PR
acqu
isiti
on
App Store Optimization
Content Marketing
Performance Marketing
DistributionDeals
(pre-installs, traffic
exchange)
Viral loops (Invites & content sharing)
Inte
rnatio
nal
First Time User Experience
(onboarding, tutorial)
User Accounts(Data persistence, sign-in services)
Deep LinkingLifecycle Marketing
Activity Notifications
Community (Engagement
& Support)
Payment Processing
(Carrier Billing, PayPal, Offer walls,
Credit Card, etc.)
Pricing (Fixed, Dynamic,
Regional, Virtual currency)
Merchandising
Ad Inventory Management(Native Ads,
Sponsorship, Direct Sales, Ad Exchange)
Re
targe
ting
Partn
ersh
ips
Cross-sellContent Indexing
(Mobile SEO)
Deeplink Attribution
Event Tracking
Campaign Measurement
A/B testingTouch
HeatmapsScreen Flows
Conversion Optimization
Push notifications
In-App Messaging
SMS
Search
Social
Mobile Ad Networks
TV, Print, Radio
Owned Channels
App Store Listing
App Store Analytics
ASO Keyword Performance
Tracking
User Segmentation
Conversion Funnels
Billing & Revenue
Reporting
Performance Analysis
(CPU, Battery, Network, Bug
Tracking)
Cohort Analysis
Content Analytics
Sentiment Tracking
LTV Modelling
Growth Accounting(Growth rate,
Churn, Sessions)
Growth Modelling
(Opportunity identification)
Install Attribution
User Testing
Channelsen
gage
men
t&
rete
ntio
nan
alyt
ics
& in
sig
hts
Revenue Model(Freemium,
Paid, Ad-Supported, Subscriptions, Virtual
Goods)mon
etiz
atio
n
TargetSummit Berlin@andy_carvell
Mobile Growth Process
www.mobilegrowthstack.comwww.phiture.com
● Growth Modeling● Experimental process● Hypothesis definition● Using the Mobile Growth Stack to identify opportunities● Running a growth meeting● Growth team setups● Examples of real growth teams: how SoundCloud did it
Growth Process
Everything comes back to the Growth Model
● Start with nailing the qualitative model● Figure out (and agree internally on) your ‘growth story’● Next: Quantitative spreadsheet to model these steps.● Refine model over time as user understanding deepens through
experimentation, user testing, interviews and questionnaires.
Src: http://abovethecrowd.com/2014/07/11/how-to-miss-by-a-mile-an-alternative-look-at-ubers-potential-market-size/
Uber’s Growth model
SRC: http://www.mobilegrowthstack.com/analytics-insights/growth-modeling/mobile-growth-model/ (William Gill)
Build a Basic Growth ModelBuild a quantitative, input-driven model that captures the drivers of acquisition and retention as you understand them.
Start simple. All models are abstractions; it will never be perfect.
Use your growth model to do scenario planning:
See how numbers develop if you focus on acquisition and increase installs by 10%.See if you would get more active users in 3 months by making retention improvements affecting new users, etc.Identify the biggest opportunities.
Identify Opportunities and Build a Growth Backlog
What’s the current process for defining & refining hypotheses & predictions?
Is there an experiment backlog? How do new ideas make it into the backlog?
What are the currently active experiments? What are they teaching you about your users?
Audit your growth efforts to identify areas for development.
PR
acqu
isiti
on
App Store Optimization
Content Marketing
Performance Marketing
DistributionDeals
(pre-installs, traffic
exchange)
Viral loops (Invites & content sharing)
Inte
rnatio
nal
First Time User Experience
(onboarding, tutorial)
User Accounts(Data persistence, sign-in services)
Deep LinkingLifecycle Marketing
Activity Notifications
Community (Engagement
& Support)
Revenue Model(Freemium,
Paid, Ad-Supported, Subscriptions, Virtual Goods)
Payment Processing
(Carrier Billing, PayPal, Offer
walls, Credit Card, etc.)
