The Lean Startup: How Today's Entrepreneurs Use Continuous Innovation to Create Radically Successful Businessesby Eric Ries
Eric Ries with this book started a big movement around the world among startup development processes. This book presents an innovative approach to startup creating. The main object is to stop building products that no one wants. Instead, discover what people really want on every step of the product development.
VISION // Start
“The goal of a startup is to figure out the right thing to build”
The Lean Startup has its origin in lean manufacturing practices invented by two people in Toyota, Taiichi Ohno, and Shigeo Shingo. It shows the difference between value-creating and waste.
The Lean Startup is about measuring productivity in a different way. It’s not about time you put into the project, it’s not about executing times itself. It’s about finding out what is the right thing to build, that will really be used by the customers.
The Lean Startup method is about creating small improvements and learning from them on the way, which is completely opposite of building a big product, to realize later, that no one needs it and all that time and money are wasted. This method is called the Build-Measure-Learn feedback loop, that will be described in more detail in later chapters.
In the Lean Startup method, the role of the leader changes from being a Caesar that decides what stays alive, to building a company culture that encourages the innovation among the team. As the leader, it’s your responsibility to build the experimentation system.
The experimentation system is a culture in the company that is focused on fast building, deployment, and analysis. System, that makes testing the hypothesis easy and fast, for pretty much everyone in the team. It’s called a system because the entire company must be involved in it. When marketers want to test something in the product, they need a certain process in the company to make it happen.
“Lean thinking defines value as providing benefit to the customer; anything else is waste.”
The base of the Lean Startup is to learn as fast as possible. So, instead of building a complicated product with all the features, it’s better to build something simple, a prototype, then show it to people, and learn from it. Learn from the way they use the product, what they are saying about it, etc.
In order to make it happen, it’s important to talk to customers.
“If you cannot fail, you cannot learn.”
In order to achieve good learning, experimenting is crucial. One type of learning is to talk to potential customers and listen to what they are saying. This is good to start, however, once you design your product or service, you want to validate if this is what users would use. You want to show them the product and observe how they are using it. The whole process of experimentation is about creating value, validating that the feature has a real value, and remove everything that does not contribute to that value.
Building a product is about doing experimentations. And experimentations should be made in a real scientific way. First, define the hypothesis and then test the prediction empirically.
Here are a few very good questions to ask:
Do consumers recognize that they have the problem you are trying to solve?
If you offered them a solution, would they buy it?
Would they buy it from you?
Can you build a solution to that problem?
The core of the Lean Startup model is the Build-Measure-Learn feedback loop. None of the elements of this feedback loop are enough by themselves. Entrepreneurs must use all 3 elements to create great products.
In the next chapter, you will learn techniques that will help you minimize total time through the Build-Measure-Learn feedback loop.
“For startups, the role of strategy is to help figure out the right questions to ask.”
In the beginning, there is a business plan with many assumptions. We assume that our product will be likable by customers, or we assume that the popularity of the product will grow by itself. These assumptions are called leaps of faith because the future of the entire venture rests on them. People make assumptions based on many things, for example on analogies. If company X is using marketing strategy Y, we also can grow exactly like this. It’s just an assumption, and it can be dangerous for the future of the venture, as this assumption might not be true.
In the Toyota Production system, they use a term genchi genbutsu, which means “go and see for yourself”. This method says that if you really want to understand any part of any business, you should get the information by yourself, directly from your potential customers. You don’t get the information from the reports, instead, you go out of the building, and talk to people.
Talking to people firsthand will not get you the direct answers, but those are the basics to understanding your potential customers and what problems they have.
Avoid being a just-do-it entrepreneur, who cannot wait to build his product and get it out to the market. Instead, validate your assumptions first.
Actions to take
“A minimum viable product (MVP) helps entrepreneurs start the process of learning as quickly as possible.”
Many companies that you know, such as Groupon or Dropbox, did not decide to build the product when they started. Instead, they tried to validate their assumptions in the easiest possible way. In the case of Groupon, they opened a simple site on the Wordpress and started to publish one offer daily. They sent coupons to people manually. They were experimenting with what was working and what was not.
Guys from Dropbox did it in an even easier way. They published a video showing all the features they were going to build. The video went viral, and this proved the hypothesis that there was a need for such a product in the market.
As you can see, it’s possible to validate many assumptions at more than 10x lower the cost of building a full product. Why not build the full product in the first place? Because it will save you time and money. You will be able to iterate fast and learn from your mistakes along the way.
