DOREMON DEN CLASS #1

AGENDA

1. Favourite App
2. What PM is? What PM is not?
3. Sample Case Studies
4. Chain game
5. Concept #1
6. Example

Doremon Den = It's a den of PMs who help Doremon make gadgets full of utility.

What is PM

● Anyone who knows to capture the applause of an audience is a great PM: Kunal Shah
● Not a specialist role, but a generalist role.
Chain Game - How is Whatsapp used (Not the WhatsApp Business app)
Ans:
Chat
For text messaging people
For talking to business numbers for customer supports
For sending marketing messages
For collaborating in groups
For payments
For sending voice notes, documents, location, contacts
For broadcasting messages to many
For transferring files across devices
For personal note taking

Call
For voice, video calls to international numbers
For group calls

Status
For publicising events in statuses
For notifying our circle through statuses

How did we bucket the above list items?

● core/2ndary/tertiary features
● chat/voice/status
● based on % penetration of features
● based on chronology of features introduced
● different UI screens
● 1 to1/ 1tofew / 1toMany
Concept: Bucketing
Principle: MECE - Mutually exclusive, Collectively Exhaustive

Example -

1. Cricket Test Match Result for Team A - Win/Loss - Is it MECE? No. Not CE.
MECE - Win/Loss/Draw/Tie/No result

2. Revenue from products A,B,C from countries X,Y. (Assumption - A,B,C all are in X & Y)
● Rev(A) + Rev(B) + Rev(X) + Rev(Y) : Only CE.
● A+B+Y : None
● C+X+Y : Only CE.
● A+C : Only ME.

MECE = X+Y (basis = country) or A+B+C (basis = product category)

MANTRA TO REMEMBER WHILE BUCKETING:

Definition of 3 things - 1. Problem Statement 2. Buckets 3. List items

Other Possible Bucketing Ways For the Problem statement - How is Whatsapp used ?
1. By Contacts Categories - Not ME.
Family
To send text messages
Relative (Distanced family members)
Friends
School friends
College friends
Workplace friends
Teachers/Professors
Work Colleagues
To send text messages
Customers
Business bots
Myself
Unsaved contacts

2. By communication channel - MECE & the best way to ans.

1. Chats
● Talking with family, relatives, Friends
● Sharing pics, videos, and docs
● To talk to school friends groups
● To communicate with office teams
● To purchase clothes, jewelry, food medicine
● To talk to a business bot
● To receive BookMyShow movie tickets
● To share multiple photos
● To share contacts.
● To share location
● To take notes
● To compress images and media
● To transfer media from mobile to desktop
2. Status
● To share content with all
● To market some link
● To check what others are doing in life
● To circulate covid emergency posts
3. Call
● To voice call office colleagues
● To voice call friends group
 ● To video call mummy
● To keep track of calls
● To conference call

3. By no of people communicating with - Is MECE but not the best way

1. To communicate with 0 other person
● To take notes
● To store messages, media from other chats
● To compress media
● To transfer files to the desktop or vice-versa.
● To see profile picture
2. To communicate with only 1 other person
● To answer kurti queries by a friend
● To check what others are doing in life by watching status
● To video call 1 person
● To share an image of a coding bug to a developer
3. To communicate with 1+ people
● To send kurti catalog in a group
● To share content with all
● To video call a group
● To market some link
● To put statuses
● To circulate covid emergency posts
● To voice call office colleagues
● To voice call friends group
● To conference call
● To share an image of a coding bug to the tech team
Other common bases for buckets
○ Profitability Calculation

○ User journey - user enters, the user interacts, user exits

○ Format/method - text/video/voice
○ Purpose - boards, competitive exams
○ Frequency - active, medium active, casual users
○ Stakeholders - food orderers, delivery Partners, restaurants
○ Stakeholders’ might - B2B (business to business) mei small, medium, large businesses
○ Target community by age - 0-5, 5-15, 15-25, 25+
○ Target community by geography - MH, GJ, Others
○ Monetary perspective - paid, free

THE END
THANK YOU!
 DOREMON DEN CLASS #2
AGENDA
● What are Metrics
● Funnel Analysis
● Models to understand User Psychology

Metrics = mathematical numbers/fractions/percentages

What are Metrics?

Quantifiable data points tracked & analysed

AARRR A+AARRR

Acquisition = app download Awareness

Activation Acquisition = app download
Retention = coming again n again Activation
Revenue Retention = coming again n again
Referral Revenue
Referral
By Dave McClure.
Pirate, because pirates exclaim Arrr!

Engagement = Retention ?

Retention Engagement

No of sessions per user (each session might not Total time spent on the product (while completing
mean actions on the app) actions)

Every day notification sent “New offer unlocked” User visits once in 3 days -> checks 3 offers,
-> users opens the app -> doesn’t like the offer -> explores other sections on the app -> drops in 90
drops in 9 sec min
More exhaustive:

AAAERRR
Awareness
Acquisition
Activation
Engagement
Retention
Revenue
Referral
Eg: Swiggy Business

Awareness Acquisition Activation Engagement Retention Revenue Referral

No of app Total app Total Time spent No of AOV (Avg order No of users
page visits downloads, signups/ on each app users value) - ORDER referred/No of
on App week, visit, with 1+ BASED current users,
CAC(Cust
store, Play
orders,
acquisition Total No of ARPU (avg rev per NPS (Net
Store
cost) users restaurant user) - USER promoter
No of
no of social with at pages visited BASED score) =
users
media posts least 1 before
with MRR (Monthly 9/10 -
impressions restaurant adding to
multiple Recurring Promoters
page cart/user
viewed app Revenue), 7/ 8 - Passives
sessions 6 and below -
LTV (LifeTime
Detractors
Value) =
Avg amt paid per
NPS =
order*no of
%pro-%det
orders made in
the whole life

Eg:
Monthly rev 1000, orders 10, users 5
AOV = 1000/10 = 100
ARPU = 1000/5 = 200
 USER PSYCHOLOGY

Weapons of Influence

1. Reciprocation
Reciprocation recognizes that people feel indebted to those who do something for them or give them a gift.
Examples - linkedin Endorsement, zerodha, prime, ux pin, cashback.

