iterate. calculate. communicate

37
ITERATE. CALCULATE. COMMUNICATE.

Upload: laure-parsons

Post on 27-Jan-2017

108 views

Category:

Data & Analytics


0 download

TRANSCRIPT

Page 1: Iterate. Calculate. Communicate

ITERATE. CALCULATE. COMMUNICATE.

Page 2: Iterate. Calculate. Communicate

CLOSING THE GAP BETWEEN MEASURE + LEARN

Page 3: Iterate. Calculate. Communicate

LinkedIn, Medium & Twitter: @laurex

LAURE PARSONS SENIOR PRODUCT MANAGER

Page 4: Iterate. Calculate. Communicate

“THE GOAL OF BUILD-MEASURE-LEARN IS NOT TO BUILD A FINAL PRODUCT TO SHIP OR EVEN TO BUILD A PROTOTYPE OF A PRODUCT, BUT TO MAXIMIZE LEARNING THROUGH INCREMENTAL AND ITERATIVE ENGINEERING”

Steve Blank, author of The Four Steps to the Epiphany

BE ACTIONABLE!

Page 5: Iterate. Calculate. Communicate

1. WHAT SHOULD YOU TRACK? 2. WHAT MAKES A METRIC “GOOD?”

3. HOW CAN YOU GET ALIGNMENT WITH YOUR TEAM AND LEADERSHIP?

4. WHAT SHOULD YOU DO TO ENSURE YOU LEARN?

‣ CASE STUDY: BUILDING A NEW FEATURE!

Page 6: Iterate. Calculate. Communicate

ACQUISITION ACTIVATION REFERRAL RETENTION REVENUE

AARRR: PIRATE METRICS

Page 7: Iterate. Calculate. Communicate

ACQUISITION ACTIVATION REFERRAL RETENTION REVENUE TEAM

AARRRT: TEAM METRICS

Page 8: Iterate. Calculate. Communicate

THE ONE METRIC

From Lean Analytics by Benjamin Yoskovitz & Alistair Croll

Page 9: Iterate. Calculate. Communicate

UX TESTS MVP A/B TESTSCLASSIC LEAN MEASUREMENT

Page 10: Iterate. Calculate. Communicate

WHO, WHAT, WHEN, WHERE, WHY & HOW

▸WHO CUSTOMERS ARE

▸WHAT CUSTOMERS DO AND/OR TRY TO DO

▸WHEN DO USERS DISENGAGE

▸WHERE CUSTOMERS COME FROM

▸WHY CUSTOMERS CONVERT

▸HOW MUCH YOU CAN BUILD

Page 11: Iterate. Calculate. Communicate

GOOD METRICS

Page 12: Iterate. Calculate. Communicate

GOOD METRICS

COMPARABLE

Page 13: Iterate. Calculate. Communicate

RATIOS & RATES

Page 14: Iterate. Calculate. Communicate

GOOD METRICS

UNDERSTANDABLE

Page 15: Iterate. Calculate. Communicate

GOOD METRICS

ACTIONABLE

Page 16: Iterate. Calculate. Communicate

BE ACTIONABLE!

TWO TYPES OF ACTIONABLE DATA▸“Accounting” metrics can make your

predictions more accurate

▸“Experimental” metrics help optimize product, pricing, or market. Agree on what will change before you collect the data.

Page 17: Iterate. Calculate. Communicate

“IF YOU MEASURE SOMETHING AND IT’S NOT ATTACHED TO A GOAL, IN TURN CHANGING YOUR BEHAVIOUR, YOU’RE WASTING YOUR TIME.”

LEAN ANALYTICS by Ben Horowitz and Alistair Kroll

BE ACTIONABLE!

Page 18: Iterate. Calculate. Communicate

ALIGNMENT & COMMUNICATION

Page 19: Iterate. Calculate. Communicate

OKRS

Page 20: Iterate. Calculate. Communicate

WHO OWNS THE DATA?

Page 21: Iterate. Calculate. Communicate

WHAT ACTIONS WILL YOU TAKE?

Page 22: Iterate. Calculate. Communicate

WHO NEEDS TO KNOW?

Page 23: Iterate. Calculate. Communicate

HOW WILL THEY KNOW?

Page 24: Iterate. Calculate. Communicate

THE DATA TOOL YOUR TEAM WILL ACTUALLY USE

Page 25: Iterate. Calculate. Communicate

‣A single source of truth

‣Company-wide common understanding and definitions

‣Appropriate permissions

‣Transparency

‣Collaboration via shared data and commenting

‣Automated integration from your tools

‣Sharing in Slack, where your team is already talking

‣Dashboards to track the important things for each team

‣One-click reports to share with management and stakeholders

Page 26: Iterate. Calculate. Communicate

CASE STUDY

Page 27: Iterate. Calculate. Communicate

CASE STUDY QUESTIONS

DISCOVERY & LEAN CUSTOMER DEVELOPMENT▸ Based on our problem

hypothesis, what percentage of the 6-10 interviews we conduct confirm what we think is the problem?

Page 28: Iterate. Calculate. Communicate

CASE STUDY QUESTIONS

MVP

▸ Solves a problem in the simplest way possible

▸ Could actually be better than the intended product, with “concierge onboarding"

▸ Asks a question, which should be clear and measurable.

Page 29: Iterate. Calculate. Communicate

CASE STUDY QUESTIONS

EFFORT TO BUILD▸ How much dev effort is

anticipated in points or cost per developer hour?

▸ Reduce scope to the most essential parts of the feature. Build “nice to haves” when you have signal

▸ Track velocity, throughput, cycle times to anticipate & understand building costs

Page 30: Iterate. Calculate. Communicate

EFFORT TO BUILD

Page 31: Iterate. Calculate. Communicate

CASE STUDY QUESTIONS

WHO CARES?▸ Beta test the feature with customers

who’ve requested it and/or with freemium or opt-in customers

▸ If you have a large enough customer base, survey your customers with a tool like Intercom or Qualaroo when they use the feature

▸ Measure feature adoption in Mixpanel or other event-based tools. See how often customers try the feature and how many times they re-use it.

Page 32: Iterate. Calculate. Communicate

CASE STUDY QUESTIONS

COSTS BEYOND DEVELOPMENT

▸ Track support requests tagged with the feature name

▸ Track marketing spend on launching the feature

▸ Track escaped bugs

Page 33: Iterate. Calculate. Communicate

CASE STUDY QUESTIONS

THE BOTTOM LINE

▸ Do a cohort analysis of customers before and after the feature was introduced.

▸ Track number of signups, churn, and referrals.

▸ If the feature is plan-gated, you can also track by pricing segments.

Page 34: Iterate. Calculate. Communicate

WHAT ACTION FOLLOWS?

Page 35: Iterate. Calculate. Communicate

COMMUNICATION CHECKLIST

ARE YOU COMMUNICATING?▸ Does your whole team and company know why

you are tracking these metrics?

▸ Do you have agreed-upon actions based on the results?

▸ Do you have to track down people to get the data?

▸ Does someone own the metric?

Page 36: Iterate. Calculate. Communicate

DON’T STOP NOW!

LEARN MORE

▸Lean Analytics by Benjamin Yoskovitz & Alistair Croll

▸Getting Started with OKRs & Metrics Guide

▸The School of Little Data

▸The AARRRT of Product Metrics

Page 37: Iterate. Calculate. Communicate

THANK YOU!