Yesterday I had the pleasure to visit TechHub to listen to Eric Ries talk about the lean startup. Like some others attendees I could not attend Business of Software earlier in the month so when I found out Eric was coming to London I booked a ticket immediately.
If you’ve read Eric’s blog you’ll have a reasonable idea of what the talk is about. What surprised me is that I’ve only read a few bits of the blog and only managed to tip my toes into one of his books but I seemed to have stumbled onto the key points all the same. Listening to someone talk is always interesting, the asides, the questions from the audience and the clarifications and answers. Not the same as reading a blog or the book that comes from the blog. Eric’s style is confident, engaging, witty, intelligent. Frankly everything you could want from someone that knows their topic really well. Really on the money.
What is a startup?
A startup is a human institution designed to create something new under conditions of extreme uncertainty. Nothing to do with industry, economic sector, or how well funded you are. A startup is an experiment to determine if a product should be built.
Most startups fail
One of the purposes of lean is to give you more chances to succeed. This means you stop wasting people’s time developing the wrong product.
What is success?
Traditional answer: Success is because it made money. Wrong answer.
Eric thinks that success should be defined as “Success is because the product improved other people’s lives”. I think he has a point. There are plenty of rubbish products that make money but make your life a misery.
Manufacturing history. How did we get here?
Eric’s talk covers the pioneer of modern industrial management Frederick Winslow Taylor and also touches on the work of W.Edward Deming and a Japanese gentleman (whose name I can neither spell nor pronounce, sorry) in Japan which ultimately lead to the lean manufacturing movement. Eric talks about how some of the initial ideas that worked for manufacturing no longer work. I hadn’t heard of Frederick Winslow Taylor but I had heard of W. Edward Deming as he is referenced in Quality is Free (a rather dry read I remember).
The waterfall development method is history (its OK, no one expects you to use it). Waterfall provides the wrong type of feedback. You think you are acheiving because you complete each step. But there is no customer feedback in the loop. So you can get right to the end of product development before you discover that the customer will not even use your product (not even in user testing when you are paying them!). Thats a real life lesson from some of Eric’s work. Thats a painful lesson.
Entrepreneurship is management
If you are leading your business then you are managing the direction of your business. You do not have to think of management is uncool stuff that folks in grey suits do. You are managing, but you are doing a different type of managing in a startup than you do in an established business.
Comments and jokes about pivots. You can read all about pivots all over the web.
Whats the problem? Whats the solution?
The Donald Rumsfeld moment. “We have known unknowns and unknown unknowns”. Yowza. How did he get that in there?
Traditional development is about trying to create a solution to a known problem. The problem that most startups have is they are trying to create a solution to an unknown problem. The reason the problem is unknown is because you are not exactly sure what the problem is what the customer wants solved. The lean method with its continual build, test, measure iteration allows you to burn less as you try to refine the unknown problem into a known problem and at the same time build the correct solution for this soon to be known problem. You also need to correctly identify the correct customer for your problem. Think about that last sentence for a while. A lot of value in there.
You must question your assumptions at every stage of development. Everything from who the customer is, what the problem is, how the problem should be solved, what each step should be along each stage of development. All of these must be constantly re-evaluated. Eric gave a great example from IMVU where they had instant messages solutions and assumptions about that market and how all the assumptions were wrong. Painful. Frustrating and so on. But ultimately very enlightening (learn from the mistakes, insight follows).
Lean startup optimizes for the whole process of product and customer development. You want to make the time to go around the loop of ideas, build, test, measure is as fast as possible so that you can iterate rapidly. This is to ensure you develop the right product that customers want with as few wasted (the test results were not good) iterations as possible. A side effect of this is that underlying sub stages are not optimized (you can’t optimize for the whole and optimize all the sub parts at the same time).
After the 40 minutes of talking, Eric then answered questions for another 30 minutes. Unfortunately I can’t quite make out what the questions were (not all people had a microphone when talking) so I can’t summarize those here. After the questions lots of people made a beeline for him to have a chat.
On the train home I had a chance to reflect about the talk and the chat with Eric.
Although it wasn’t stated outright in the talk I thought another way of looking a lean startup is that it prevents premature optimization (building the best wrong product you can) because it forces you to build, test, measure each feature of the product with a view to what the customer thinks is useful (their benefits). When at IMVU Eric wasted 6 months creating a feature rich set of messaging protocols that no one wanted. Why do I call that premature optimization? Well if you don’t know what you should be building you can’t just steam ahead and build it. You have to experiment. If you think know what you should be building you should still test it before building. Anything else and you’ve come to a premature conclusion on what you should be doing and you’ll focus on optimizing that (most likely wrong) conclusion.
Hare and tortoise
Lean startup is also a bit hare and tortoise. You get to the end (the correct product for the correct customer) faster by working slower. Slower? Yes. If you are going to take careful, thoughtful, deliberate steps, with tested measurable results, then the steps may well be slower than sprinting to the finish. But you’ll have some idea that you are on the right track courtesy of the testing and measurement. And if you get to the finish first but have arrived at the wrong destination (customers don’t like your offering) have you really finished first Mr Hare? I think that Tortoise is looking mighty fast.
One thing that I used to do many moons ago was keep a development diary. I stopped some time ago. I can’t remember why. I do write copious amounts but its unstructured stuff just to help the current task get done. I couldn’t help thinking that part of what Eric was talking about would be helped by a more formal development diary separate from the unstructured notes. Reviewing that diary every month or so would be a good way to get some sense of progress and journey out of what often seems day to day “right whats next?” all the time.
I had a brief chat with Eric but didn’t feel the occasion was right for what I wanted to discuss in detail. Perhaps I’m wrong on that. I just felt it would take more time than would be fair for the other people waiting to talk to Eric. So I kept it brief and left. We’ll make contact another way. I live in the fens north of Cambridge and rarely go to London (I like people, but just too many people in London), so it was very interesting how trains have changed since I last used them. Nearly everyone had their head in a mobile phone, a Kindle or a laptop. Hardly anyone read newspapers or books. I had a book (which would have been a Kindle if it had arrived last week).
The book I was reading is 59 Seconds (think a little, change a lot). It was really interesting how some of the topics in the book resonated with Eric’s talk. The book is not about software. The book is described as “At last a self help guide that is based on proper research. Perfect for busy, curious, smart people”. It really is interesting. Just about every topic discussed also references studies by various people. Often there are exercises to do to see how you score. The motivation topic was really interesting. I now know why I get things done. I also know why when things don’t turn out as planned, the next day I am right back on it. I suspect these traits were shared to a greater or lesser degree by most of the people in the room last night. I think most people could read this book and take something positive away for their personal or business life.
I can’t remember how I came to purchase the book, but I know it came up in a list of books on Amazon, even though I’ve never purchased a self help book on Amazon. I guess other technies are already reading it. I think part of what sold it to me was the endorsement by that master of deception, misdirection and illusion, Derren Brown.
Recommended (that is not an affiliate link above, I get nothing for recommending this book).
If you get a chance to see or listen to Eric talk, take it. I recorded the talk. 40 minutes of talk, 30 minutes of Q&A.
I hope I’ve given you a sense of what the talk was about and that I haven’t distorted any of the ideas too much (or even at all!). As always if you want the exact ideas, visit the startuplessonslearned blog or buy Eric’s books (there are 3 books in total, the link is to the most recent book).