2. Commitment and Consistency

After making a choice or stating a commitment, both personal and interpersonal pressures drive us to
behave consistently with it even after the initial motivation is no longer there.
Examples - Duolingo (since you started this, complete it), Udemy payments.

3. Social Proof
When someone is uncertain about a course of action, they tend to look at those around them to guide their
decisions and actions.
Examples - Like, respond & comments on social media, polls, reviews and ratings of Products imdb etc

4. Liking
People are easily persuaded by other people that they know and like.
Examples - Referrals from friends, spotify friends list, brand ambassadors

5. Authority
People respect authority and want to follow the lead of real experts.
Examples - Crypto promoted by Musk, Verified account

6. Scarcity
The thought of losing something motivates us more than the thought of gaining something of similar value.
Example - Clubhouse, leher, billion day sale

Read - Influence: The Psychology of Persuasion.

Fogg Behaviour Model

B=MAT

Behaviour = Motivation * Ability (Simplicity) * Trigger

Behaviour = an action that someone might do

Types of motivation

Sensation - physical level of motivation.

We’re motivated to seek pleasure (youtube) and avoid pain (calm).

Anticipation - emotional level of motivation, specifically, hope (broking apps) and fear (savings acc)

Belonging - social level of motivation

People want to feel like they belong and don’t want to feel rejected (insta).
Ability (How can things become simple)

How can things become non-Able Examples

Time Opening a Demat Acc

Money iPhone for commoners

Physical Effort Physical KYC

Thought Paytm Homescreen

Social Deviance Sharing a video showing physical assault

Non-Routine Gym at night

Types of Triggers

Motivation Ability Type of Trigger Examples

No Yes Spark - can create motivation Zerodha Varsity

Yes No Facilitator - can simplify the action eKYC

Yes Yes Signal - can launch the rocket Zomato notifications

Hook Cycle - by Nir Eyal

S Stages Goal Example - Insta Reels

No

1 Trigger To use external triggers to form internal Showing reels posted by friends on
triggers in the long run. in the long run. Feed Tab.
a. External - Ads, PR, email &
push notifications.
b. Internal - self initiated,
emotions or situations

2 Action To remove as many steps as possible & Getting the user to click on it &
make it a thoughtless journey for the user. smoothly land on Reels Tab.

3 Variable To promote anticipation & build interest. Giving the user content according
Reward to mood; Tweaking the algorithm
such that the user doesn’t know
which Reel is coming next.

4 Invest-ment 1. To let the user invest something on the 1. Making him follow a
hashtag, save an
app which makes him/her return audio/reels, etc.
2. Showing him an ad.
2. To get him to return the favour

THE END
 DD Call #3

Agenda

1. Users vs Customers
2. What is User Journey
3. UJ Activity
4. DD Syllabus

User v/s Customer

● User is anyone who uses your product (not necessarily a customer, but can be a potential
customer).
● Customer is anyone who pays you for your product. (will be a user).
Example :-

If you are a normal YouTube user you aren’t a customer for YouTube, but are a potential customer

(can buy YouTube Premium)

USER JOURNEY

What is UJ? Assuming uska koi goal hai, vo kon kon paths se jaa raha to achieve that goal.

UJ = Source -> A -> B -> Destination

Eg:

Product: Swiggy

Task for the user:

To compare restaurants & pick the one with the optimised cart value of your order for ordering a meal

(paneer & roti)

User Journeys:

1. Touchpoint Version

Used mainly in RCA.

2. Overview Version

Used mainly in:

1. Product Ideation
2. User Journey Optimisation
3. Explaining to Users
 Doremon Den Session #4
AGENDA

● Use Case v/s User Goals

● User Personas
● Product Requirements Identification

Use Cases User Goals

How do we use the product Why do we use the product

How I use WhatsApp Why I use WhatsApp

Sending Texts & Making Voice/Video Calls To stay connected with people

Why are user goals important?

Basis for making USER PERSONAS and identifying PRODUCT REQUIREMENTS

User Types User Themes User Personas

Order Makers Different subtypes of order Details of each theme

Restaurant Partners makers based on goals
Delivery Agents
PS: Food ordering App User Personas
Note: food ordering , pre-covid scenario

