I'll start off with my review of "The Lean Startup" and then my notes/ book report is below. Definitely worth the read and you can buy the book on Amazon here. Or better yet shop at a local independent bookstore like the Brookline Booksmith, where I got my copy.
This book is a mini business course providing quantifiable ways to innovate and manage companies in uncertain climates. From recognizing which growth strategy you are using and aligning that with the correct actionable metrics to make sure you are driving the factors of growth to how to manage people in an environment that is different than a traditional established business. It cuts through the feel good idea of entrepreneurship and innovation, as if just sitting around an amazing idea will pop into your head, and provides a structure to go about testing different hypothesis and enabling growth.
“The Lean Startup” thesis: Entrepreneurship is management. “A startup is an institution, not just a product, and so it requires a new kind of management specifically geared to its context of extreme uncertainty (8).” “Old management methods are not up to the task. Planning and forecasting are only accurate when based on a long, stable operating history and a relatively static environment. Startups have neither (9).”
Goal of “The Lean Startup”: “Comprehensive theory of entrepreneurship which addresses vision and concept, product development, marketing and sales, scaling up, partnerships and distribution, and structure and organizational design.
- Entrepreneur – someone within the “startup ecosystem regardless of company size, sector, or stage of development (27)”
- Startup – “human institution designed to create a new product or service under conditions of extreme uncertainty (27)”
- Validated Learning – “the process of demonstrating empirically that a team has discovered valuable truths about a startup’s present and future business prospects (38).”
- Build-Measure-Learn feedback loop (BML) – Ideas –>Build–> Product–>Measure–>Data–>Learn–>Ideas
- Minimum Viable Product (MVP) – the version of the product that enables a full turn of the BML feedback loop with a minimum amount of effort and development time (77)
- Pivot – “a special kind of change designed to test a new fundamental hypothesis about the product, business model, and engine of growth (173)”
- Value Hypothesis- “whether a product or service really delivers value to customers once they are using it (61)”
- Growth Hypothesis – “how new customers will discover a product or service (61)”
- Traditional Rate of Growth – Profitability of each customer, cost of acquiring new customers, and repeat purchase rate of existing customers (116)
- Churn rate – the rate at which a portion of the customers for a certain period fail to stay engaged (210)
- Viral coefficient – “measures how many new customers will use a product as a consequence of each new customer who signs up (213)”
- Lifetime Value – revenue per lifetime of each customer (216)
- Cohort analysis “looks at the performance of each group of customers that comes into contact with the product independently (123)”
- Success: “Success is not delivering a feature; success is learning how to solve the customer’s problem” – Mark Cook, Kodak (66)
The book is divided into three sections: Vision, Steer, and Accelerate. Vision takes a look at traditional management and prescribes a new way forward. Steer provides the operating processes to create the new structure. And Accelerate refers to fine tuning operations to make it as efficient as possible.
Most traditionally managed companies are organized into function based departments such as sales, accounting, marketing, human resources, production etc (19). With this setup people are responsible for good performance in whatever specialized area they are a part of (19). Work is passed between departments and often requires sign off or approval before moving on.
On the strategy side traditional management is good at identifying the assumptions in the current business which are often drawn from well known facts (81). Their goal then is optimization: traditionally incremental benefits for incremental effort, better design, more efficient engineering etc…(126). Success is measured as creating a feature or product and passing it off. Individuals see success or efficiency as one in which they did their specific function all day with few interruptions. “Failure to deliver results is due to either a failure to plan adequately or a failure to execute properly (24).”
While traditional management has been very effective and brought about much success it doesn’t align with a startup’s goal. The goal is to innovate and create and this is an extremely uncertain environment. “Because startups often accidentally build something nobody wants, it doesn’t matter much if they do it on time and on budget (20).” It is a likely possibility that the company is executing a plan with discipline that doesn’t make sense (127). “Which of our efforts are value-creating and which are wasteful (47)?”
This brings us to the second section, “Steer,” where a new management system called Validated Learning is used. To create a concrete operating structure data or some sort of feedback is needed. Feedback (qualitative and quantitative) is generated when customers interact with products. Here Eric introduces the BML feedback loop: build a minimal viable product, evaluate leap of faith assumptions using rigorous testing, and a learning method to determine pivots. This is what he calls validated learning.
“Although we write the feedback loop as Build-Measure-Learn because the activities happen in that order, our planning really works in reverse order: we figure out what we need to learn and then work backwards to see what product will work as an experiment to get that learning (201).”
We need to hypothesize the answers to the two riskiest assumptions first in order to build the MVP that will allow us to test those answers. Inherent in the creation of the MVP are qualities that allow for measurement of impact (77) such as building into the new feature split testing (140). Here are four questions designed to answer the value and growth hypothesis’:
- Do consumers recognize that they have the problem you are trying to solve?
- If there was a solution, would they buy it?
- Would they buy it from us?
- Can we build a solution for that problem?