Pricing (Fixed, Dynamic,
Regional, Virtual currency)
Merchandising
Ad Inventory Management(Native Ads,
Sponsorship, Direct Sales, Ad Exchange)
Re
targe
ting
Partn
ersh
ips
Cross-sellContent Indexing
(Mobile SEO)
Deeplink Attribution
Event Tracking
Campaign Measurement
A/B testingTouch
HeatmapsScreen Flows
Conversion Optimization
Push notifications
In-App Messaging
SMS
Search
Social
Mobile Ad Networks
TV, Print, Radio
Owned Channels
App Store Listing
App Store Analytics
ASO Keyword Performance
Tracking
User Segmentation
Conversion Funnels
Billing & Revenue
Reporting
Performance Analysis
(CPU, Battery, Network, Bug
Tracking)
Cohort Analysis
Content Analytics
Sentiment Tracking
LTV Modelling
Growth Accounting(Growth rate,
Churn, Sessions)
Growth Modelling
(Opportunity identification)
Install Attribution
User Testing
The Mobile Growth Stack 3.0 — A strategic framework for mobile growthBy @andy_carvell with contributions from @moritzdaan
enga
gem
ent
& re
tent
ion
anal
ytic
s &
insi
gh
tsm
onet
izat
ion
Channels
Growth Backlog Best Practice
● Everyone encouraged to contribute ideas● Backlog groomed by PM + reviewed with growth team● Every experiment driven by a hypothesis
○ Result should lead to better understanding of user psychology & possibly refine the growth model, even if there is no positive impact on metrics.
● Prioritised by Projected Impact, Estimated Effort, Level of Confidence (i.e. do you think it will work?)○ Use scenario modelling to work out which initiatives have the most upside
for overall growth● Aim to iterate fast… quick wins and small tests can be good for gaining
momentum in the process
Experiment Metric(s) Impact Confidence Ease Score
Add focus on security in onboarding flow to accommodate German-specific concerns
D-1 Retention 4 8 5 =AVG()
Frame App Store screenshots in phone to boost conversion Downloads 3 5 10 =AVG()
Partner up with a German Telco with low data offerings (ie. AldiTalk)
MAU 10 7 3 =AVG()
From Stack to Ideas
Considering REACH
● Consider Reach through the lens of retention curve… if you’re targeting only engaged, retained users, the group (at least in cohort terms) is very small compared to a feature / action that reaches NEW users before they churn.
● Better to ask new users to do things while they are still there, even if conversion would be better with highly engaged users (since the higher volume x lower conversion rate still = more impact). NB: Not in every case, but be sure to do the math.
● Fewer than 25% of A/B tests produce a meaningful upside. Often useful to test extreme ideas - as long as the upside case is there.
How to Run a Weekly Growth Meeting
Brian Balfour (HubSpot) on running a growth meeting http://www.coelevate.com/essays/growth-meeting
● Celebrate IMPACT over everything else● Keep weekly meeting tight● Maintain a backlog of ideas that anyone can
contribute to (company wide)○ PM / growth manager grooms the
backlog● Encourage environment of experimentation
(failures are important)● Aim to increase the speed that experiments
(and hence failures) that can be executed: faster iteration.
● Kill initiatives that don’t work quickly○ SCALE those that do work
Possible growth team structures
Src: https://medium.com/swlh/how-do-you-choose-the-best-growth-team-model-632ad5a85be9#.bjwodayow
Newly formed growth team in April 2015, replacing earlier attempt. (Restructuring again since Sep 2016)Total team size: 4 PMs, 8 Eng, 1 Designer, 2x Data Scientists, 1x Growth Marketer
Acquisition handles inbound referrals, SEO, ViralRetention handles CRM (inc. activity notifications), Re-Activation, OnboardingSign-In is a small team focused on Sign-In… likely to expand to pick up Activation
Some Email marketing and ASO support from Marketing Team.
Partnership BD is done by a separate team out of SF.
Sub-Teams
● Acquisition● Sign-In● Retention (inc. reactivation)● International