Different psycho/demography -> Different goals -> Different pain pts -> Different PRODUCT
requiremen
SN Persona Person Demographic Details Psychographic Details Goals Pain Points in achieving Goals (hurdles)
o Theme a Name (Age (Personality,
Gender Beliefs,
employment, education, Interests,
income, Lifestyle)
marriage status,
etc)
1 Not Rajat Rajat is 20 yr old. Rajat is tech-savvy & has a lot of Alternate for mess - Doesn’t get discounts elsewhere.
earning He doesn’t earn right time in hand. food, (skip + close + - Doesn’t get small cart value orders
students now & is unmarried. He He lives in a hostel & is a part of night) - Doesn’t get different payment modes
goes to college and many friends groups, usually - Can’t order at midnight.
lives in a hostel. skips breakfast
2 Unmarried Aditya 24 yrs of age, male, bachelor lives in a metro, shares a flat with 2-3 unwilling to cook, skip his - a dearth of healthy food on the app.
Working degree, unmarried, earns colleagues, ,9-6 job Mon to fri, fixed dabba, or uses swiggy as - pre-scheduling options not available
Professiona 50-60 k per month meal timings, has subscribed to dabba dabba, house parties - cheaper options with good discounts unavailable
l or maid comes, usually free on - Pickups & BNPL not available"
weekends, likes to go out & party or - splitting the bill & group ordering
hangout - ordering takes a lot of time & difficult to order
during working hours
- no subscription options available
3 married Neeti 35 yrs, male & female , lives in a loaned flat, nuclear family, unwilling to cook, food - difficulties with overwhelming options
working mohan married, graduate, lives in a both work on weekdays, wife cooks needed at home while they - healthy & hygienic food choices
couples metro earns 1-2 L post-work, has a kid 5-6 yrs of age, are away - multiple cuisines/restaurant ordering not
parents are quite concerned about kids available
health
4 family ka Rahul 23, unmarried, tier2, lives in a joint family likes snacking but as a change of cuisine ya - ordering from multiple restaurants not possible
vaaris unemployed, college-going parents prohibit, rich hai family gatherings - discounts on bulk orders not available
- COD options not available a* experience is bad"
- lesser street food partners
How to Present in a Deck
Pain Points Prod requirements

Incremental Soln Bridge Soln Moonshot Idea

(Basic increment, so not very Something that connects incremental & (very ambitious idea. But will show that you are

innovative) moonshot) visionary)

No hygiene Include Hygiene rating in Restaurants Include a ‘Hygiene Lounge’:

info of rating A VR experience to check out the actual ambience

restaurant of a restaurant

Include a ‘Hygiene Cell’ where we include data

pts like:

● Density of workers
● Frequency of replacing kitchenwears
● Freq of cleaning hands
● Freq of cleaning ambience
● Overall hygiene rating
Additionally, include Photos.
No Mark a dish as ‘Healthy’ Mark a dish with ‘Exact kcal & carbs,proteins,etc.’

calories-relate + +

Create a Healthy Section where such Create a Healthy Section where users can enter
d info on food
dishes will be filtered. max limits & thus filter.

Create a Healthy Section in each resto menu

where users can enter max limits & thus filter.

Mark a dish with ‘Low/Med/High Calorie’.

Create a Healthy Section where such dishes will

be filtered.
Difficulty in Displaying ‘Suitable for X’ based on Displaying quantity for each dish type (to be input
by restos) :
user input:
deciding food ● Rotis - 1 Swiggy Pc
● Biryani - 1 Swiggy Box/Suitable for X
quantity User orders >
● Gravy - 1 Swiggy Bowl/Suitable for X
Ask a Dishwise QnA in the tracking
● Starters - 1 Swiggy Piece
screen ‘Ordered for X people’ >
Data feeds the algo.
Swiggy will provide standardised packaging & users
can complain about the lesser volume.
“Most people find this dish suitable for
2 people” shown on the menu page.
Input by resto + fulfilment by Swiggy .

Input by users .

Displaying

a. Pieces for discrete dishes

b. ‘Suitable for X’ for non-discrete dishes

Input by resto .

(Prone to non-fulfillment)
 DOREMON DEN - Class 5

AGENDA:
● Application Process
● Resume making
● Deck making

The Process:

Pre - Resume

1. LinkedIn profile
2. Case Studies
3. A project
a. A people-oriented hobby (blog for readers, community for communication, etc.)
b. A product prototype
c. A product
4. A startup

(Try executing Product sides of things for your historical works.)

Resume

1. Skills
a. Figma
b. Adv Excel - Pivot Table, Macros, Graph types, Functions
c. SQL - www.w3schools.com, hackerrank, sqlzoo.net
d. Google Analytics, Clevertap, Amplitude, Mixpanel
2. Internships
3. Pre-resume (execution)
4. Achievements - Cases, Hackathons
5. PORs
6. Interests

Tests

1. Aptitude - time, pariksha

2. Case home assignment

Interview

1. Product Design
2. Product Metrics
3. Product Evaluation/Critique (Favourite Product) + Improvement
4. Product Strategy
● GTM (Go-To-Market) strategy - GTM of Insta Reels?
● Market entry - Should Flipkart enter the Gift delivery market?
5. Pricing
6. RCA - Instagram is facing a 10% decrease in DAU since the last 2 days. Why?
7. Problem Solving
 8. Technical
a. Internet basics
b. Databases & APIs
c. System Design
d. DSA - data struc & algo - gfg
9. Guesstimates

Resume tips

Structure

1. 1-page maximum (especially for freshers)

2. 1 - column preferable
a. ATS scans don’t always parse information from columns properly.
b. Is not followed by senior applicants, so difficult to follow for human eyes.
3. Try listing in reverse chronological order (within sections)
4. Relative font sizes should be close. It’s not like you use 20 for title and 8 for description

5. Each bullet should have max 1-2 lines

6. The order of sections depends on your history and JD
7. White spaces Laxman Rekha: Inexperience | Bad design
8. Name it like ‘Rahool_Jaen_Resume’

Content

1. 'Objective' can be avoided if you have other points to cover.

