What Is A Lean Startup?

Editor's Notes

• #2: Hi, I’m Ash with WiredReach and I’m going to be talking about Lean Startups. To get a sense of who’s in the room, How many people here have heard about Lean Startups before? How many people are in a startup or have been in a startup? While lean startup came out of the startup up world, the concepts themselves are fairly general and can be applied to companies of any size or industry that are building new products or services.
• #3: Here’s what I’m going to be covering today. I’ll talk about why building new products can be so challenging, briefly cover some lean startup theory, and then illustrate them with a case study from my startup. Let’s first look at how software is built.
• #4: This is the classic waterfall model. It only works if you fully understand the problem and solution upfront. Most software is not built this way anymore because it’s too hard, expensive or both to define the solution upfront. Which is where Agile comes in.
• #5: Agile takes a problem and helps you iterate your way to the right solution. A key concept in agile is that of the “product owner” or in-house customer that drives the definition of the problem. But what if you are building a new product or service and don’t yet have any customers or even know what problems to solve?
• #6: That is the startup condition and what Lean startup tries to address.
• #7: This is how most startups typically build new products.
• #8: There are 2 main concepts in lean startup: The first is building products with parallel customer and product development tracks. The second is the emphasis on validated learning about customers over achieving product milestones or lines of working code.
• #9: There are 2 main concepts in lean startup: The first is building products with parallel customer and product development tracks. The second is the emphasis on validated learning about customers over achieving product milestones or lines of working code.
• #10: This is what a combined customer development + product development feedback loop looks like.
• #11: This is another way to visualize iterations in a lean startup. You start with some ideas which you build into code then measure and learn from it. Build/Measure/Learn is the fundamental iteration loop and startups that succeed are those that manage to iterate enough times before running out of resources.
• #12: Taking this back to the Agile world, another way of thinking about validated learning is adding a fourth state to a story card. After the feature is tested and deployed, someone still has to verify if the feature was any good.
• #13: With that I’d like to jump right into a case-study of how I have been applying these techniques in my startup. I have been in business for several years and have launched 2 products. The first product was built using release early, release often… It started out as a very simple application but quickly grew out of hand because we were trying to be all things to all people. I eventually hit the reset button and stripped about 60% of the features that weren’t getting used. I had started reading about customer development and lean startup and decided to apply those techniques to my second product. You can see some of the differences but the biggest payoff for me was being able to define and prioritize my success metrics more clearly.
• #14: With that I’d like to jump right into a case-study of how I have been applying these techniques in my startup. I have been in business for several years and have launched 2 products. The first product was built using release early, release often… It started out as a very simple application but quickly grew out of hand because we were trying to be all things to all people. I eventually hit the reset button and stripped about 60% of the features that weren’t getting used. I had started reading about customer development and lean startup and decided to apply those techniques to my second product. You can see some of the differences but the biggest payoff for me was being able to define and prioritize my success metrics more clearly.
• #15: The approach I had taken with the first product was building something I thought people wanted and then testing it. Keeping it small helped but a lot of my underlying assumptions were wrong and had to be reworked over time. Steve Blank literally wrote a book on customer development where he asserts that all the answers lie outside the building and emphasizes engaging customers even before the product is built. I used a problem presentation to identify the top 3 problems most important to my customers which helped me define what needed to go into a minimum viable product.
• #16: The approach I had taken with the first product was building something I thought people wanted and then testing it. Keeping it small helped but a lot of my underlying assumptions were wrong and had to be reworked over time. Steve Blank literally wrote a book on customer development where he asserts that all the answers lie outside the building and emphasizes engaging customers even before the product is built. I used a problem presentation to identify the top 3 problems most important to my customers which helped me define what needed to go into a minimum viable product.
• #17: The approach I had taken with the first product was building something I thought people wanted and then testing it. Keeping it small helped but a lot of my underlying assumptions were wrong and had to be reworked over time. Steve Blank literally wrote a book on customer development where he asserts that all the answers lie outside the building and emphasizes engaging customers even before the product is built. I used a problem presentation to identify the top 3 problems most important to my customers which helped me define what needed to go into a minimum viable product.
• #18: We released BoxCloud on a biweekly schedule and used feature requests as the primary means to drive the product. Unused features are a form of waste and Eric Ries emphasizes the importance of validated learning through fast build/measure/learn loops. In CloudFire, we use Continuous Deployment to release software almost daily and incorporate both qualitative and quantitative metrics to measure, learn, and occasionally kill features that don’t measure up.
• #19: We released BoxCloud on a biweekly schedule and used feature requests as the primary means to drive the product. Unused features are a form of waste and Eric Ries emphasizes the importance of validated learning through fast build/measure/learn loops. In CloudFire, we use Continuous Deployment to release software almost daily and incorporate both qualitative and quantitative metrics to measure, learn, and occasionally kill features that don’t measure up.
• #20: We released BoxCloud on a biweekly schedule and used feature requests as the primary means to drive the product. Unused features are a form of waste and Eric Ries emphasizes the importance of validated learning through fast build/measure/learn loops. In CloudFire, we use Continuous Deployment to release software almost daily and incorporate both qualitative and quantitative metrics to measure, learn, and occasionally kill features that don’t measure up.
• #21: Anyone who has used google analytics knows it’s very easy to get lost in a sea of numbers. Dave McClure built a model using just 5 key metrics which I’ve implemented using home-grown and off-the-shelf tools.
• #22: Of those 5 metrics, not all of them need to be optimized day 1. Product/Market fit is the first thing that matters. But how do you measure it? Sean Ellis measures it using surveys to gauge initial user gratification. In Dave’s model, those translate to Activation and Retention. It is only after product/market fit that you scale up user acquisition.
• #23: Of those 5 metrics, not all of them need to be optimized day 1. Product/Market fit is the first thing that matters. But how do you measure it? Sean Ellis measures it using surveys to gauge initial user gratification. In Dave’s model, those translate to Activation and Retention. It is only after product/market fit that you scale up user acquisition.
• #24: Of those 5 metrics, not all of them need to be optimized day 1. Product/Market fit is the first thing that matters. But how do you measure it? Sean Ellis measures it using surveys to gauge initial user gratification. In Dave’s model, those translate to Activation and Retention. It is only after product/market fit that you scale up user acquisition.
• #25: The fundamental mind shift in applying lean startup is going from thinking you know something to testing everything you know.
• #26: The fundamental mind shift in applying lean startup is going from thinking you know something to testing everything you know.
• #27: The fundamental mind shift in applying lean startup is going from thinking you know something to testing everything you know.