Whether it is through one MVP or several a minimum viable product needs to released to establish a baseline (118). “An MVP allows a startup to fill in real baseline data in its growth model — conversion rates, sign-up and trial rates, customer life-time value, and so on… (119)”
Measuring the data or the first learning milestone is the second part of the BML loop. In contrast to traditional gross metrics such as revenue and customer growth we would need to use cohort metrics. Traditional gross metrics or what the author calls vanity metrics don’t lead to action. They just provide an overview. Cohort analysis “looks at the performance of each group of customers that comes into contact with the product independently (123).”
Split testing (one design versus another) shows what is and what is not important to the customers rather than trying to infer a cause and effect relationship after the fact. Sometimes this is called A/B testing as well. “Lean Startups incorporate it directly into product development (137).”
This is innovation accounting –> actionable data that demonstrates clear cause and effect (143) rather than after the fact educated guesses. Two key factors of the data: it must be accessible and as simple as possible (144). To make sure everyone can trust the data it must be audit-able and can stand up to spot checks (147).
The last step in the loop is Learn. Learn = pivot or persevere. The goal is to learn whether to pivot or not as soon as possible. This creates capital efficiency, both financial and intellectual.
Using this data we need to determine if we should pivot or persevere. “A pivot requires that we keep one foot rooted in what we’ve learned so far, while making a fundamental change in strategy in order to seek even greater validated learning (154).” Once the two main assumptions are answered each further iteration is aimed at one of the drivers of the growth model. If the drivers of the business model are not moving than we are not making progress → time to pivot (120).
There are many types of pivots as seen below:
- Zoom in – a single feature becomes the whole product
- Zoom out – the whole product becomes the single feature of a much larger product
- Customer segment – realization that the product solves a real problem for customers but not the kind originally planned to serve
- Customer need – discover more important needs than the ones originally being solved
- Platform – change from an application to a platform or vice versa
- Business architecture – high margin, low volume (B2B) or low margin, high volume (B2C)
- Value capture – how to capture value
- Engine of growth – viral vs sticky vs paid growth
- Channel – different delivery channels
- Technology – switch technologies
The last section is about accelerating, or getting through the BML loop as quickly as possible. This is the hard work that truly creates innovation.
Here he takes some cues from lean manufacturing principles. The critical question being how to understand which activities create value and which are wasteful (181). Here are some techniques to increase the efficiency of value creating activities (BML –> validated learning).
He introduces just in time scalability: “conducting product experiments without making massive up-front investments in planning and design (182).” This means batch size should be as small as possible. “Batch size refers to how much work moves from one stage to the next (184).” Large batch operations are a traditional management approach with a lot of work being done upfront and hindering the ability to catch any issues until almost done. “What if it turns out that the customers have decided they don’t want the product (185).” “The biggest advantage of working in small batches is that quality problems can be identified much sooner” and these questions are easily answered (187).
This can be misinterpreted as pulling/creating only what the customer wants (201). However customers don’t always know what they want so the goal of the Lean Startup is “responding to pull requests in the form of experiments that need to be run (201).”
Sustainable growth comes from one of three engines: paid, viral, or sticky. “New customers come from the actions of past customers (207).” Growth comes from four places: word of mouth, paid advertising, repeat purchase, and as a side effect of product usage (208). The goal is not to make mere optimizations but each experiment needs to be aimed at a factor of a engine of growth. Actionable metrics need to be matched up to the appropriate growth strategy of a company.
- Sticky – attract and retain customers for the long term. Growth comes when the rate of new customers exceeds the churn rate (211)
- Viral – “person to person transmission as necessary consequence of normal product use (212)”. Rate of growth is driven by the viral coefficient (needs to be greater than 1 for exponential growth)(213).
- Paid – rate of growth depends on the cost of acquisition for each additional customer being less then their LTV.
A note from the author: “…in my experience successful startups usually focus on just one engine of growth (219).”
He then gives us the “5 Whys” framework to manage a team as they try to solve problems as they come up. It is simply asking “why” five times to get to the heart of the problem. In doing so one often finds the root cause moves away from a technical fault to a human error (231). One advantage is that it is an automatic speed regulator (233). However once you truly evaluate an issue teammates point fingers to determine blame. He gives several pointers to short circuit this.
Make sure everyone affected by the problem is in the room for the meeting (whoever is left out is often the target for the blame). The position of management’s position should be shame on us for making it so easy to make a mistake. The environment must create trust & empowerment (236). To do this be tolerant of all mistakes the first time but never allow the same mistake to be made twice.
And finally the last chapter tackles what everyone is looking for, innovation. Hand offs and approvals slow down the BML loop and inhibit learning and accountability. Move to a cross functional structure so that a rep from every department is at the table (255). A sense of ownership is needed for everyone to feel respected and involved. Give credit where credit is due.
He suggests using an innovation sandbox; protect the parent org from the startup (261-2). Create a true split-test experiment that affects only the sandboxed parts of the product, service, customer segment etc…However:
- One team must see the experiment through
- No experiment can run longer than a specified amount of time
- No experiment can affect more than a specified number of customers
- Evaluate on the basis of a single standard report of 5-10 actionable metrics
- Every team and every product must use the same metrics for evaluation
- Any team that creates an experiment must monitor metrics and customer reaction while in progress and abort it if something catastrophic happens
And that’s how to create a Lean Startup…