2. Can follow STAR (Situation, Task, Action, Result)
3. Action words (improved, collaborated, analyzed) -> to show action
4. Quantitative data points -> to show impact
5. Use PM buzzwords at moderate level
 Incorrect way:

Correct way:

Some keywords for different roles:

Strategy keywords

a. Developed a product vision and strategy

b. Created product personas
c. Built a product roadmap
d. Conducted a strategic review
e. Recommended strategic initiatives
f. Etc.
UX / UI design keywords

g. Designed initial wireframes

h. Gathered UX feedback from users
i. Researched competitors' UX
j. Provided feedback to design team
k. Contributed to the development of a design system
l. Etc.
Coding keywords

m. Wrote detailed user stories for engineering team

n. Made initial tech difficulty assessments for features
o. Negotiated timelines with Lead Engineer / CTO
p. Helped make engineering trade-off decisions
q. Built front-end / back-end features for app XYZ
r. Etc.
Digital marketing keywords
 s. Created value props with marketing team
t. Provided feedback on copywriting
u. Coordinated launch with marketing team
v. Tested five acquisition channels
w. Scaled a channel to X number of users
x. Etc.
Data analysis keywords

y. Set key success metrics

z. Measured metrics' improvement
aa. Performed analysis on large datasets
bb. Generated actionable insights
cc. Summarized data findings
dd. Etc.
Leadership keywords

ee. Lead a team

ff. Aligned stakeholders
gg. Managed a project
hh. Took an initiative
ii. Set a vision
jj. Etc.
Communication keywords

kk. Communicated next steps

ll. Summarized a complex situation
mm. Ran product meetings / reviews
nn. Presented in front of a large audience
oo. Wrote product requirement documents
pp. Etc.
Organization keywords

qq. Delivered a project

rr. Ran two-week agile sprints
ss. Organized an event
tt. Improved shipping cadence
uu. Optimized a process
vv. Etc.
How to present Internships

Actual Title Title in Resume Explanation in the

Interview

Software/FE/BE Developer Product Developer

UI/UX Designer Product Designer

Should know product
Data/Business Analyst Product Analyst concepts written in the
description
Community manager Growth Manager

Social Media/SEO Manager Product Marketing Manager

Operations Manager Generalist

Understand the difference between - Tax Avoidance & Tax Evasion

Note - If mentioning 3 or more Experiences, keep 1 Experience with the ‘Actual Title’.

Incorrect way:

Correct way:

General Gyan:

1. You can customize based on job requirements. Education can be relevant in cases JDs have asked
specifically, but, if not, you can also highlight your best work at the top.
2. Question each point by thinking from the reader's POV
3. You are targeting 2 people with the same resume: 1) HR, 2) Profile specific senior. Serve both.

Sample Resume: ​https://drive.google.com/file/d/1lWm1ImwwGsn1Dfx-33ktV-LXqkWia3_s/view?usp=sharing

 Where to apply

MNC

Established Startups

Unicorns

Soonicorns

Series A+ Startups

Other Startups

Where to find where to apply

1. Recently Funded Startups - https://yourstory.com/search?category=Funding

2. Unicorns - https://www.ventureintelligence.com/Indian-Unicorn-Tracker.php
3. Funding Data - https://www.crunchbase.com

Cover letter (was used for colleges/professors):

1. Brief
2. Catchy
3. Personalised

Eg:

Hello from Rahool Jaen!

● I am an APM guy with 4 months of Product learning.

● I drove 71% engagement at a previous Product intern.
● I won IIMA Red summit Product Case.
● I steered my team to the National Finals in EV case.
● I can join immediately.

PFA my resume for my application as an APM at Peesho.

Hoping to hear back from you,

Rahool Jain.
 How to reach out to recruiters

How to cold mail:

1. Use tools to find email IDs of people

2. Add a catchy subject
3. Use cover letter & resume
4. Follow-up (generally, 2 times in 2 days gap each)

How to network on LinkedIn:

A. Prepared Dekisugi
1. Find the most helpful & active person in the Product team
2. Personal invite
3. Offer help (asking gyaan/appreciating them/discussing a common interest/suggesting productivity
resources)
4. Ask for a referral

B. Desperate Nobita
1. Send connection requests to all PMs in your desired companies

APM PM Sr PM Lead PM Director, PM VP, PM CPO

2. Start a convo
3. Share a crisp cover letter asking for help
4. Share resume

LI Personalised Invites:

Content

1. Who are you? (Optional)

2. Why should that person accept your invite? (Compulsory)
Structure:

1. Short (for lazy people) to medium (for people who appreciate efforts)
Eg:

1. (Medium)

Hi, Arvint!

This is Rahool from NIT Warangal!

I am a budding Product Manager & you seem to be at a place I’d want myself to be in future. Looking

forward to connecting with you so I can enjoy your company, actively or passively.

2. (Short)

Hey, Revadi!

Completely in awe of your post on APIs for PMs. Don’t want to miss any future posts. Please accept my

request to avoid my FOMO!

What if you’re from non-Tech?

1. Read Tech - hackernoon, TechCrunch, Abhishek R. (Swiggy)

2. Explore Tech -
a. Explore Products
i. Pick the top 2 products from your favourite 5 industries.
ii. Explore them across AAAERRR.
iii. Note Product/Tech innovations down.
iv. Mention those in your resume.
b. Find ways to automate things
i. Eg: Send personalised attachments - YAMM browser extension
ii. Eg: Send Auto Emails to form responses - Google Form Extensions, Google Apps Script
3. Execute Tech
a. Build a video resume - https://www.loom.com/
b. Build a No-code Landing Page/App

How to Present a Problem + Solution in a Deck

https://www.canva.com/design/DAEturlpTiA/uZWuxwjQ2wPI1_8pbDWUXg/view?utm_content=DAEturlpTi
A&utm_campaign=designshare&utm_medium=link&utm_source=sharebutton
 Doremon Den Session #6

AGENDA

● Guesstimates
● Sample Problem
● How to track Metrics

Guess + Estimate.

