The Lean Startup Summary
Here's the great startup myth of our time. If you only have determination, brilliance, great timing, and above all, a great product, you too can achieve fame and fortune. A related misconception is that ideas are precious. Generally, people hesitate to reveal their ideas in public - even among friends! There's this nagging fear that someone can steal the idea from you. Please ... Sorry, but an idea is never that great. I would bet my right arm that during a lifetime, the average person has at least 20 awesome ideas that could be turned into commercialized and successful startups. And that's even when removing Elon Musk from the sample, which otherwise would have skewed the average upward by a disproportionate amount. The issue is not ideas, determination, brilliance or great timing even - its execution. We will now look at 5 advice regarding the creation of a successful business from The Lean Startup", that the multi successful entrepreneur Eric Ries has written.
Takeaway number 1: The build-measure-learn feedback loop Planning and forecasting are only useful when your operations of business have been rather stable and where the environment is static. Startups have neither of these. Planning for several months and releasing a product only after many thousands of hours of perfecting it, is a game of very high stakes. What if you build something that nobody wants anyways? Many entrepreneurs have realized this, and as a consequence, they take the approach of the opposite side of the spectrum: If you can't plan, then they think that you can only operate in chaos. So they adopt the just do it strategy instead. Nike might be correct when they tell you to put on your running shoes and just do it, but in building a successful startup, this plan is usually flawed. So what can we do instead? The solution is the build-measure-learn feedback loop. Too much planning makes the process rigid and very risky. Just doing it tends to turn everything into chaos. Quickly going through cycles in the build-measure-learn feedback loop is the solution. The goal of a startup is to figure out the right thing to build - that customers want and will pay for - as quickly as possible. We must be able to quickly kill off that which doesn't make sense, and double down on that which does. Although the order of the actions in the build-measure-learn framework is, well ... build, measure, learn, the planning happens backwards. We must first figure out what we want to learn. We can do this by forming hypotheses. Here are a few examples: Commuters want to be able to order food from their cars; People are willing to assemble their own furniture at home. People are willing to pay monthly for being able to stream unlimited music online. In hindsight, these hypotheses seem trivial, but before these companies proved them right - they you weren't. After formulating these hypotheses that your startup revolves around, it's time to validate or reject them, which is our next takeaway.
Takeaway number 2: Everything is a grand experiment Okay, okay. So learning as quickly as possible about your customers needs and willingness to pay for your product is important, but how can you do this empirically? The answer is experiments. Experiments are run together with real or potential clients. The common denominator among the successful ways of experimenting is this: Observe, don't ask. Customers usually don't even know what they want! But if you show them a product and they are interested, well done! Then you just gained some validated learning. First we should always ask ourselves: Do we even have to build anything at all? An experiment can be as simple as setting up a landing page, stating that you have a product that can do something - say - improving your chances of getting a match on Tinder by 500%, driving some traffic to the site through Google Adwords, and then seeing how many people that would sign up for it. At other times we must be more rigorous, and develop a so-called MVP, which is short for Minimum Viable Product. In Lean Production, a concept that originally stems from Japanese car manufacturers, waste is defined as something that doesn't create value for the customers of the company. In Lean Startup, waste is defined in another way: Everything that doesn't lead to validated learning is waste. This is the reason why it's called a Minimum Viable Produc. It should only contain the features of utter importance, and it shouldn't be polished any more than what's absolutely necessary for us to prove or reject our hypothesis. If you are embarrassed releasing the product to your potential customers, you are on the right track.
Takeaway number 3: Different types of MVPs There are many different types of MVPs that can be constructed in order to learn what solves the problems of customers, and what they're also willing to pay for. Eric Ries gives three different examples: The video MVP A famous example here is Dropbox. Drew Houston, the CEO of the company, produced a video showing an extremely easy to use file sharing tool, and published it online in communities with tech savvy members. The sign up list for the product went from 5,000 to 75,000, overnight! The best thing about this story is that the video was fake. It didn't even exist such a product yet! The concierge MVP Another way of testing your initial hypothesis empirically to gain validated learning about customers is to give them the concierge treatment. This means focusing on a single or only a few customers, and develop the product according to their preferences. Slowly - but surely - after being certain that you solve the problem of a single early adopter (often manually) you can expand to more customers and automate more and more of the process. The Wizard of Oz MVP Aardvark, a search engine for subjective inquiries, such as: Which YouTube channel is best - PewDiePie or T-Series?, famously use this approach with great success. Basically, it's all about pretending that you have developed a fancy technical solution, while, behind the curtains, it's operated by humans. In the case of Aardvark, they had employees who were coming up with the answers to inquiries from customers in the beginning, before they knew that the product was something that people actually wanted.
