In this presentation, Annabell Satterfield, Senior PM, Growth at BitTorrent, talks about how to engage Customers To Find Your Next Big Growth Opportunity.
Watch the full video at: www.Growthhackers.com
AI, Machine Learning, and their Application for Growth - #GHConf18GrowthHackers
Contenu connexe
Similaire à [GrowthHacker Conference '16] Annabell Satterfield Senior PM, Growth at BitTorrent: Engaging Customers To Find Your Next Big Growth Opportunity
Create a happy ending for data-driven decision-makingAman Sandhu
Similaire à [GrowthHacker Conference '16] Annabell Satterfield Senior PM, Growth at BitTorrent: Engaging Customers To Find Your Next Big Growth Opportunity (20)
7. Agenda
Processing data > insight with
Bucketing
Screening + Tradeoff decision modeling
Quantify utility
A/B test to maximize profit
In order to:
Prioritize next steps
Internationalize
Prioritize (paid) features
Choose prices
Getting from data > insights > action
Annabell Satterfield @als355
13. Work backwards.
1. Define the question
2. Define outcome
3. Is what I need out there? Go get it.
Data
Annabell Satterfield @als355
14. Work backwards.
1. Define the question
2. Define outcome
3. Is what I need out there? Go get it.
4. Process the data
Process
Annabell Satterfield @als355
19. Question
“What features will you build to drive paid app revenue?”
Among these, what will drive the most paid app revenue:
price testing, adding features, building awareness?
Annabell Satterfield @als355
31. Outcome
#1 Country 1- Language
#2 Country 2- Language
#3 Country 3- Language
--
#X Country X- Language
Highest-potential… based on our goals.
Annabell Satterfield @als355
32. Data
Does this data exist? Yes! Let’s go get it!
Annabell Satterfield @als355
34. Data
Choose your data based on those elements:
Market size: Installed Android user base (all apps),
Install base for the BitTorrent Desktop clients
Likelihood to monetize: Relative value of mobile adverts
Annabell Satterfield @als355
35. Process
Choose a weighting for each:
Market size: Installed Android user base [3],
install base for the BitTorrent desktop clients [4]
Likelihood to monetize: Value of adverts by geo [2]
Annabell Satterfield @als355
36. Step 1- Screen
Min Criteria 1
Min Criteria 3
Min Criteria 2
Meets
all 3
Annabell Satterfield @als355
37. Step 1- Screen
Mobile ad Value > $X
eCPM
Android User
Base >ZMM
BitTorrent
Desktop Users
>YMM
*Fake list
Annabell Satterfield @als355
38. Step 1- Screen
Android User
Base >ZMM
BitTorrent
Desktop Users
>YMM
*Fake list
Mobile ad Value > $X
eCPM
Annabell Satterfield @als355
39. Step 2- Prioritize
(Decile x
(1
..
(10
Android User Base
Weight)
3)
..
3)
Top
90%
..
10%
=
3
..
30
=
=
x
x
Annabell Satterfield @als355
40. Step 2- Prioritize
(Decile x
1
..
10
Android User Base
Weight)
3
..
3
+
Ad value
Top
90%
..
10%
+
Desktop Users
=
Annabell Satterfield @als355
41. Step 2- Prioritize
(Decile x
1
..
10
Android User Base
Weight)
3
..
3
+
Ad value
Top
90%
..
10%
+
Desktop Users
=
Value
68
..
90
Annabell Satterfield @als355
42. And… do your due diligence
Think about re-adding:
Brazil
Russia
India
China
...others?
Ordered by value…*
#1 United States- English: 90
#2 Canada- English: 89
#3 Spain- Spain Spanish: 70
#4 S. Korea- Korean: 68
…
*Fake list
Annabell Satterfield @als355
43. Spain +93% w/w
Brazil +43%
Italy +32%
Russia +23%
Installs in high-potential geos targeted for growth with
translation:
Annabell Satterfield @als355
44. Case 3- [Paid] Feature Prioritization
Data: Survey Process: Quantifying Utility (+ ROI analysis )
Annabell Satterfield @als355
45. Question
Will more users be willing to buy Pro if we released more features?
Which ones?
Annabell Satterfield @als355
46. Outcome
Feature #1- Utility value of X
Feature #2- Utility value of Y
--
Feature #N- Utility value of Z
Annabell Satterfield @als355
47. Outcome
Feature #1- Utility value of X
Feature #2- Utility value of Y
--
Feature #N- Utility value of Z
Impact of new feature on likelihood to buy
Audience:
Core users, top geos
for paid product
Pay for upgrades
Annabell Satterfield @als355
48. Data
Does this data exist? No. Let’s go make it!
Annabell Satterfield @als355
54. Calculated weight of each Pro feature:
Feature X: 5.78 Ranked #1
Feature Y: 5.59 Ranked #2
Feature Z: 5.20 Ranked #3
10+ features considered and prioritized with data… priceless
Annabell Satterfield @als355
55. Outcome…
Battery-Saver: +47% Daily paid Revenue
Auto-shutdown: +20% Daily paid Revenue
uTorrent Android- top 5 grossing (for paid
upgrades] in its category in 71 geos
including US, UK, CA, AU*
Happy paying users
* App Annie, April 2015 (before moved to desktop team)
Annabell Satterfield @als355
66. These are not substitutes for experiments or MVPs.
They inform experiments.
Annabell Satterfield @als355
67. Don’t let anything keep you from being data-driven where it counts.
Annabell Satterfield @als355
68. Qualitative data can fill in the gaps where other data ends.
Annabell Satterfield @als355
69. Processing data > insight with
Bucketing
Screening + Tradeoff decision modeling
Quantify utility
A/B test to maximize profit
In order to:
Prioritize next steps
Internationalize
Prioritize (paid) features
Choose prices
These tools are multifunctional
Annabell Satterfield @als355
70. Processing data > insight with
Bucketing
Screening + Tradeoff decision modeling
Quantify utility
A/B test to maximize profit
You can also:
Isolate problems with feedback
Prioritize B2B features w/survey
Develop keep/kill feature list
Pretotype unbundled paid
products
More…
These tools are multifunctional
Annabell Satterfield @als355
72. These are not substitutes for experiments or MVPs.
They inform experiments.
Annabell Satterfield @als355
73. Don’t let anything keep you from being data driven where it counts.
Annabell Satterfield @als355
74. Qualitative data can fill in the gaps where other data ends.
Annabell Satterfield @als355
75. Processing data > insight with
Bucketing
Screening + Tradeoff decision modeling
Quantify utility
A/B test to maximize profit
In order to:
Prioritize next steps
Internationalize
Prioritize (paid) features
Choose prices
These tools are multifunctional
Annabell Satterfield @als355
76. Processing data > insight with
Bucketing
Screening + Tradeoff decision modeling
Quantify utility
A/B test to maximize profit
You can also:
Isolate problems with feedback
Prioritize B2B features w/survey
Develop keep/kill feature list
Pretotype unbundled paid
products
More…
These tools are multifunctional
Annabell Satterfield @als355
82. Annabell Satterfield @als355
Bingo
Talking to users
Open-ended surveys
OPR (OP Research)
Product data
Databases
Product data
Surveys*
Experiments
OPR
Talking to users
Databases
Experiments
Pretotypes
Surveys*
Databases
*term used loosely
Choose data based on what you understand about your problem…