Basically making guess to estimate particular values

Common Structuring Techniques

1. Top-Down (Start Big)

2. Bottom-Up (Start Small)
3. Constraint-Based

Common Splitting Techniques

1. Rural-Urban (Geography-wise)
2. Gender Split
3. Age Split
4. Income Split
5. Willingness
 Question - Estimate the number of news swiped on Inshorts in a day.

Clarify

1. Reading from notifications or via app - landing & reading the article == news swiped
2. Per user or total no of swipes - total
3. Weekday or weekends - both
4. Are we considering all categories of news - yes
5. App or website - app
6. Geography - India
7. What is the criteria to consider the news as read - reads through 80% of the content
HOW TO DEFINE METRICS
 Sample Metric Formats:

Metric
Index
Entity

Timeline

Slicer

Sample

Entity == User Entity != User Slicer added

No of Resto pages visited Time spent on the app No of Resto pages visited

__________________ ______________________________
________________________________
User * Week User * Month * State
Session * Week
 Function User Activity Business Tech

Goal How is the user behaving with the How much is the business generated by the How is the performance of tech in facilitating the
product? product? product?

Example DAU, Time spent, No of downloads, ARPU, Avg revenue per month Latency, initial load time, uptime, data loss rate,
etc. http errors,...

Optimisation Category Positive/Optimisation Metrics Negative/De-optimisation Metrics

Definition Metrics whose value should increase as time passes. Metrics whose value should decrease as time
passes.

Example DAU, ARPU No of days taken to make the 1st purchase,

Latency to open a restaurant page

How to solve a Metrics Question

1. Define Product - Top Level Mission/Vision, How does the product work, User types
2. Decide Goal for Measuring (AAAERRR), if not there
3. Map out the user journey from all users types’ perspectives
4. Define Signals/Reasons
5. Define Metrics for all Function (if not mentioned otherwise in the ques)
6. Prioritise
7. Repeat 4,5,6 for Positive & Negative Metrics
When goal is not given, identify goals using this table (can put constraints based on the problem):

Product Example A A A E R R R
Life

Early Tyke Invest

Stage

Growth Paytm Money

Maturity Instagram Stories

Question: Define Engagement Metrics for Instagram Reels

1. Product
Mission (IG)

A platform to help people communicate, express & stay in touch with people virtually by

sharing photos & videos.

Mission (IG Reels)

A tool to express & communicate with a wider audience using short videos.

How The Product Works

A create tab to record, edit & upload videos, and a Reels Tab to swipe & see the created videos

User Types

Content Consumers

Content Creators

2. Goal

Track Engagement

Link to the Metric Sheet -

https://docs.google.com/spreadsheets/d/1W_y2NjSrVV_EsY6f0ZYrlkhja6gGesgoTVKG33I8-qs/edit?usp=sharing
 Doremon Den Session #7

AGENDA

● Prioritisation Frameworks
○ Value Effort
○ RICE
○ RUMERCOTH
● Sample Problem

Prioritisation Frameworks

1. Value vs Effort

2. RICE
Flaw with RICE: No consideration of Monetary ROI

Soln:
1. Ask the Goal:
a. No of users
b. Monetary Returns (Revenue / Profit)
2. Ask the timeline to achieve the goal
3. The priority will be depended now on : Impact/Monetary Input
a. Impact = No of Users Affected OR Monetary returns (Revenue / Profit)
b. Monetary Input = Tech Stack + HR + Other Costs

3) RUMERCOTH

Reach ✖ Impact✖ Confidence

RICE Score =
Effort

Reach ✖ Efficacy ✖ UX ✖ Risk ✖ Money ✖ Confidence

RUMER-COTH =
Tech Stack + Human Resource + Other Costs

Absolute
Term Explanation Units
/ Relative

Number of users who were affected

Reach Absolute No of people
(see/understand/interact) by the feature.

- Percent
Efficacy Percentage of users for whom the goal is met.
(Multiples of 5)

User How easily/quickly can the user see/understand/interact with Relative (1 to 5, Multiples
Experience the solution of 0.5)

How likely is it that other product(s) will witness a negative

impact, when this feature is introduced/adopted?

Risk = cannibalisation Eg: Relative (0 to 1, Multiples

Risks
1. Walnut SMS- low usage of the product of 0.1)
2. Fk flights cart - low completion of the payment
3. Paytm launching a chinese game - negative publicity
(low retention)

Money What is the average revenue/profit per user? Absolute Rupee

Proportional to the reliability of the data your team has to Percentage

Confidence -
support one or more factors of the score. (100%/80%/50%)
 Human Salaries of Tech, Product, Design, Sales, Marketing, CSM,
Absolute Rupee
Resources etc.

Fixed cost - Domain name, Hardware

Tech Stack Absolute Rupee
Recurring - Notion, Jira, cloud, analytics tool, dev ops tools

Operations + marketing cost + sales cost + others if any

● Operations - logistics or purchase of any other
resource
Other Costs ● Marketing - spend on ads (offline + online) + Absolute Rupee
promotional offers
● Sales Cost - money spent on agent field trips,
vouchers, presents etc.

Sample:

Feature Reach Impact Confidence Cost Scor

e
Efficacy Mone User Risks Human Tech Others
y Efforts Resources Stack

A 1M 100% - 1 0.4 100% 50K 10K 2K x

B 2M 60% - 4 0.5 80% 100K 1.5K 1K y

C 5M 50% - 2 0.7 50% 80K 5K 5K z

Problem Statement - Prioritise the following 3 solutions.

1. Explanatory i Button - A pop-up explainer for key terms used in Jira Issues (Tickets)
2. Video Explainer - A video walkthrough for issue creation with key terms explained.

