Lean Analytics
M. Thiyagarajan, Sr PM QBO
A product manager’s key role is decision making
Analysis precedes decision making
Analysis
◉Ad hoc
◉Human Driven
◉Aggregated
Analytics
◉Programmatic
◉Machine Driven
◉Granular
Analytics - Making sense of raw data
Data led decision making, more objective
“What gets measured, gets managed”
-Peter Drucker
Black
Case of Good, Noisy & No Data
Traditional
Analytics
Lean Analytics None
NoisyGood No
H1, H2 & H3 Offering
H1
H2
H3
Source Geoffrey Moore – Horizon Planning of offering
Uncertainty
increases
Lean
Analytics
Brand New Idea
Lean Startup navigates the unknown
In startup goal of analytics - get to Product/Market
fit fast
Startup Big Business
Idea
Problem
/Solution Fit
User
/Experience Fit
Product
/Market Fit
Business Model
/Growth Fit
Uncertainty
Decreases
Risk
Decreases
Lean Analytics Cycle
Metric -> Hypothesis -> Experiment -> Act
Metric -> Hypothesis -> Experiment -> Act
◉Metric
Figure out what to improve
◉Hypothesis
◉ Make an educated guess
◉Experiment
◉ Test your guess
◉Act
Decide whats next
Metric depends on the stage
Product Market
Fit
Source: Lean Analytics by Ben Yoskowitz,
Alistair Coil
Metric depends on the stage
Product Market
Fit
Source: Lean Analytics by Ben Yoskowitz,
Alistair Coil
And Industry Too
Source: Lean Analytics by Ben Yoskowitz,
Alistair Coil
For Saas – Dave Mclure’s AARRR Pirate Metrics
A Saas Startup Data0
100
200
300
400
New Signups
0
0.5
1
1.5
2
2.5
Monthly churn Rate (%)
0.0%
10.0%
20.0%
30.0%
40.0%
50.0%
Mar/14
Apr/14
May/14
Jun/14
Jul/14
Aug/14
Sep/14
Oct/14
Nov/14
Dec/14
Jan/15
Feb/15
Mar/15
Apr/15
May/15
Jun/15
Free to Paid Conversion (Acq/Act)
0
10
20
30
40
50
60
70
80
Referrals
Numbers from a startup ->
Help Focus on right Metric
Succession Metrics
◉– used within Intuit
Used in Intuit
Empathy & Stickiness Relate to Value Hypothesis
Source: Lean Analytics by Ben Yoskowitz,
Alistair Coil
What is not Lean Analytics ?
◉A/B Testing, Conversion Rate Optimization
Data tells you whats happening, not why ?
◉ReferSpark, an example.
Great Google Ad CTR, Landing Page Signups
However attracted all wrong audience
◉ Operation success -> Patient Died
It is not about the numbers, its what is between them
When Ad CTR – 2.5% & Bounce Rate –> 75% - Good or Bad
OMTM - One metric is GOOD
◉Entire Org aligns towards it.
◉Critical in a startup team
Need two balancing metric
◉ In another Saas startup
Had acquisition rate 2x benchmarks
Churn also 2X of the average.
◉GNS vs Churn - tracked metric
User Generated Content Offering
Setup
Sharing
Jan 15 Mar 15Aug 15Apr 14
VOC of
Customer
Feature Built
Platform
Increase sample size
to 200
Jun 15
Share at ~1%
Increase targeted
sampling to 5,000
Different messaging
trigger
Behavior from 33 users
1 Share
Statistically Insignificant
Jan 14
Base Metric – 0.8%
Changed
Early Adopter User
Share at >1%
Absolute number still statistically
insignificant
Lean Metric will provide directions, not absolute or statistical
significance
Hano JourneyLine
Hano.in
Online Tools
Ease Payment Collection in SMM,
Easy Photo management
Track Orders , Online + Offline
Slow Education
- FTU challenges
- Inventory – seems key
Hano Mobile App
May15 Jun 15Apr15Nov 14
- Marketplace &
Community
- Repeat purchase online, not first
- Brand important , artist centric
- Between Seller, Buyer side. Focus on Seller
Photo management
not key
Narrow segment –
Jewellery
Mar 15
Top pain point seem
getting new business
Behavior slow
Broaden segment
Jewellery
Offline
Not Useful
ArtIlaka – Unit of One
Crafters – US, Great VOC
All Online Sellers
Voice Of Customer
Jan 14
Lean Analytics will help you from lying to yourself.
PreHano JourneyLine
BridgeVendor
Google Ads Test
Positive
Interview with
Customer, not a big
problem
Pivot to horizontal
purchase
SafeSME
5 Customer
Interviews
Problem not top 5
FlowPal
Surprises from customer interview
“Eliminate non core distraction to create
flow for startups”
Is non core problem (accounting, marketing,
legal right)
Is TAM big ?
Is NEP right solution ?
10 customer interviewed ( tech, non-
tech SMB)
Few have problem, not top 3 or 5
Hano
Nov13 Dec13Oct13Sep13Aug13
Customer Interviews
FMH (SoulSante,
Kitsch Mandi)
When doing Innovation Accounting through Lean
Analytics
◉Watch Out for what you are not measuring
( 2 weeks)
Overall Learning Goal
Goal/Objective Learning Goal/Objective Customers Spoken/Experiment Validated Learning Surprise Learnt
Summary
◉Lean Analytic is for when no or scant data
◉Metrics Function of stage and industry
◉Have one metric, two is good
◉Despite data, subjectivity in decision making, be
aware about bias
“
◉Thanks

