My pick: The Lean Startup by Eric Ries
Approximate read time: 10 minutes
Approximate read time: 10 minutes
This is How to Think Like a Startup
I was nineteen when I moved to Seattle. Not having a job, a phone, or place to live, I did what any enterprising young lad would do moving to a new city alone- I moved into a pay-by-the-week former-sorority-house-turned-neglected-tenement. My roommate was a Vietnam vet who was out of his goddamned mind. He yelled in his sleep a lot. I didn’t sleep much. Had we bonded more, perhaps I would have had more insight into his triggers and moderated the little waste created by his affinity for oranges. He was, after all, a very tidy roommate.
Turns out, the mild weather of the Pacific Northwest is great for accommodating fruit flies in November. I returned to the room one afternoon to find Charlie had crossed the wire. We were in the shit. The fruit flies had come, and it was either kind of annoying, or a race against time before they organized and armed themselves. Roomie contracted Malaria in Nam, we could not lose ground.
I know, Drosophila as an infectious vector. But I had to sleep next to this guy. My job was to stand by and confirm that his plan of action was sound, that a baited container propped up with a stick tied to a string was indeed our best shot at neutralizing the threat. Sure, there were, of course, easier and more sure-fire ways to exterminate a swarm of fruit flies, faster ones too, but this was his delusional ride and I could not be more entertained.
"Work smarter, not harder"
All startups experience a fog of unknowns in the search for product/market fit. Absent a systematic approach for effectively minimizing the unknowns, the chances of viability for the business are slim. It’s often too late before many realize they were trying to catch fruit flies with a rodent trap.
Lean Thinking
When businesses invest in complicated assumptions, tiny errors can lead to catastrophic outcomes. At its core, the Lean Startup method’s aim is to discern the value-creating activities of a startup from the waste, exploiting those activities that create value and eliminating all those that don’t. It emphasizes fast iteration, agile development, and customer insight to accelerate turning a vision into a business. In short, if it doesn’t provide value to the customer, it’s waste.
The Lean Startup is not a collection of tactics, nor is it a blueprint of steps to follow. It’s a framework, a principled approach to new product development. There are five principles of the Lean Startup method:
1. Entrepreneurs are everywhere
The Lean Startup approach is applicable to all entrepreneurs, whether starting a business working out of the garage or as a team of innovators inside a large company.
2. Entrepreneurship is management
A startup requires a new kind of management geared for an environment of extreme uncertainty. It’s the failure to manage startups, as startups, that so many flame out. Old management methods fail here. Innovation is not a centralized thing and entrepreneurship must be cultivated within the organization. Planning and forecasting, and a solid strategy based on thorough market research, rarely translate to success when starting from zero.
3. Validated learning
If you’re introducing a new product into a new market, things are going to get weird. Customers rarely know what they want, and your business plan relies on the assumption that you know what they want. Without validating the truth of what assumptions you’ve made, all you’ve got are best guesses. Building a business on that is just crazy.
So, minimize the unknowns- What the customers’ needs are, if your product matches those needs, if you’re in the right market, if elements of your strategy are working... It’s a lot less expensive to know now whether or not you’re on the right track or simply delusional, and how to adapt strategies accordingly.
The lean thinker validates the elements of the founders’ vision scientifically by experimentation, straight out of the gate. The earlier you get the data, the sooner you know where to invest your energy. It requires blowing a business plan up to empirically demonstrate or refute as many of the assumptions as possible. Small-scale, micro-experimentation reveals costly inefficiencies in the operation equating to wasted work, and small problems that could develop into very big problems with a lot of momentum.
The objective is to get through the Build-Measure-Learn feedback loop as quickly as possible, so you know you’re not in crazy town and following the delusional path that leads to achieving failure, i.e. successfully executing an erroneous plan.
4. Build-Measure-Learn
The products a startup builds are experiments. The outcome is learning how to build a sustainable business. By measuring the interactions between the products and customers, we get data that we can learn from to validate or refute hypotheses to calibrate our products or elements of the business plan. Develop a new idea from what has been learned, and the feedback loop starts over.
The Build-Measure-Learn Feedback Loop
The more times we go through the Build-Measure-Learn feedback loop, the more data we collect, and the more we learn. Time and capital are scarce resources for a startup, so it follows that the faster we’re able to go through the cycle, the more quickly we can adjust course, with these limitations in mind.
Minimize total time through the feedback loop.
Because most professional training focuses on one element of the loop, it’s easy to see how a hang-up can slow down the process, and what that means for the operation as a whole. Forget traditional roles and departments until you have a reliable and repeatable roadmap- they’ll only slow you down. They’re expensive and time-consuming, problems which are compounded when momentum swings in the wrong direction.
Getting through the loop quickly requires an expedient build for both product and non-product development. Efficiency requires quick iteration, and a focus on rapid release and implementation rather than a perfect design.
For a product, it’s the simplest and barest you can get away with- buggy and low-quality, only the essentials, and probably not the one you envisioned. Early on you don’t know what quality is anyway, you don’t even know who the customer is. It serves its purpose- a working model to put in the hands of innovators and early adopters, to test as fast as possible, and to serve as a model for future iterations based on customer feedback. This minimum viable product (MVP) is your first experiment. It will teach you what the customers want, and what is a waste of time and resources. If any feature, process, or effort does not contribute to your learning, remove it.
Consider what this means for a process like batch sizes. Large batch sizes are all produced at one time, lengthening the manufacturing process. They also increase inventory, and large inventories have a way of getting larger- not to mention one defective unit could turn all inventory to waste. Small batch sizes, on the other hand, produce a finished product in a short amount of time. It allows for faster testing, with multiple iterations possible among early adopters. Mistakes are found earlier, and if a customer cancels the order, you’ll know sooner. It also sets the business up for pull manufacturing with an ideal goal in mind: single-piece flow along the entire chain, resulting in a lower total system cost.