Actions to take
“A startup’s job is to (1) rigorously measure where it is right now, confronting the hard truths that assessment reveals, and then (2) devise experiments to learn how to move the real numbers closer to the ideal reflected in the business plan.”
The development team needs to have good metrics that will prove that the product is getting better. Having more users or even positive opinions from some of them is simply not enough. Actually, if a company does not have proper analytics, the changes in the product may cause harm. It’s because, statistically, most of our hypotheses are wrong.
Three learning milestones.
The first one is establishing the baseline. It’s learning where your product is right now. So, if you build a prototype, you will get the first metrics on all your assumptions. It will tell you a lot about your current stage.
The second one is to tune the engine into the ideal one. As you have all the metrics from the first step, now make it better until you get to the point that you expected the metrics to be at the very beginning.
The third and the last one is the decision - should you preserve it or pivot into a different idea. If you get to the point, that you cannot improve the product anymore to your ideal numbers, maybe it’s time to pivot.
On the early stage of the product building, be sure that you are focused on learning, not just optimization. It’s easy to confuse those two. Optimization is making a better conversion rate on every state of the customer journey. Learning is understanding your customers, their needs, how they use your product, and so on. There is always room for optimization, but what you want to focus on is the big vision and developing the company towards it.
Bear in mind, that you always want to observe the conversion. The total number of users or purchases are not really relevant in the early stage. Instead, you want to observe what’s the increase in the percentage of people who registered or purchased your product. In order to do it correctly, use split testing - show one version of your product to 50% of your users, and the second version to the other half. The groups of people must be the same. Looking at metrics, you will know which version is better. To learn more, read some articles about it on the internet.
The three A’s
There are three important elements of having good metrics.
Actionable - everything that you’re measuring should be an effect of the actions of your team. For example, the total number of registrations is not good enough, instead, you should check the number of registrations per marketing channel with an interpretation of what actions your marketing team did to lead to that.
Accessible - The metrics should be easy to understand for every manager in the team. Very often, companies have complicated reports, that many people in the company don’t know how to interpret them. It’s better to make metrics simple, so everyone can understand them and act on them.
Auditable - Make sure that the data is credible to everyone. Sometimes things don’t go well, and employees must understand why that is.
Actions to take
Pivot (or Persevere)
“Every entrepreneur eventually faces an overriding challenge in developing a successful product: deciding when to pivot and when to persevere.”
For every startup the moment comes when it has to make this hard decision - are the metrics good enough, and can we keep continuing with the current vision of the product, growth strategy, etc? Or are we far away from the ideal number, and do we have to adjust the vision? This change is called Pivot.
Pivots are good because they help with growing a sustainable business.
Pivots require courage. Most of the entrepreneurs, who pivoted, tell that they regret not make that decision sooner. There can be a few reasons why this is so. For example, vanity metrics may lead to wrong conclusions. Or, when an entrepreneur has an unclear hypothesis, it’s impossible to experience a complete failure. And pivoting is scary, and entrepreneurs are afraid to acknowledge the failure.
Delaying the pivot is dangerous because the startup may run out of money before it validates the business model. And pivots are sometimes necessary to do so.
Take into account, that pivot may not only be necessary for the early stages of the startup, but also when the startup is growing well, and getting more and more customers. For example, you may grow virally and gain popularity fast in a certain demographic that is very limited. That means, you may grow fast for one year, and yet that may end eventually. Don’t put too much attention on the vanity metrics, but work on the whole engine altogether.
There are following types of pivots.
Zoom-in Pivot - A single feature of a product becomes the product itself.
Zoom-out Pivot - The whole product you build becomes just a feature in a much bigger product
Customer Segment Pivot - When you realize that your product solves a problem much better in a completely different segment group.
Customer Need Pivot - When you realize that the problem you’re trying to solve is not worth solving at all. But you encounter a similar problem that is worth solving.
Platform Pivot - When you change the platform.
Business Architecture Pivot - It’s when you change the business model from the high margin and low volume into the low margin and high volume. Or vice versa.
Value Capture Pivot - It’s when you change the monetization strategy or the revenue model.
The Engine of Growth Pivot - when you change the way you acquire new customers.
Actions to take
“The critical first question for any lean transformation is: which activities create value and which are a form of waste?”
The last part of this book is about growing your startup without sacrificing the agility and speed. Yes, you can keep your lean structure, the culture of innovation, and learning orientation even when you’re entering the enterprise level.
“The ability to learn faster from customers is the essential competitive advantage that startups must possess.”