3. Guided Tutorial - A step by step user driven tutorial with explanations

Step 1
 Step 2

Last Step
Current solution
 Solution

Goal: improving activation for new users who want to create Jira Issues. Making them create their first Jira
ticket
Time period: 3months
Product: Jira Website
Assumption: these features come up after the user clicks on the Create button on the top nav bar.

Absolute/
Term Explanation Units
Relative

Number of new users used the feature to understand issue

Reach Absolute No of people
creation.

Percentage of users getting activated post feature usage / - Percent

Efficacy
% of users creating the jira just after the feature usage. (Multiples of 5)

-
User How easily/quickly users used the i button/ video/ guided Relative
(1 to 5, Multiples
Experience tutorial
of 0.5)

-
How likely is it that those users won’t try other features of Relative
Risks (0 to 1, Multiples
Jira?
of 0.1)

Average profit/Revenue generated per user from the

Money Absolute Rupee
implementation of the i button/ video/ guided tutorial

Percentage
Reliability of data to support reach, efficacy, UX, risk,
Confidence - (100%/80%/50%
revenue, of the features.
)

Human
Absolute Rupee
Resources

Fixed cost - Nil

Tech Stack Absolute Rupee
Recurring - Server storage & computation costs.

Operations - any logistics costs involved

Marketing – cost of marketing the new guide/tutorial
Others Absolute Rupee
feature if any
Sales - cost of selling the new feature to users
 Impact Cost
Feature Reach User Confidence Human Score
Efficacy Money Risks Tech Stack Others
Efforts Resources

Explanatory (i)
7.5k 75% 1 - - 100% 39K 0 0 5.1/39 = 0.12
button

Walkthrough 0.18/72.5 =
12k 50% 4 - - 80% 71K 1.5K 0
animated pages 0.027

Assisted tutorial 15k 90% 3 - - 80% 68K 0 0 32.4/69 = 0.48

New users acquired (who create jira) per month - 6K

New users acquired (who create jira) per month (who never used Jira before in another org) - 5K

Impact
Solution Reach Confidence
Efficacy User Experience Risks Money

50% - shifting user behaviour

Ways to learn 75% - user driven and
100% ( percentage
1. Youtube videos - 30% > 5% multiple visits possible,
Explanatory 1 - clicks on i button completion of issue
2. From colleagues - 70% > 45% but still the wording may - -
i button then read the text creation post clicks on
Total share = 50% be confusing to some,
the current i button)
so taking 3/4
= 3* 0.5 * 5k = 7.5K

80% - some might ask their seniors or 50% - info provided in

4 - no user actions
Video they are given onboarding training bulk, there can be gaps 80% (either user
just watching & - -
explainer in understanding and research or derived data)
reading screen text
= 3* 0.8 * 5k = 12k confusion persists.
 100%- Max reach
Guided 90% - there can be a 3 - just reading the 80% (industry data on
User driven & very intuitive - -
tutorial small drop off, text guided tutorial screen)
= 3*(5k) -= 15K

Tech Setup Human Resource Other Costs

Feature
One Time Recurring One Time Recurring Sales Marketing Operations

1 content creator for 1 day - 1.3K

1 Designer for 2 days - 3.3K
0 BE Developer -
1 FE developer for 2 days - 8K
1 QA for 1 day - 2.5K
Explanatory i
0 0 1 PM for 6 days (1+2+2+1) - 20K 0 0 0 0
button
1 Devops for 1 day - 4K
0 Sales
0 CSM

= 39K

1 video editor for 3 days​​- 5K

1 content creator
(script+recording) for 5 days - 6.6K
Video file size - 1 Designer for 2 days - 3.3K
300 MB (length 1 BE Developer for 1 day - 4K
Video
0 5 mins) 1 FE developer for 2 days - 8K 0 00 0 0
explainer
1 Devops for 2 days - 8K
= 1.5K 1 QA for 1 day - 2.5K
1 PM for 10 days - 33.3K

= 71K
 1 content creator for 1 day - 1.3K
1 Designer for 3 days - 5K
0 BE Developer
1 FE developer for 5 days - 20K
Guided tutorial 0 0 1 Devops for 1 days - 4K 0 0 0 0
1 QA for 2 day - 5K
1 PM for 10 days - 33K

= 68K
 Doremon Den New Session

AGENDA

1. How does the Web work?

2. All about APIs

HOW DOES THE WEB WORK?

1. Enter the URL of the website.

URL - Uniform Resource Locator, a string containing details about

● Protocol - Set of rules for sending Requests & receiving Responses.

● Domain Name - The name of the website which is mapped to the IP of the server where the
website is hosted.

● Resource - The exact location of the requested page on the web server.

2. The browser communicates with the ISP (voda, jio) to do a DNS lookup to find the IP address of the
web server hosting the website.

● DNS connects to the root name server to get the IP address of the .com name server.
● Contact the .com server to get the IP address of the web server hosting the website.

ISP (Internet Service Provider) - Your Internet Provider

 DNS (Domain Name System) - A distributed database of domain names of the websites & their IP

addresses

IP Address (Internet Protocol) - a set of 4 numbers separated by dots which uniquely identifies a

device (server, laptops, desktops, mobiles, printers)

192.111.4.x

3. The ISP returns the IP address of the web server to the web browser.

4. Now the web browser uses the IP address & the port to open a TCP socket connection with the server.

5. Now, http (or any such protocol) request is sent to the browser to GET the desired webpage.
 6. On receiving the request it looks for the desired resources and sends a response (the HTML page in
this case)

How is the HTML file sent to the web browser? - Data packets

7. On receiving the HTML file the web browser parses the file and makes new requests to the server to
provide assets mentioned in the HTML file.
Assets are basically css & js files, images and other media files.