Non-product development activities should also be as stripped-bare as possible. Anything that doesn’t contribute to learning- junk it. Align marketing and sales objectives with the validated learning from the MVP/customer interaction. As learning evolves, strategies become more sure-footed, tactics more precise. You build marketing and sales plans based on what you know, not riddled with uncertainties that when refuted make the wheels fall off.
Product and non-product activities go through the loop together, they work parallel to each other. What is learned from one informs the whole. A forced time constraint is helpful here- new information and problems can result in over-planning and getting bogged down in details. Limiting the time in the loop to one-month sprints, for example, and using actionable metrics (those that give a real-world picture of the business’s progress rather than ones that make you feel good about yourself but teach you nothing) accelerate cycle-time and learning, and the decision to pivot or persevere.
5. Innovation Accounting
A business can only know whether or not it’s improving by measuring its performance over time. The right metrics give insight, they show what’s working for the business and what’s not.
Innovation accounting simply means using the right metrics to measure progress, set up milestones, and prioritize work. A metric should be actionable, accessible, and auditable.
Actionable- it demonstrates clear cause and effect.
Accessible- It is in plain language and easy to read by everyone on the team.
Auditable- Managers should be able to spot check the data with actual customers, and it gives managers first-hand insight into customer behavior. Metrics that fall under the guidelines of the Three A’s help the team tune the engine of growth: improving customer behavior for the better, e.g., reducing customer acquisition costs while increasing customer retention.
Accessible- It is in plain language and easy to read by everyone on the team.
Auditable- Managers should be able to spot check the data with actual customers, and it gives managers first-hand insight into customer behavior. Metrics that fall under the guidelines of the Three A’s help the team tune the engine of growth: improving customer behavior for the better, e.g., reducing customer acquisition costs while increasing customer retention.
There are three steps to innovation accounting for every cycle through the feedback loop: The first is using an MVP to establish real data. Second, using the data you collect as a baseline to move toward the ideal- improving customer behavior for the better. Third, Pivot or persevere.
Once we have a working prototype in the hands of customers, we can begin to collect data and analyze. Using major, informative, and real metrics such as churn rate, cohort analysis, viral coefficient, and cost per acquisition give the business a real understanding both of how to grow and where growth is coming from. They also reveal when to focus on customer retention without further investment in sales and marketing.
The Three Growth Engines for Business
Sticky- retaining customers long-term, focusing on maintaining low customer attrition
Viral- Word of mouth, the product sells itself (very difficult to pull off)
Paid- Advertising and storefront. Money is paid to attract and maintain customers
Each growth engine has its advantages depending on what the product is and what customers it is meant to attract. All can be used, but a company will best optimize by focusing on one growth engine at a time. A kind of testing triage must be deployed, giving priority to the riskiest assumptions first- after all, if you can’t mitigate these, why test the others? This means performing big experiments that lead to leaps in learning, and focusing on the metrics that matter. The goal as with everything else, is sustainability. This means knowing the thresholds of the engines employed, and knowing what the indicators are of a growth engine that’s being exhausted.
The right market is everything. The best product in the world is going to fail in the wrong market. When reviewing the data sets keep in mind that poor numbers don’t indicate a problem with the product in most cases. Knowing the metrics lets you know if you’ve got an effective growth engine, or if product development is responsible for growing numbers. Constant tuning and tweaking of the product produces mediocre results. It’s often not a question of adding features to attract new customers, it’s a matter of positioning and marketing.
Pivot when stagnant or negative results show the drivers of the business model aren’t moving- one foot in the ground, planted by what testing has showed us is true, the other changing direction toward progress. The sign of a successful pivot: new experiments are more productive than the preceding ones. Each pivot unlocks new information, and the cycle continues and establishes a rhythm. Baseline, tune engine, pivot or persevere. Wash, Rinse, Repeat. The sooner we recognize it is time to pivot, the more efficient we become as a business- we waste less time and money.
Pivots allow for faster iteration, hypothesis creation and validation. Pivoting essentially splits the operation into a control/independent variable experiment. A customer segment pivot, for example, would keep the functionality of a product, while changing the audience focus. For a successful pivot, the team has to be open to accepting whatever the data says. Keep the faith. It’s important to remember your company won’t be the same company you started as. Even when you’ve achieved product/market fit, the time will come when you must pivot again.
####
Why this book is my pick
In The Lean Startup Eric Ries combines the Customer Development methodology outlined by Steve Blank in The Four Steps to the Epiphany with the Lean Manufacturing concepts employed by Toyota to make it the world’s number one manufacturer of automobiles, to provide such a methodology for startups. It’s a guide for finding customer/market fit, and calibrating a business model to steer a startup toward its purpose- to learn how to build a sustainable business.
A killer manual and reference, The Lean Startup is an essential read for managers of startups, and product managers and marketers in general. I also got a nice Zen and The Art of Motorcycle Maintenance vibe from it. If you’re going to ride the machine, you should know how to be your own mechanic. If you’re going after a platoon of fruit flies, all you need is a bottle of Windex and a glass of bourbon.
A killer manual and reference, The Lean Startup is an essential read for managers of startups, and product managers and marketers in general. I also got a nice Zen and The Art of Motorcycle Maintenance vibe from it. If you’re going to ride the machine, you should know how to be your own mechanic. If you’re going after a platoon of fruit flies, all you need is a bottle of Windex and a glass of bourbon.