Review: “The Lean Startup” by Eric Ries
In the presence of uncertainty, teams are building products a lot of times that nobody wants or needs. The work can be done on time and on budget with the right team and processes in place. However, it doesn’t matter much because the end result is a considerable waste of resources and efforts. Eric Ries provides a different way to build products that should address this problem. “The Lean Startup” is one of the classics so I would like to share a review on it.
Summary
The book enumerates two different ways of building products that usually end up in failure:
- Building a good plan, strategy, and thorough market research. This is the most common way to build products at large corporations. However, planning and forecasting are only accurate when based on a long, stable operating history and a relatively static environment.
- “Just do it”: This is the most common way to build products at startups. Many times the management and processes are the problems in the traditional businesses so chaos is the answer for startups.
Both approaches can provide the desired outcomes however they pose significant risks to the execution. The first approach definitely helps to avoid the risk of doing the work that doesn’t have a product-market fit. However, it starts the waterfall process and as a result, it is not capable to quickly adjust to the changing realities of the environment typical to the tech startups. The second approach is great at producing work quickly. However, it highly depends on the personalities that reign in the chaos to stay on track while searching for a product-market fit. And as a result, the organization can easily pivot in the wrong direction. The Lean Startup framework provides a different way of building products that should help mitigate those risks.
The Lean Startup is based on five principles:
- Entrepreneurs are everywhere — create new products and services under conditions of extreme uncertainty
- Entrepreneurship is management — a startup is an institution, not just a product.
- Validated learning — running frequent experiments that allow testing the elements of the vision
- Build-Measure-Learn — turn ideas into products, measure how customers respond, and then learn where to pivot or persevere
- Innovation accounting — how to measure progress, how to set up milestones, and how to prioritize work
The core idea is the concept of validated learning. Overall, learning is a common excuse for a failed execution. It is always easy to rationalize the results after the fact. Learning should be focused on figuring out which parts of the strategy are working to realize the vision — learn what customers want, not what they say they want, or what you think they should want. Most of the time customers don’t know what they want in advance so asking them might not yield the desired learning. Validated learning is focused on the idea of making a hypothesis and validating it quickly in the live environment.
The main hypotheses to validate are value hypothesis and growth hypothesis. The value hypothesis tests whether a product or service really delivers value to customers once they are using it. The growth hypothesis tests how new customers will discover a product or service. Both of them give rise to the tuning variables that control startup’s engine of growth. It is a mechanism that startups use to achieve sustainable growth.
There are three engines of growth defined in the book:
- sticky engine of growth — attract and retain customers for the long term
- viral engine of growth — customers do the lion share of the marketing
- paid engine of growth — costs to sign up a new customer
More than one engine of growth can operate in the business however, the author suggests that it is always better to focus on one engine.
At any point the company can spend its energy in the following activities:
- finding new customers
- serving existing customers better
- improving overall quality
- driving down costs
Metrics that define the engine of growth help to figure out which activities to prioritize and focus the energy on. That triggers the new set of hypotheses. Their validation should provide positive improvements in the core metrics of the engine of growth.
The typical execution at the startup should start with figuring out what needs to be learned. Then, we need to figure out what needs to be measured to produce validated learning. And finally, figure out what product needs to be built to run the experiment to get that learning. The goal is to keep the feedback loop very short in order to avoid waste.
That process is formalized as the Build-Measure-Learn feedback loop. Build phase is focused on producing the minimum viable product (MVP) — learning milestone. Measure phase is focused on innovation accounting. Innovation accounting tracks the metrics that drive the growth of the business. The metrics should be actionable, accessible, and auditable. Learn phase is focused on the interpretation of the results: either pivot or persevere depending on the impact on the core metrics.
The book describes different types of pivots that can be taken based on the results of the experiment. A few of the most interesting pivots are:
- zoom-in pivot — single feature in a product becomes the whole product
- customer segment pivot — the product solves the real problem but for a different customer than originally planned.
- business architecture pivot — switch between “high margin, low volume” and “low margin, high volume” business architectures
- engine of growth pivot — switch between the viral, sticky, paid growth models.
The author claims that the Lean Startup framework can be used in large corporations as well though it is much harder to do so without the proper support from the senior management. In general, failure to deliver results is due to either a failure to plan properly or a failure to execute. You can’t come up empty-handed and ask for another year citing the amount that was learned and that you are close to a new line of business. Therefore, large corporations should cultivate entrepreneurship by providing certain conditions. Eric provides an example of SnapTax product built by Intuit. Intuit didn’t hire external people such as star entrepreneurs to build that product — they used a local team. The team didn’t face constant meddling from senior management — an island of freedom was created. There was no huge budget or team — the product started with a team of five.
My experience
Consumer products seem more favorable for the Lean Startup than enterprise products. You are trying to throw things at the wall and see what sticks on it. So it is essential that you can iterate as fast as possible and produce MVPs before the runway is over. However, it is more challenging to build quick prototypes for large enterprises due to the quality expectations, sheer challenge of integration, and long procurement cycles. The best solution is to focus on the early adopters. They are more inclined to sacrifice the quality and more willing to run the experiments. This applies to both types of products but even more so to the enterprise products where it is extremely difficult to identify early adopters.
Another challenge is to find the fine line between the speed of execution and the quality of the MVP. If the team is only focused on the speed, the technical debt will start to accumulate. It will start a snowball effect of spending more and more time on the maintenance and emergency calls. At some point, it will require a significant amount of resources to address it.
Finally, the process reminds me of a gradient descent algorithm where a local minimum is a product-market fit and the quality of the hypotheses defines the speed of the convergence.
Verdict
Recommended for entrepreneurs and product managers