Takeaway number 4: The three engines of growth Now, once we have confirmed that the product creates value for someone, it's time to shift focus to growth. Eric Ries argues that there are three different types of engines of growth that a company should concentrate on. It should pick only one primarily, and try to improve the metrics associated with that engine through iterations in the build-measure-learn feedback loop. The sticky engine These types of businesses are designed to attract customers for the long term and never let go. A company of this type should focus on two variables primarily - the new customer acquisition rate, and the so called churn rate, which is the fraction of customers that failed to stay engaged with the company's product. The sticky engine of growth is used by, for instance, repeat purchase companies like Gillette, and also by pretty much all online gaming companies. The viral engine These are businesses that focus on spreading like epidemics. For every new customer joining, the idea is that that person should invite one or more friends as well. Success in growing a company of this sort depends on something called the viral coefficient. This is calculated as the number of customers recruited on average by every recruited customer. If every customer on average, say, attracts half a new customer, it's not an efficient engine. On the other hand, If every customer can attract, on average, 1.1 new customers - the product will spread like a virus. The viral engine of growth is used by all social networks and by multi-level marketing businesses, among others. The paid engine A company using the paid engine of growth realizes that primarily, it will have to focus on advertising in some form to reach customers. The two most useful metrics to a company of this kind are CPA (short for cost per acquisition), and LTV (short for lifetime value). It's simply the difference between how profitable a customer is over its entire lifetime, minus the cost of acquiring a customer, that will determine the growth rate of such a company. An example is e-commerce businesses. Eric Ries mentor used to tell him the following: Startups don't starve - they drown He's referring to that there's so much that a start-up CAN do, but it's much more difficult to determine what a start-up SHOULD do. The three engines of growth are a way to focus the energy of the startup in the right place. Takeaway number 5: Pivot or persevere? Successful entrepreneur does not give up after facing a little bit of headwind, but at the same time, he doesn't stubbornly hold on to his idea until he crashes either. You must possess both perseverance and flexibility. Sometimes a pivot is necessary, or in other words, a change in this strategy on how to achieve the overall vision that you have for the startup. But how do we know when to pivot? It's a three-step process.
- Establish a baseline of the current situation using an MVP.
- Attempt to tune the MVP towards the ideal, through multiple iterations in the build-measure-learn feedback loop.
- Pivot or persevere? The issue is that there will always be a gray zone. If our attempts to tune the MVP into something that is improving our engine of growth is not helping, it's a sure sign that we need to pivot. But is a 10% improvement per year enough? Unfortunately, there's no clear-cut threshold here, intuition is needed, and that we can only gain from experience. Here are some of the most common pivots to make: Customer segment pivot The initial group of customers that you intended to focus on, might not be as interested as you thought. But maybe another group showed a lot of enthusiasm? Let's focus on those people instead. Value capture pivot Deciding to give away the product for free, and then charge through advertising instead, is a typical pivot that many companies using the viral engine of growth tends to do. Engine of growth pivot Perhaps your hypothesis that the product would spread virally like an epidemic was rejected, but in the meantime, you discovered that the lifetime value of every customer is much higher than the cost of acquiring a customer. Then you may pivot from the viral engine of growth, to the paid engine of growth instead. One thing that is important to notice here is that a pivot is not a failure. Discovering that something doesn't work, might not be as useful as discovering that something does, but it's way better than trying to dig yourself out of a hole.
Time to wrap it up! The goal of a startup is to figure out the right thing to build, that customers want and will pay for, as quickly as possible. To confirm or reject your hypothesis, experiments must be run to gain validated learning about potential customers. An MVP usually forms the basis of such experiments. Three popular types of MVPs are: video, and The Wizard of Oz. Every startup should focus on an engine of growth, either the sticky, the viral or the paid one. Successful entrepreneurship is about perseverance and flexibility at the same time, pivoting an idea does not equal failure. In fact, almost every successful startup has done a major change to its strategy at some point Have a good one!