8. Painting the final picture

a. The web browser parses the HTML file and creates a tree structure of the DOM (Document
Object Model) The tree is basically a method to structurally represent the objects in the
HTML file. Each Object in the tree is called a node.

b. The CSS file is then parsed to apply styling to each node

c. Finally the page is ready to view.

Languages used at The Client Side

a. HTML - basic physical structure

 b. CSS - Look and Feel

c. Javascript - User Interactions

WHAT ARE APIs(Application programming interface)?

Keyword = Interface
1. Why should you learn?
Opportunities for ur product

API Audit – what APIs are currently being maintained but are not used?

Confidence – while talking to the Tech team

2. Definition

System 1/Customer (requesting) -----API / Waiter----- System 2/Kitchen (responding)

a. Kitchen
The database - no customers allowed, to protect data integrity.

b. Waiter
The API - a middleman that knows how to serve data from the database without

disrupting its functioning.

c. Customer
An external system - that wants to get their data

d. Menu
The data format - reference the external systems have to use to perform their

operation.

e. Order
An actual single API call.

3. Eg:
● I send a request to the Spotify API
 ● As part of my request I tell Spotify which album I want to retrieve a list of songs for
using an ID e.g. Spice, by the Spice Girls (a classic) has a unique ID
● Spotify responds with a list of songs from the Spice album in a language called
‘JSON’
● The API documentation will specify what the request should be and what you can
expect in the response.

4. Uses:
Integrating with third party APIs. Eg: Google maps, Facebook login

Speed of product delivery Dependencies

Low effort Risks

Costs

5. Guess the types of APIs

Private/Partner/Public

6. Question: Heard ‘We’ll HIT gmail login API to get that data’ What do we hit here? - Endpoints
Endpoints = Doors which can be opened and get what you want.
7. API Calls
8. Payloads - It refers specifically to the meaningful information in a given set of data.
9. Authentication and authorisation:
Authentication – proves who you are (Username, pw, access token)

Authorisation – defines the what you are permitted, or authorised to do

10. OAuth - open-standard authorization framework for APIs. It offers secure and restricted access to the
end-users’ data which is to be used by the requester without accessing their password.

Receptionist ➖ Authorization Server

Room key/Card ➖ Access Token

Door lock ➖ API

 11. Rate-limiting - controlling the rate of incoming as well as outgoing traffic is called rate-limiting. It’s also
defined as the total number of requests that a user hits to the API.

12. API throttling - refers to the process in which the usage of APIs by consumers is controlled for a specific
duration.

13. Pagination - process to determine as to how much data should be displayed and at what frequency.

Insta Frontend (tanay.username) —-- Insta Database (meme1, photo2,....)

Eg: You need 15 albums per day.

Current Spotify API allows only 5 API hits per day till March
1 hit gives you 1 album only.

14. Wrappers - packages that wrap sets of API calls into easy-to-use functions. The wrapper programmatically
calls multiple API calls without requiring user interaction, further automating projects.

15. REST (Representational State transfer) APIs

category of APIs that follow a certain guideline

SI unit = rest api

CGS unit

16. The basic characteristics of a RESTful API are:

1. Resource based
2. Representations – JSON (Javascript object notation), XML
“Key”:”value”
Name, std, gender, tanay, 11, m, priya, 12, f,....
{“Name”:Tanay
“Std”:11
“gender”:M },
“Name”:priya
“Std: 12
“gender”:M

3. Constraints –
1. Uniform interface (Typically uses HTTP resources, HTTP verbs as the actions: GET, PUT
(update), POST (create), DELETE

2. Stateless – each request has enough information for the server to process the message.
The server never relies on information from previous requests from the client. If any such
information is important then the client will send that as part of the current request.

Request - URL, query parameters, HTTP body, or HTTP headers

Sample Request:
curl --location --request GET 'https://api-staging.xyz.com/communication/v1/chat/send/' \
--header 'Content-Type: application/json' \
--data-raw '{
"userId": {1001692722} ,
"placeHolderData": {"reliance": 5

},
"moduleName1": "CHAT",
"processName": "CHAT",
"source": "xyz",
"templateName": "welcome",
"channelId": "ce6427bf-b635-4e09-a4c9-8c6350a7d37f"
}'
Response : which date user bought

Request: which website is asking to integrate

Response: linked.com/login

Parikshit, parikshit123 -> login

Request: Parikshit, parikshit123
Response: parikshit ki profile ka sara sections

Data: { “userid”:1001692722,

“date”:”29/01/2022”,
“Price”:2000 }

200 series
400 series
500 series
Response - HTTP headers, status code, and HTTP response body
REST API Methods:

Method Status Code Common Status Request Idempotent

Code A^n = A

GET 200 (OK), 404 500 (INTERNAL Yes

(NOT FOUND) SERVER ERROR) if

the processing fails

POST 201 (CREATED) No
and the server errors
(Create a along with a When we invoke the same POST
out
resource) location header request N times, we will have N new

with the link to the resources on the server.

newly created

resource

PUT 200 (OK), 404 Yes

(Update a (NOT FOUND) If you invoke a PUT API N times, the

resource) very first request will update the

resource; the other N-1 requests will

just overwrite the same resource state

again and again – effectively not

changing anything.

Update (order 11) to (order 12)

DELETE 200 (OK), 404 Yes

(NOT FOUND)

Read: SOAP APIs

 Doremon Den New Session

AGENDA

● RCA (Root Cause Analysis)

● Problem Solving

RCA

Eg: Insta reels has 10% dec in DAU in the last 2 days. Why?

