The Lean Startup

By: Eric Ries



Each and every day, new entrepreneurs around the world are starting their own enterprises. Most of them are armed with amazing ideas for new products or services (or so they think), but not necessarily the tools required to building an enduring business. 

Enter Eric Ries. 

Ries has seen numerous businesses start and fail - most notably his own. Through these false starts, he has learned many lessons and has gone on to have great success in building a multi-million dollar enterprise and coaches others to do the same.

For the next 12 minutes or so, you will learn the key components to a system that - almost without fail - can lead you towards building a successful business. 

What does it mean to run a Lean Startup?

According to Ries, there are five principles that are critical to the success of a startup, and what makes a startup a “lean” one.

First is the idea that entrepreneurs are everywhere. They are the person who just lost their job in the recession and have struck out on their own. They are the person who has started and sold their first five businesses, and is now onto their sixth. These are the people we traditionally read about in magazines and books, and who are self-made success stories. Entrepreneurs are also found in global corporations, working on the next big idea.

The second idea is that entrepreneurship is management. The goal of an entrepreneur is to build a sustainable enterprise, and so there needs to be a new (and predictable) method of doing so – especially in the middle of the digital revolution we find ourselves in.

The third idea is that startups exist – not to make money or even to serve customers – but to learn how to build a sustainable business. This idea is the most critical to Ries’ entire premise, and it’s a revolutionary one.

Most entrepreneurs start a business with a new idea thinking (most of the time, incorrectly) they have an idea that will be a huge success. I mean, why else start a business and take all that risk, right?

They then plod along, hustling the hell out of that idea until it either succeeds or fails – usually in spectacular fashion. Ries however, argues that if an organisation can learn as quickly as possible what the marketplace values enough to pay for, they will be able to adapt their business and grow it into a sustainable enterprise. He calls this validated learning.

The fourth idea is the method in which companies should approach this task: build, measure and learn. The idea is to get back to “build” as quickly as possible after learning from the marketplace. The quicker you can get through this cycle, the faster you’ll learn what the market values, and the better chance you have of surviving and building a sustainable business.

Lastly is the idea of innovation accounting. Although it sounds sexy, this is actually the boring stuff that will make a company successful. It is how you measure the milestones you set and how you prioritise the work you get done.

What these five principles add up to is a new way of thinking about management. If you value innovation as a company – whether as a startup or a multi-billion dollar corporate behemoth – this is how you will accomplish it.

Build a product

The first thing you need to do is to build a quality product or service for the marketplace to create what is called the customer archetype. 

The purpose of the archetype is to humanise the target market for your business. By doing so, it will help guide the decisions you make about product development and allocation of resources moving forward. So, before you make anything, make sure you know exactly who you are making it for.

Next, know you will need to take a leap of faith at some point. No matter how much research you’ve done, and how certain you are of your chances of success, your new venture is going to have to make some assumptions concerning some very important things. The key here is to acknowledge and know when part of the plan requires you to take that leap of faith. 

A simple tool to help with this decision would be to analyze analogs and antilogs. According to Ries, there's no problem basing the strategy for your new business on the success (or failure) of others. 

For example, when Apple was building the iPod, they knew people would listen to music in public places wearing earphones based on the success of the Sony Walkman. This in itself answered a critical question for Apple. What they didn't know however, was whether or not people would pay for the music they were listening too. 

The antilog analysis for this question was to examine what other companies had done. Napster for example had just proven that people - in record numbers – would stop paying for music when offered a free (albeit illegal) alternative. So taking this into consideration, Apple decidedly built their now insanely successful business (iTunes) on a leap of faith, but knew exactly where risk lay. 

The next step on this process is to build a rapid prototype. Most of us know of or have purchased items from Zappos. Zappos is a billion dollar-a-year online shopping portal that started out as a rapid prototype by founder Nick Swinmurn.

Swinmurn's original idea was to build a brand new retail experience, which he could have pursued at great cost and risk. Instead, he chose to run an experiment to see if people would buy shoes online. He visited local shoe stores in his community and asked if could take pictures of the shoes they had in stock. He would then take those photos and put them up on a website. If people bought the shoes from him via the website, he would return to the store and buy the actual shoes at full price. There, for a cost of next-to-nothing (except time and energy), Nick had figured out that people would indeed buy shoes online.

There are a few important lessons to glean here:

First, always build what the startup community now calls a “minimum viable product” (MVP). It’s the smallest product or service that you can create and start generating learning from. Nick didn’t need anything more than a simple website to start Zappos, and it’s likely that you need a heck of a lot less than you think to launch your new product or service.

Second, you should be attempting to attract the early adopter market with this MVP. Since these early adopters know they will almost always get a product with “bugs” in it, you don’t need to worry about having the best possible product to launch. In fact, any effort beyond what you need for an MVP is considered waste because it wasn’t driven in a response to the marketplace in the first place. 

Measure and Learn

The job of a start-up is to figure out where they are at any given time, confront the cold hard facts, and then design experiments to move the numbers closer to what they've laid out in their business plan.

These come together in what Ries calls the 3 Learning Milestones – (1) establish the baseline, (2) tuning the engine, and (3) pivot (or persevere).

In establishing the baseline, you need to make sure you are setting the right metrics. One thing to be wary of are “vanity metrics.” In the web startup world for example, these metrics might include “website visitors," and in some cases even “registered users." These metrics are easily manipulated, and do not necessarily correlate to the numbers that really matter. In almost every case, they can and will lead you to focus on actions that, at best, limit your chances for success.

In order to prevent this from happening, you need to ensure your metrics meet the "3 A's Test." This stands for Actionable, Accessible and Audit-able. 