Engineers from Toyota discovered already a long time ago, that working with small batches is much better than one big batch. It’s faster, safer, and very often more efficient. For a startup, it means to have releases of new versions very often, with just a few changes. There is less room for mistakes, the learnings are faster, and the whole development of the product is faster because of this.
It doesn’t mean that you should ship your product once a day, but go through the Build-Measure-Learn feedback loop faster than competitors, and gain a competitive advantage.
Actions to take
“Engines of growth are designed to give startups a relatively small set of metrics on which to focus their energies”
Eric Ries distinguishes 3 engines of growth.
The Sticky Engine of Growth.
This engine is about engaging the current customers and keeping them as long as possible. It’s the retention. If you don’t know how to engage your customers with your product, it doesn’t really matter how many new customers you acquire. To improve this engine of growth, look at your churn rate. It’s the percentage of users that stop using your product every month. Churn rate should definitely be low. Other metrics, like a number of acquired users, activation rate, or even the purchase rate, have nothing to do with this one.
The Viral Engine of Growth.
Social networks are a great example of how people share stuff. What’s important to notice, viral marketing is not a side effect of people using the product. It’s a strategy. A great example of this comes from the ‘90s when Hotmail started to grow. For each email that was sent using their platform, they put on the bottom “P.S. Get your free e-mail at Hotmail”.
Just like every other engine, in order to build this one, you need the feedback loop. And the metric you want to look at is called the viral coefficient. It’s the number of people that each user brings to your product. For example, viral coefficient = 0.1 means, that every user brings 0.1 more users to your product. Make sure that you have correct metrics in place, and you will be able to measure it correctly.
Be careful with your business strategy, charging users in the early stage, or even showing them advertisement on your platform may kill the viral growth. In order to make users talk about your product, you must make it as user-friendly as possible.
The Paid Engine of Growth.
This one is the most obvious engine of growth - you spend money to get more customers. The most important element here is that you spend less money to acquire a new customer than the total value of the customer is. The total revenue you get from one customer is called the lifetime value (LTV). The money you spend to get a customer is called the cost per acquisition (CPA). If the CPA is much lower than the LTV, your company will grow fast. If the CPA is equal to the LTV, your company is not growing.
As there are 3 different engines of growth, a well-established company should focus on all of them. However, for a startup, that is too much to do at once, so startups should put their focus only on one at the time.
Actions to take
“Should a startup invest in a training program for new employees?”
Having no training program in your company is bad, but creating one big training program at once is also not a perfect solution. Eric Ries recommends building, as he called it, an adaptive organization. One, that automatically adjusts process and performance to the current conditions. Improve the processes gradually, when you see they will improve the company.
In a startup, new employees make many mistakes. It’s important to notice those mistakes and ask why they really happened. If the mistakes are serious, consider making a training program to prevent them in the future. And yes, there are always many much more important things to do, but not having the training system in place may slow you down more than creating one.
The power of five whys
Leaders know this technique very well. It’s very powerful to find the root cause of any problem. Very often, with asking just one “why”, we see the problem on the surface. When we ask this question 5 times, we get to the root. For example:
Why do servers not work? -> Because there was an error in the source code.
Why did we allow them to have an error? -> Because nobody tested it.
Why did we have nobody run a test? -> Because it’s time-consuming, and we no one has time for it.
And so on.
When you use this method in your team, try to avoid pointing your finger at other team members. You want to find the root problem and create a solution to prevent it in the future, not blame others for their mistakes.
Actions to take
Companies, as they become larger, very often lose the capacity for innovation, creativity, and growth. But they don’t have to. In fact, they should aim to keep their structures in such a way, so that they keep innovating, creating, and growing.
Disruptive innovation means, that you do not only meet the current needs of the customers, but you also anticipate their unstated and future needs. Three elements to make it in your organization.
First, make sure that the innovation team has enough resources to innovate, but not too much. Having too much money is more dangerous for a startup to have than too little.
Second, the innovation team must be independent of the rest of the company. Also, it should be cross-functional - it should be able to do all the product development and marketing things by itself.
Third, the leader must have a personal stake in the outcome. There could be some stock options or other forms of equity ownership. Although, it doesn’t have to be financial. The important part is that the leader must feel that this is his creation and he has 100% responsibility for it.
When you have an innovation team in a big organization, make sure you protect the main organization from the team. This team will experiment with many things, and those things should not influence the main organization until they are ready to do so. When the reputation of your organization is important, or your competitors are watching you, it’s easy to make a mistake, that is difficult to reverse.