Key Learnings:
-Understand what your users are doing on your app once you’ve acquired them
-Use the Recency, Frequency, and Monetary (RFM) analysis framework for segmentation to convert loyal customers into influencers
-Apply cohort analysis, and intent-based segmentation to fuel growth
3. You’ll learn about…
Using the Recency,
Frequency, and
Monetary (RFM) analysis
framework to convert
loyal customers into
influencers
Applying cohort
analysis, and intent-
based segmentation to
fuel growth
Leveraging analytics to
understand what your
users are doing on
your app once you’ve
acquired them
4. Our Speakers
Backed by an engineering degree from IIT Bombay,
Shobhit spent his 20s in three entrepreneurial ventures
and finally took up his first job at the age of 30. Shobhit is
presently responsible for growing Disney+ Hotstar –
India’s largest OTT platform and the future of
entertainment.
https://www.linkedin.com/in/shobhitsinghal1/
SABYASACHI SAHU
Customer Success Manager
Sabyasachi is a senior consultant with 5+ years of
experience that spans multiple functions in product.
At CleverTap, he’s been responsible for helping
clients across India and Europe, from early stage
startups to multi-billion dollar enterprises for driving
growth.
https://www.linkedin.com/in/sabyasachisahu/
SHOBHIT SINGHAL
Associate Director, Product
6. What does Growth mean at D+ Hotstar
How do we ascertain which Metrics to grow
How do we set up Analytics
How do we build Segmentation frameworks &
Cohorts
How does analytics translate into User
Intelligence
How does insight inform Personalisation
How does personalisation create Growth Loops
Here’s the Line Up
7. What does Growth mean at D+ Hotstar
How do we ascertain which Metrics to grow
How do we set up Analytics
How do we build Segmentation frameworks &
Cohorts
How does analytics translate into User
Intelligence
How does insight inform Personalisation
How does personalisation create Growth Loops
Getting Started
8. Grow What Exactly?
I want to maximise time spent
- Time spent is a metric that many
consumer internet apps relying on ad-
led revenue care about
I want to maximise revenue
- Revenue could be a metric mature
companies post Product Market Fit
may care about
10. What does Growth mean at D+ Hotstar
How do we ascertain which Metrics to grow
How do we set up Analytics
How do we build Segmentation frameworks &
Cohorts
How does analytics translate into User
Intelligence
How does insight inform Personalisation
How does personalisation create Growth Loops
Sounds Simple, But...
11. Keep It Simple
Simple Equations >> Complex Equations
Define a North Star
Extract L1 and L2 metrics from the North Star
12. Measure of Success = Subscriber Growth
We want users to become paying customers because higher LTV + more loyalty
Subscriptions North Star → L1s
Acquisitions ChurnSubs Growth
13. Acquisitions = Convert free users + Acquire users via marketing + Organic
➔ Convert free users → Show relevant paid catalogue to free users
➔ Marketing → Optimise each step of the performance marketing funnel
➔ Organic → Easy and frictionless journey to subscribe
Churns = Users who cancel + Non-renewers on non-recurring plans
➔ Users who cancel → build more engagement with Hotstar
➔ Non-renewers on non-recurring → convert both negatives to positives
Subs L2
14. Measure of Success = Watch Time
As an OTT platform, the amount of time spent by users watching videos is a strong indicator of the health of our platform
Engagement North Star → L1s
Acquisition
No. of Users
Frequency of
visit
Time Spent
per visit
Retention Engagement
Watch Time
15. No. of Users = MAU = Acquisition
Frequency = Active Days per Week X Active Weeks per Month
Time Spent per Visit = (1/ bounced) X Time spent engaging
Engagement L2
16. What does Growth mean at D+ Hotstar
How do we ascertain which Metrics to grow
How do we set up Analytics
How do we build Segmentation frameworks &
Cohorts
How does analytics translate into User
Intelligence
How does insight inform Personalisation
How does personalisation create Growth Loops
Getting the Basics Right
17. Early Stage
You probably have 1-2 apps
Talk to peers/leaders about metrics
Talk to industry experts. Work with
Clevertap CS teams (paid plug 😅)
Focus instrumentation on key insights
You support 95% of the platforms you
want to be on; for us that’s 12 apps
Source-of-truth Data Dictionary is
mandatory
Start assigning a risk value to events
and properties. Talk to peers
Eg: Shobhit will lose his job if the
“Watched Video” event breaks
Set up alerts on event volumes and
values
Growth Stage Late Stage
Everything you are doing for data
quality is a year late
Remove dependency on client apps
and release cycles
Automate schema quality; social
development practices
Invest in anomaly detection based on
how many PMs have lost their jobs -
Talk to people :-)
Talk to People
19. What does Growth mean at D+ Hotstar
How do we ascertain which Metrics to grow
How do we set up Analytics
How do we build Segmentation frameworks &
Cohorts
How does analytics translate into User
Intelligence
How does insight inform Personalisation
How does personalisation create Growth Loops
Making Cohorts Cool Again
20. Worthy Segments
Hotstar 2018 - Business Insight
MAU
● IPL would drastically move up the
baseline
● Many users would churn out when
IPL was over
● New baseline of active users =
step function growth
● Users would stay back for non-
cricket
Worthy Segment:
Users who first watched cricket && then
watched non-cricket content (C2E)
21. 2020: No IPL, Now What?