  • In order a metric to be actionable, it must demonstrate a clear cause and effect relationship so that you can take definitive action in response to it.
  • In order for a metric to be accessible, it must be easily understood and available widely to people in the company.
  • In order for a metric to be audit-able, you need to be able to go back to the source of data to prove that the metrics were telling the true (and entire) story.

One example of these kinds of metrics would be the ones used by Ries' company, IMVU during their startup phase. The company sold a 3D avatar/social networking service that could be described as a chat service where you can dress up your character. Using $5 a day in pay per click advertising, they were able to get 100 visits to their website to test their product. They considered each day’s visitors to be one cohort, and tested each cohort based on the following data points:

  1. Registration – How many people signed up;
  2. Activation – How many people then went on to actually login to their account; and
  3. Retention – How many people had one chat, how many people had five chats, and how many people became paying customers.

A good way to do this for your own business is to pick metrics in the following buckets: registration, activation, retention and referral.

Another great example is Grokit, which is a an online learning company. Grokit used this model in one-month sprints, using “split tests” (sometimes referred to A/B tests) to determine the effectiveness of the changes they were making to the product. Quite often, products get “improved” on when the CEO says that he heard from somebody that they didn’t like feature X, or a product engineer says that they can improve the product by doing Y.

Most often these changes have no effect on customer behaviour at all, and most times that critical fact goes unnoticed. This is a critical point to understand. When making improvements to your product, the only arbiter of whether or not it was successful is the metrics.

And when you are implementing an improvement to your product, you should be testing that improvement against a baseline to see what – if any – impact the change has on your business results.

This is the only way a company should be implementing a product development strategy.

Unless, of course, you somehow have millions of dollars somewhere that doesn’t need to be accounted for to anybody, and that you don’t need to provide a positive return on.

Pivot or Persevere

At some point, you will need to make a decision about whether or not your business strategy has a reasonable chance of success. This is the time where you will need to decide to either “pivot or persevere." Of course, if things are working well and you can see a path the great success and profits, keep working on the idea you’ve started with. Just be sure to base your decision on the cold hard facts.

A pivot is a fundamental change in business strategy. If you conclude that your business strategy isn’t likely to succeed, you can it. This is where the mindset of an entrepreneur truly comes into play. A true entrepreneur is always learning how to build a sustainable enterprise, not trying to make a single product idea a success.

There are many kinds of pivots your company can make: 

  • Zoom-in: Where a single feature of your product becomes the entire product.
  • Zoom-out: Where your product is too narrow to support a business, and you decide to make a broader product.
  • Customer segment: Where you realise that you are building a product that solves a need for a segment of customers that is different than the one you started with.
  • Customer need: Based on an intimate understanding of the customers developed during your iteration process, you realise that the need you were solving for isn’t very important. But you find new needs you can solve instead.
  • Business architecture: This is where you go from a high margin, low volume solution to a low margin high volume solution.
  • Technology: Where you realise you could solve the exact same problem using a completely different (and usually less expensive) technology.

Ries discusses other pivots your business could make in the book, but these spoke the most strongly to us and our business. It is important to realise that whatever pivot you make, it’s only the next hypothesis in your business model, and that it should be rigorously tested just like everything else.

Technology company Aardvark is another great example that took this idea and ran with it. An alternative to other search engines where the answer needs some form of human interaction, Aardvark allows people to search on questions such as, “where’s the best place to get a drink after the movies tonight?” This is not something Google will give you a great answer for.

Aardvark however will - but this wasn’t the first product launched by its founders Max Ventilla and Damon Horowitz. It was their sixth! And it was their sixth product in less than six months!

Because they had used the Lean Startup model of the MVP along with rapid prototyping and measuring the results, they found out very quickly that their first five products were destined to become flops. Aardvark was the first and only product that pointed to success, and not surprisingly, it’s the product that they are now growing using the same principles we’ve discussed so far.


Now that you’ve found the product that you know will help you create a sustainable business, you need to have sustainable growth.

According Ries, the only way you can build a sustainable business is when your new customers come from old customers and there are three ways to do this:

First, you can create a “sticky” growth engine. This depends on having a product or service that customers will continue to pay for over time. In this model, if you can bring in new customers at a faster rate than your old customers leave the service, your business will grow. The metric that you’ll want to pay the most attention to is your retention rate.

Second, you could create a “viral” engine of growth. In this model, you depend on your current customers to bring in your new customers. The most famous example of this is Hotmail, which was once a slow growth business struggling to get traction. That was until they decided to append each mail message you sent with an invitation for other people to sign up for the service, with a link directly to the sign up page. The metric for this engine is something called the “viral loop." If you can get each new customer to bring in one or more customers, the viral growth will continue.

The last engine for growth is the “paid” model. In this model, you take the profits you’ve earned from your old customers, and invest it into advertising (or any new business development tactic) to attract new customers. The metrics to pay attention to in this case are the Lifetime Customer Value (the profits you’ll make off each customer over the lifetime of doing business with you) and the Customer Acquisition Cost. As long as your LCV exceeds your new customer acquisition costs, you will grow.

Ries makes the point that most companies only have the bandwidth to specialise in one of these engines of growth, and the time it takes to test and fine-tune everything in any particular engine is too great to split your focus. So make sure you pick the engine that is best for your business.

Of course, if you start on one of the engines and you find it isn’t going to work out as you’d hope, you can always do the “growth engine” pivot and focus on a different one moving forward.


So there you have it – everything you need to know in order to begin your Lean Startup journey. If you are truly an entrepreneur, you’ll take Ries’ advice and get passionate about building a sustainable enterprise rather than seeing your new idea succeed at all costs.