the lean startup - eric ries 5 principles 1) entrepreneurs are everywhere 2) entrepreneurship is management a kind for high uncertainty 3) validated learning startups exist to learn how to build a business 4) build-measure-learn loop fundamental activity 5) innovation accounting how to measure progress, set milestones, prioritize 3 main sections vision define entrepreneur, startup, way to gauge progress steer lean method in detail, b-m-l loop, mvp tests, when to pivot accelerate how to speed up b-m-l loop, lean manuf. concepts, org design, prod growth .suggests silod departments think they are more productive because of local measurement vision, strategy, product pyramid strat is the pivot level, product is the optimization level shift decisions from projection and politics to cheap exploration (because the customer or feature set is unknown) *validated learning rigorous method for demonstrating progress in uncertain conditions empirically discover truths about business prospects value creation vs wasting effort (risk is a liability) value hypothesis test whether customers actually want it growth hypothesis how they will discover it vapor, video, manual mvps ideation period collection of experiment products to test solutions *innovation accounting in the measure part of the b-m-l loop quantitative approach, set learning milestones and actionable metrics 1 rigorously measure current state 2 devise experiments to match reality to business plan three learning milestones 1 use an mvp to get data about now, test critical assumptions aim to derisk the point of mvps is to match early adopters 2 tune toward ideal and get to a decision point hypothesize and test directions to improve 3 pivot or continue reconsider viability use cohorts and split tests user stories should not be considered complete until they lead to learning 4 states: backlog, being built, tech done, validation kanban board: limit how many stories can be in each state (eg 3) encourages culture oriented around testing 3 As of metrics actionable clear cause and effect accessible use tangible units auditable data needs to be defensible, crosschecked eg with users *pivots shifts or refocuses in fundamental ideas about strategy or growth designed to test new fundamental hypothesis new mvp real measure of runway is how many pivots are still possible consider pivots when experiments are declining effectiveness or dev feeling unproductive make decision by looking at feedback, reports vs expectation pursue alternatives so there's prepared things to pivot to even successful companies need to pivot when customer base expands early adopters --> mainstream requires shift in personas kinds of pivots zoom in on feature, zoom out to broader product, customer segment, customer need, platform (app <--> plat), business arch (high margin low volume <--> low marg high vol), value capture, growth engine, channel, tech *small batches sometimes streamlining attempts ignore side effects that actually make things slower (eg letters one at a time vs in bulk steps) postpones feedback andon cord - letting anyone interrupt the process analogous: testing suites and metric-driven auto rollbacks "continuous deployment" advantages of large batches are sometimes too local doesn't make sense in uncertain conditions or when returns are likely accountability -> process -> culture -> people 4 primary types of sustainable growth word of mouth side effect of product usage funded advertising repeat use 3 growth engines sticky, retention based viral paid long term paid growth requires unique ability to monetize because of CAC competition (eg imvu allowed mailed cash from teens without credit) growth engine determines product/market fit eventually growth hits a wall because of size of market so there will need to be pivots or simultaneous new products do root cause analysis on everything include people natural speed regulator, slows you down when you go too fast or commit too many errors might need to start small to transition into this practice or have it triggered in certain situations so it doesn't get out of hand can mix in 'proportional investment' to the steps of 5 whys thing eg 'why didnt this work, because this, we can do this at that level' 'but why didn't that work' two outputs of cause analysis learning and doing *ways to nurture disruptive innovation scarce but secure resources independent development authority (suggests crossfunctional) personal stake in outcome *making a platform for experimentation shield the parent/child organizations from each other rational fears (other business depts worry about the competition) danger of hiding innovation from other parts of company *innovation sandbox any team can create a split test on their sandbox (eg pricing page) one team must see it from end to end timeboxed customer count limited evaluated by clear, actionable metrics those metrics are uniform for that sandbox's products the team is responsible for monitoring and aborting clear leader *entrepreneur as a job title creative managers have different skillsets from optimization managers these methods can/should be applied to internal company structure