com/2009/12/business-model-example-xer...
Zynga continues to implement Lean with it's products. They don't develop any new feature in any product without concrete proof of user interest. Put up the "store" sign, make it clickable, if enough users click it, develop the feature.
Simple, straight forward, validated.
Google: They built the product first and scaled massively before they had a business plan. Not a lean process.
Facebook: Product first and scaled massively before they had a business plan. Not a lean process.
Dropbox, Mint, Heroku, I can't comment but I suspect you're massively misunderstanding what a lean startup is.
A lean startup is not starting in a garage on a shoestring, it is poorly named. The point of a lean startup is to identify a profitable product using a very low burn rate. Google and Facebook's burn rate was massive. It is a methodology, finding a winning product by using an iterative feedback loop of customer, prototype, adoption. Only once you have a profitable product that provably sells do you scale. That is definitely not Google or Facebook.
This diagram illustrates the concept:
http://steveblank.files.wordpress.com/2010/04/lean-startup-2...
Again, that does not at all describe the process that Google or Facebook went through. I'm not spending time researching the other companies you mentioned but for example I'm pretty certain Dropbox was written without the first three stages, so again can't be classifed as a lean startup.
>It is a methodology, finding a winning product by using an iterative feedback loop of customer, prototype, adoption.
Ok fair enough.
Google: What were gmail, gmaps, gdocs doing in BETA for several years?
Dropbox: You do realize the online course on the lean start-up uses dropbox as an example right? http://www.socrated.com/user_courses/226?content_item_id=168... (you have to click start learning first and then use this link to go directly there)
> I'm not spending time researching the other companies you mentioned but for example I'm pretty certain Dropbox was written without the first three stages, so again can't be classifed as a lean startup.
Which is why I'm shocked your comment is upvoted more than mine...
The big red flag for me was you calling Google and Facebook lean start ups when they weren't. I didn't have to do any research for that. Google's entire 20% thing flies in the face of lean methodology.
None of the other google prdoucts you mention are lean anyway, gmail was built for internal use and then made public. Gmaps and Gdocs were bought, so no lean methodology by Google there.
I'm also not sure what you mean about them staying in beta for so long? Google's beta's are practically meaningless. Was there anything special about them taking the beta off gmail?
> I think you're misunderstanding what a lean startup is.
It also irks me that one of the most well-known cases of MVP/Pivoting discussed on HN has been by Dropbox, and yet you were completely unaware of it. I'm sorry if I'm coming off crass but the lack of research on your part doesn't help your argument. If I've goofed somewhere I'd like empirical evidence. It's important I understand this fully.
> The point of a lean startup is to identify a profitable product using a very low burn rate. Google and Facebook's burn rate was massive.
Really? When was the burn rate massive? From Steve Blank:[0] The key contributors to an out-of-control burn rate is
1) hiring a sales force too early,
2) turning on the demand creation activities too early, [1]
3) developing something other than the minimum feature set for first customer ship.
Google:[2]
1) Google hired their first head of sales nearly 7 months after it incorporates. I wouldn't really call that hiring a sales force (nor is one guy a force) too early.
2) AFAIK Google wasn't out there actively promoting with their 8 member team. They continued to develop the product.
3) I'm pretty sure there is nothing more minimal than this: http://web.archive.org/web/19981111183552/google.stanford.ed...
Gmap - Yes, they bought a very early staged mapping software and transformed it into their own. But it was very much the baby project of Lars, whom they acquired. They continued to reiterate through the development of it, with sometimes weekly, if not monthly updates.[3] Yet, the source of revenue for gmaps has continued to evolve (adwords, enterprise, premiere-api, and now featured listings)[4] gmail was built for internal use and then made public. And how is that not lean? They had zero burn rate and they developed a product with the minimum features for the first customers. CLASSIC case of lean principles, if not THE BEST.
Facebook:[5]
1) Did Facebook even need sales people? It spread word of mouth.
2) Read #1
3) "Membership was initially restricted to students of Harvard College, and within the first month, more than half the undergraduate population at Harvard was registered on the service."[5] - It literally was minimal for several years. Do you know how many features they've added over the past 7 years? Too many to count, I could write a book on it.
Other notes: Facebook wasn't available to the mass public. You needed to have an EDU e-mail. They did a MASSIVE pivot making it public to anyone.
[0]- http://steveblank.com/2009/11/02/lean-startups-aren%E2%80%99...
[1]- http://www.quora.com/What-is-the-definition-of-demand-creati...
[2]- http://www.google.com/corporate/history.html
[3]- http://en.wikipedia.org/wiki/Google_Maps#Development_history
[4]- http://blumenthals.com/blog/2009/05/13/google-maps-developme...
I'm intrigued by the Xerox example, because it seems like a disruption, but in fact it was much higher performing than the incumbent technology. It reminds me of how the iPhone beat the Nokia with a better product (instead of a worse one, which is usually how disruptions start).
In short , the path from the first mp3 , via the ipod to the iphone is a disruptive path. there's no motivation for nokia to compete , until it's too late.
Unless you're measuring battery life ;)
are these the examples of startups that when being early stage proved their concepts with minimum viable products, pivoted, failed, 'reiterated', etc?
maybe they were lean, but didn't they hit the right target from first attempt? (it's a question, not a statement)
Maybe it would be better discussion to talk about companies who aren't lean?
Simply put lean startups result from the combination of lean manufacturing principals and maneuver warfare (agility) principals combined and used in business. Which if you want to boil it down further could be thought of as using the scientific method as quickly and efficiently as possible to create organizations and products where teams have fully adopted a culture of continuos improvement.
Lean is probably best known for its ability to enable Toyota to be the amazing manufacturing organization they have been (ignoring recent growth related fuckups). See the book Lean Thinking or the many Lean Software books to see the ways in which people have translated Lean concepts to software development. Of course if you want to really grok lean in the context of making products you should study http://www.amazon.com/Principles-Product-Development-Flow-Ge...
Agile principals as codified by the god father of agile strategic thinking, John Boyd would be the other crucial element to lean startups. He is famous for driving maneuver warfare doctrine into the US military establishment and codifying some of the most advanced strategies and concepts regarding the nature of competition, agility and strategy. If you have seen the OODA loop or heard the phrase operating inside the enemies decision cycle then you have seen his work in action. See the book Certain to Win by Chet Richards for a good primer on Boyds work.
These concepts rigorously and intentionally applied to developing products, teams and companies are IMO what makes a startup lean.
And of course if you are interested, you can also read Eric Ries or Steve Blanks blogs.
It's like evaluating a basketball camp for kids by seeing how many make it to the NBA.
I thought the point of lean development was to fail fast and try again.