Hotstar 2020 - Business Insight
“Avengers: Endgame” Theatre Performance
● There is an equally sized market for
content dubbed in Indian languages
● Launched AE on Hotstar in Nov 2019 in
English, Hindi, Tamil and Telugu
● Offered lower price points for HTT
● Launched Disney+ in Indian languages
Worthy Segments:
1. Users who watch entertainment content in
Indian languages
2. Users paying for higher and lower priced
plans and differences in language consumption
2x tickets sold for AE in
Indic Languages
Indic rev = English rev
25. What Can We do With Segments?
Re-activation
Facebook / DBM
Spend some money
for win-back
Analytics
BI Tools
Understand
conversion metrics
for each segment
Notifications
Clevertap
Push / nudge users
likely to perform
desirable actions
We built a platform to allow user segmentation and refreshing cohorts at scale. We deploy
connectors to send “lists” of users for various applications:
26. What does Growth mean at D+ Hotstar
How do we ascertain which Metrics to grow
How do we set up Analytics
How do we build Segmentation frameworks &
Cohorts
How does analytics translate into User
Intelligence
How does insight inform Personalisation
How does personalisation create Growth Loops
Get Smart, Quickly
27. Let’s Focus on More Subs
We wanted to identify which users can convert into paying customers
Massive free user base
● We created paid value propositions
out of our content catalogue
● Latest movies, Specials
● Disney+ :-)
Worthy Segments:
1. Users who sample, click on paid content
2. Users who watch on more than 1 device
Overall Entertainment
Value Proposition for
Indian Households
Multiple Hooks to
Subscribe
28. Some Parameters Stood Out
Device Cost
Metro Cities
Watched Cricket
&& Entertainment
WiFi vs 4G
Streaming Quality
Age
Number of Devices
29. Birthing a Model - Propensity to Pay
Predicted
Scores
DAU
Subs
Acquisitions
< 0.1 9% 0.5%
< 0.2 9% 0.3%
< 0.3 9% 0.5%
< 0.4 9% 1%
< 0.5 9% 1%
< 0.6 9% 2%
< 0.7 9% 3%
< 0.8 9% 5.5%
< 0.9 10% 12%
>0.9 20% 75%
● Break the variables down further
Example: what are the fundamental metrics that
influence device cost? Dig deeper
● Remember Branal:
Which metrics have high coverage and high accuracy
● Test early and frequently
● Most Customer Intelligence platforms help get started
quickly - Clevertap, Amplitude, Mixpanel
30. What does Growth mean at D+ Hotstar
How do we ascertain which Metrics to grow
How do we set up Analytics
How do we build Segmentation frameworks &
Cohorts
How does analytics translate into User
Intelligence
How does insight inform Personalisation
How does personalisation create Growth Loops
Pursuing the Holy Grail
31. Key Elements of Personalisation
Recommendation
Best guess, Increases
Future Gratification
Best of Hotstar for You - Masthead
Top Picks For You
You May Also Like based on X
Watch Next
3
Realisation
Deterministic, Enables
Faster Gratification
Continue Watching
Watch List
Shows / Tournament You Watched
Recent Search History
2
Customisation
User Selects, Invests
before Gratification
Preferences during Onboarding
Profiles
Content Rating (Kids safe)
1
32. What does Growth mean at D+ Hotstar
How do we ascertain which Metrics to grow
How do we set up Analytics
How do we build Segmentation frameworks &
Cohorts
How does analytics translate into User
Intelligence
How does insight inform Personalisation
How does personalisation create Growth Loops
Closing The Loop
33. Closing The Loop
App Learns More
About the User’s
Preferences
User Watches Content
OR
Does Not Engage
OR
Browse/Search
Shown Content to Watch
Based on Everything We
Know About the User
User Comes to App
CORE LOOP
37. Segmentation
● Real Time Segmentation:
○ Segmenting users who start
watching the Wimbledon
● Behavioral segmentation:
○ Segmenting freemium users
and understanding their intent
to purchase subscription
● Interest Based segmentation:
○ Segmenting users based on
the genre they watched, so
that the latest releases can be
sent out to them
39. RFM
Auto-Segmentation Tool
● Defining the power users so that
the new feature can be shared
with them
● Defining the users who are
about to churn
● Defining the loyalists and
sharing the content
41. Cohorts
● See the number of freemium users
returning after the premier
release/ sports season and
purchasing
42. Trends
● Trend:
○ Understanding the average
charged
● Trend Comparison:
○ Comparing the
video/product watch trend of
the premium vs freemium
users
○ Seeing the purchase trends
of user from different
regions
43. Funnels
● Funnel:
○ Understanding the drop off for
the new users after watching
free content
○ Understanding the median
time
● Funnel comparison:
○ Comparing users to different
regions and seeing where is
the maximum conversion
44. Pivots and Flows
● Flows:
○ Checking how freemium users
navigated once they first come to the
app
● Pivots:
○ Seeing the video consumption trend
during the day in different regions or
different segments
45. Real Impact
● Seeing average conversions for
subscription
● Measuring the primary KPIs as
compared to the control groups
○ Revenue per user
○ Stickiness of the App
○ Recency of the user