In traditional companies, the management strategy consists of developing plans, setting milestones, delegating tasks, and measuring results. This strategy works for established companies that have a long history to know what works and what doesn't. This strategy does not work for a startup because startups don't have the data to predict the future -- everything is unknown. Startups need to focus on finding the right metrics and remain flexible and adaptable to change directions quickly.
For any start-up, the ultimate objective is to establish a profitable and sustainable business model. Having great plans and efficient execution are futile if the business is not sustainable. Are people willing to pay for your product? Are they willing to pay for it years later?
Use validated learning to determine if people are willing to pay for a product. Validated learning is a process of constant learning through a scientific approach in which hypotheses are tested. Talking to real potential customers is a good way to conduct validated learning.
Zappos tested its hypothesis about customers' willingness to buy shoes online by displaying photographs of shoes in a fake web shop.
Founders create their products on a leap of faith. Since there is no concrete evidence for success, they must test their assumptions about a product's value and growth hypotheses. The value hypothesis tests whether early adopters will embrace the product, while the growth hypothesis tests if the product will find a bigger market.
Facebook was able to validate both hypotheses with its active registered user base and explosive user-activation rates, leading to early investor confidence and millions in funding.
Create a minimal viable product (MVP) to get customer feedback as quickly as you can. Do not spend too much time on developing a polished product without knowing if there's actually a demand for it.
The MVP should be as simple as possible and contain only what is needed for customers to experience how the product would work. The MVP can be a simple prototype or even a smoke test, such as pretending to sell a fake product, or in Dropbox's case, creating a video to demonstrate how the product works.
Startups should develop products quickly and iteratively through the build-measure-learn (BML) loop to advance their product.
Build: build a simple MVP or smoke test
Measure: gather feedback from customers; obtain overall data but talk to individual customers as well
Learn: use the info you collected to generate ideas to better your product; go back to the build stage to create an improved version and repeat the loop
Use split-tests to determine if a feature is of value or waste. A feature of value generates more revenue or customer retention, a feature of waste doesn't. Split-tests involve creating two versions of the product, one with the new feature and one without it, and measuring customer response.
To avoid becoming a "zombie startup" that persists in selling an unwanted product, entrepreneurs should be willing to pivot and make fundamental changes to their business.
Pivots can involve redefining the core value of a product, pursuing a different customer segment, or changing the main sales channel.
Pivots should be tested through hypotheses and data. To facilitate this process, entrepreneurs should hold regular pivot meetings to assess their business model and determine whether a pivot is needed.
Groupon was an activism and fundraising platform before it pivoted to the e-commerce marketplace it is today.
There are three engines of growth to avoid stagnation: sticky, viral, and paid. You can use all three engines concurrently, but it is more effective to focus on one engine to get more measurable results.
Sticky: retain existing customers by offering new features/improving service
Viral: word-of-mouth marketing, your customers do the work
Paid: paid investments in marketing; important to ensure user lifetime value is lower than cost per user acquisition
To achieve a sustainable business model, it is important to find the right core metrics (e.g., number of customers, customer recommendation rate) and avoid vanity metrics, which are misleading and not helpful in achieving long-term success.
The core metrics will vary from startup to startup.
Metrics such as social media attention or number of hours worked are vanity metrics and should not be used as indicators of success.
Cohort analysis (e.g., comparing a specific metric from this year vs last year) can be useful to compare how different groups of customers behave over time and determine whether metrics are improving or stagnating.