2. From vision to MVP The goal of customer development is to find a market for the product as currently specified Have a strong vision, be prepared to learn whether it makes sense Feedback from customers tells you about them not you No focus groups
3. Major Sources of Waste Building something nobody wants Adding features your current customer doesn’t want AKA adding features that don’t validate/refute current hypothesis Arguing about product priorities Flip-flopping between plans (iterating in a circle) (these are in order of costliness)
4. Problem team, Solution team Problem team: working on customer discovery Solution team: working on Minimum Viable Product (MVP) Both teams are cross-functional Commitment from solution team leader to be personally involved with customer discovery Metrics are people, too
5. Today’s Agenda Develop deep customer insight: problem presentation, “day in the life” Translate that insight into actionable per-user metrics Build a model of how those metrics lead to massive success Establish a baseline measurement using a minimum viable product … and then move to Customer Validation
6. Deep Customer Insight (Problem) Get out of the building and meet real and potential customers Figure out what problems they have Where does your problem rank on the hierarchy of pain? Figure out the how, what, where, and why of customers using a product like yours How do they describe the category, problem, and competitors – learn their language Learn to tell an early adopter from a mainstream customer
7. Deep Customer Insight (Solution) Learn how to describe your solution to potential customers Find out if they agree it solves the problem, assuming it works “by magic” Discover barriers to adoption (would they start using a magic product right away?) Offer to pre-order to discover barriers to actual purchase (and to qualify early adopters)
8. Customer Archetype Succinct description of insights Designed to be actionable for whole team If more than one, pick one target Rule: always build for the target archetype without humiliating any other archetype Big savings: avoid building features outside archetype description
9. Actionable Metrics Assume everything you learned in discovery is true – how would you know? Use customer insight to plot out the specific funnel for your customers: how they find out about your product how they acquire/try/adopt it how they pay for it how they engage over time Establish per-customer metrics for this funnel Most important: How do you know you’re making the product better?
10. Gross metrics don’t work Why not focus on gross revenue, profit, or growth rate? Impossible to predict Keeps team working on high-ROI activities, but innovation tends to be low-ROI (at first) Focus on gross numbers tends to erode differentiation (as everyone does the same “obvious” stuff)
11. Engine of Growth The goal of actionable metrics is to establish a working and growing business model Need to understand the “ecosystem” of your business at the per-customer level, and make sure it’s value-creating Marginal revenue > marginal cost High volume, low margin Low volume, high margin Need to understand how this ecosystem supports one of three drivers of growth: Paid (CPA < LTV) Viral (Viral coefficient > 1) Sticky (customer retention extremely high)
12. Build the model Create the usual spreadsheet model of your business, but Focus on inputs instead of outputs Come up with reasonable assumptions, and make sure that the outputs are sufficient Recognize that the model will probably change, so relationships are more important than specific numbers
13. Establish a baseline Minimum viable product: that version of a new product which allows a team to collect the maximum amount of validated learning about customers with the least effort. Each input in the model represents one key hypothesis about the business Use the MVP to measure each input. Eliminate any features that do not pertain to one of the key inputs. Work on the riskiest hypotheses first
14. Baseline Once you have baseline numbers for your business, you are ready for customer validation Probably, the numbers will look terrible – that’s OK Figure out what the deltas are between baseline and a good outcome Figure out which numbers are movable and which are fixed
15. Customer Validation Once you have a MVP, become more dynamic Shift from one-time activities to continuous flow, measured by validated learning As you learn, you will be able to influence the actual customer behavior in your model
16. Validated Learning It’s as important to know why a metric changed as to be able to show change Growth in gross metrics is always ambiguous, too many external factors Key validation techniques: Revenue per customer Cohort analysis Split-testing
17. BOD accountability Use validated learning to demonstrate shared sense of progress among: Founders Board of directors Investors/outside stakeholders Each baseline step is progress After baseline, each pivot is progress
18. Team accountability Charter semi-autonomous cross-functional teams, starting with just one solution team Select a mutually-agreed goal Team agrees to hit the goal or die trying Team has representatives from all functions Owns product, marketing, deployment decisions At the end of a cycle, team can achieve success by: Hitting the actionable-metric target Demonstrating deep learning about what went wrong Over multiple cycles, must show this learning is improving chances of hitting targets
19. Pivot When customer validation fails, it’s time to pivot Most pivots originate in the solution team: they cannot find a way to make the current hypothesis work. Can’t hit actionable targets Don’t improve on those targets over time Each team must bring key data to a pivot meeting: Solution team must have data about what’s not working Problem team must have evidence for a next hypothesis Both teams must have spent time with current customers Pivot: keep most of the business model the same, change one key part at a time
20. Types of pivots Customer need pivot: same customer segment, different need/problem Customer segment pivot: same problem, different segment Business architecture pivot: ie from enterprise to consumer Zoom-in feature pivot: remove features to focus on just one key feature Zoom-out feature pivot: add features to become more of a holistic solution Technology pivot: solve same problem but with different technology stack Channel pivot: same problem, same solution, different path to customers Platform pivot: open up an application to third-parties to become a platform (or vice-versa)
21. Today Develop deep customer insight: problem presentation, “day in the life” Translate that insight into actionable per-user metrics Build a model of how those metrics lead to massive success Establish a baseline measurement using a minimum viable product First up: KISSmetrics