Lean Analytics

  • 1.
  • 2.
    A product manager’skey role is decision making
  • 3.
  • 4.
  • 5.
    Analytics - Makingsense of raw data
  • 6.
    Data led decisionmaking, more objective
  • 7.
    “What gets measured,gets managed” -Peter Drucker
  • 8.
    Black Case of Good,Noisy & No Data Traditional Analytics Lean Analytics None NoisyGood No
  • 9.
    H1, H2 &H3 Offering H1 H2 H3 Source Geoffrey Moore – Horizon Planning of offering Uncertainty increases Lean Analytics Brand New Idea
  • 10.
  • 11.
    In startup goalof analytics - get to Product/Market fit fast
  • 12.
    Startup Big Business Idea Problem /SolutionFit User /Experience Fit Product /Market Fit Business Model /Growth Fit Uncertainty Decreases Risk Decreases
  • 13.
    Lean Analytics Cycle Metric-> Hypothesis -> Experiment -> Act
  • 14.
    Metric -> Hypothesis-> Experiment -> Act ◉Metric Figure out what to improve ◉Hypothesis ◉ Make an educated guess ◉Experiment ◉ Test your guess ◉Act Decide whats next
  • 15.
    Metric depends onthe stage Product Market Fit Source: Lean Analytics by Ben Yoskowitz, Alistair Coil
  • 16.
    Metric depends onthe stage Product Market Fit Source: Lean Analytics by Ben Yoskowitz, Alistair Coil
  • 17.
    And Industry Too Source:Lean Analytics by Ben Yoskowitz, Alistair Coil
  • 18.
    For Saas –Dave Mclure’s AARRR Pirate Metrics
  • 19.
    A Saas StartupData0 100 200 300 400 New Signups 0 0.5 1 1.5 2 2.5 Monthly churn Rate (%) 0.0% 10.0% 20.0% 30.0% 40.0% 50.0% Mar/14 Apr/14 May/14 Jun/14 Jul/14 Aug/14 Sep/14 Oct/14 Nov/14 Dec/14 Jan/15 Feb/15 Mar/15 Apr/15 May/15 Jun/15 Free to Paid Conversion (Acq/Act) 0 10 20 30 40 50 60 70 80 Referrals Numbers from a startup -> Help Focus on right Metric
  • 20.
    Succession Metrics ◉– usedwithin Intuit Used in Intuit
  • 21.
    Empathy & StickinessRelate to Value Hypothesis Source: Lean Analytics by Ben Yoskowitz, Alistair Coil
  • 22.
    What is notLean Analytics ? ◉A/B Testing, Conversion Rate Optimization
  • 23.
    Data tells youwhats happening, not why ? ◉ReferSpark, an example. Great Google Ad CTR, Landing Page Signups However attracted all wrong audience ◉ Operation success -> Patient Died
  • 24.
    It is notabout the numbers, its what is between them When Ad CTR – 2.5% & Bounce Rate –> 75% - Good or Bad
  • 25.
    OMTM - Onemetric is GOOD ◉Entire Org aligns towards it. ◉Critical in a startup team
  • 26.
    Need two balancingmetric ◉ In another Saas startup Had acquisition rate 2x benchmarks Churn also 2X of the average. ◉GNS vs Churn - tracked metric
  • 27.
    User Generated ContentOffering Setup Sharing Jan 15 Mar 15Aug 15Apr 14 VOC of Customer Feature Built Platform Increase sample size to 200 Jun 15 Share at ~1% Increase targeted sampling to 5,000 Different messaging trigger Behavior from 33 users 1 Share Statistically Insignificant Jan 14 Base Metric – 0.8% Changed Early Adopter User Share at >1% Absolute number still statistically insignificant Lean Metric will provide directions, not absolute or statistical significance
  • 28.
    Hano JourneyLine Hano.in Online Tools EasePayment Collection in SMM, Easy Photo management Track Orders , Online + Offline Slow Education - FTU challenges - Inventory – seems key Hano Mobile App May15 Jun 15Apr15Nov 14 - Marketplace & Community - Repeat purchase online, not first - Brand important , artist centric - Between Seller, Buyer side. Focus on Seller Photo management not key Narrow segment – Jewellery Mar 15 Top pain point seem getting new business Behavior slow Broaden segment Jewellery Offline Not Useful ArtIlaka – Unit of One Crafters – US, Great VOC All Online Sellers Voice Of Customer Jan 14 Lean Analytics will help you from lying to yourself.
  • 29.
    PreHano JourneyLine BridgeVendor Google AdsTest Positive Interview with Customer, not a big problem Pivot to horizontal purchase SafeSME 5 Customer Interviews Problem not top 5 FlowPal Surprises from customer interview “Eliminate non core distraction to create flow for startups” Is non core problem (accounting, marketing, legal right) Is TAM big ? Is NEP right solution ? 10 customer interviewed ( tech, non- tech SMB) Few have problem, not top 3 or 5 Hano Nov13 Dec13Oct13Sep13Aug13 Customer Interviews FMH (SoulSante, Kitsch Mandi)
  • 30.
    When doing InnovationAccounting through Lean Analytics ◉Watch Out for what you are not measuring ( 2 weeks) Overall Learning Goal Goal/Objective Learning Goal/Objective Customers Spoken/Experiment Validated Learning Surprise Learnt
  • 31.
    Summary ◉Lean Analytic isfor when no or scant data ◉Metrics Function of stage and industry ◉Have one metric, two is good ◉Despite data, subjectivity in decision making, be aware about bias
  • 32.

Editor's Notes

  • #3 http://images.htcampus.com/cmsmedia/uploads/ckeditor/decision-making-cartoon.jpg
  • #9 https://commons.wikimedia.org/wiki/File:Bullet-Train.jpg