RCA Framework
A. Clarification - Is there a problem? Area of the problem?
a. Problem area scoping
1. Definition of terms in the PS
2. Users - Geography, age, religion, experience with the product, etc
3. Product platform - web/android/iOS
4. Industry-wide or not
5. Time period of comparison
6. When are we discussing this?
7. Pattern of Change - progressive/sudden?

b. Sanity check
1. Frequency of occurrence - 1st time/seasonal/repeat irregular
2. Change in usage of any other feature simultaneously?
3. Definition of metric changed in the period?
4. Metric analysis tool working properly?

B. Deep dive - What change is causing the problem?

1. Competitors - new competitor/new announcement by existing competitor or a new substitute
2. PEES (Political / Economical / Environmental / Social (user behaviour)l) -
P - New policy, legal regulations
E - Inflation, unemployment rates
E - Climate, env offsets
S - geography/age/gender/religion/culture

3. BUT Changes
Business - New PR/New Pricing strategy
​UI - UI changed for that feature or near that feature
Tech - App updates/Version specific/OS specific/Platform specific/new experiment/server issues

4. User Journey

After a section has a positive response, make hypotheses there.

Question: Zomato has had a 5% decrease in signups in the last 1 day. What is the root cause?

Clarification
1. Problem area scoping
Is this decrease specific to any geography or worldwide - india
Sign up meaning - new app downloads & then sign up
gradual/sudden - sudden
Do we know the time since it suddenly started decreasing - noon
5% meaning - wrt avg of last 1 week.
Gender specific - no
Is it device specific - android
Channel specific - no (no one is able to download from play store. Some are able to download using the apk)
All food delivery or only zomato - only zomato
Age group - all
Which day - sat
Which festival/event - no religious festival in last 2 weeks

2. Sanity check
Frequency of occurrence - never seen this big dec in 1 day
Change in usage of any other feature simultaneously? - download related mainly. Yes some features might
be affected
Definition of metric changed in the period? - no
Metric analysis tool working properly? - yes

Deep Dive
Deprioritising Competitors & PEES because zomato specific problem

Business
Change in name, logo, spelling - No
Is there a change in strategy to make people download - yes/no
Is there a reduction in no of ads - no. increased.
Is there an android specific campaign - yes.
What is this new campaign?
​ Is there a business issue in play store - might be

UI
Was there a UI change 10 days back - yes

Tech
Was the server down for some time in last 1 day - no
Is there any new app update - last update was 10 days ago
App size change hua kya - No
 ​Is there a tech issue in play store - No

User Journey

Is the decrease at the download step or the signup page - download step

User opens play store

User searches for zomato/food
User goes to zomato app page
User clicks on install
____
User opens the app
User enters details
User clicks on signup

(% change is the comparison of 1-day (today) with avg 1-day in the last 1 week)

User opens play store - what is the decline % here 0%

User searches for zomato/food - what is the decline % here +2%
User goes to zomato app page - what is the decline % here -5%
App ranking changed - no
Is zomato not listed - yes
Is there any issues zomato having with the play store - yes
Did zomato violate any regulations - yes
What regulation was violated - play store doesn’t list apps that allow betting.
For how long this betting feature has been active - since the last 10 days (in the last update)
Did play store notify zomato before removing - no, as per the policy , they notify after removing.
App’s rating changed
User clicks on install - what is the decline % here -5%

User opens the app

User enters details
User clicks on signup
 PROBLEM SOLVING
Types of problem solving questions:

1 2 3

Root cause not mentioned at all Hint about the root cause given Root cause mentioned.

Framework

1. Clarify
2. Find Root Cause/causes (if reqd)
3. Make user journey/List use-cases/List stakeholders to dissect the problem
4. Propose solutions

Q (Type 2): You are a PM for Ola. There are many riders who cancel their rides after confirming. How will you solve

this problem?

Assume - cause is not due to any competitor/external factor

Clarification Questions

Geography specific – no. Across india

any particular age group – no
any particular segment - no
riders – users/passengers
decline by how much – significant
problem happening since when – 2 months
time of the day specific – dont know
weekends/weekdays – both
More data ques - I don't have any data. You tell me possible causes & then I'll tell you which one to solve.

Candidate: I’ll make 4 buckets to find causes:

1. Competitor
2. PEES
3. BUT
4. User journey
Which one should I choose?

Interviewer: we can use the user journey.

Candidate: Should we start from the download step or booking?
Interviewer: start after a ride is booked

Candidate: Let’s bucket by 2 ways:

1. By User journey of riders

Cancelled before interacting with the driver

Wrong pickup/drop location

Schedule/Plan change - .

Alternate mode of transport available -

Price updated after trip confirmation -

Longer wait time than expected -

Cab/driver not as desired (rating, vehicle) - mushkil

Cancelled after interacting with the driver

Language issues - .

Driver goals diff from rider = extra fare, ac not functioning,

Inability to find the pick location

2. By Stakeholders

1. Rider
a. Change of plans, not travelling, members rose
b. Alternate means available
i. Other cab service as it was cheaper
 ii. Personal vehicle
iii. Public transport
c. Actual time of arrival more than ETA
d. Pickup, drop location, time, car type entered incorrectly.
e. Not liked the driver ratings or car allotted

2. Driver
a. Driver’s behaviour wasn’t good
b. Driver not ready to go.
c. Driver asked for some money
d. Driver asked for a particular payment method
e. Driver couldn’t find the pickup location
f. Driver not wanting to pay commission to Ola
g. Fuel insufficient or car is dysfunctional
h. Didn't pickup the call
i. Actual time to arrive more than ETA - driver irresponsible
j. Luggage space insufficient

3. App
a. App bugs
i. Wrong driver rider matching (OTP problem)
ii. Maps couldn’t properly show pickup location or driver location
iii. ETA calculation incorrect
iv. Price increased after confirming

Interviewer: Prioritise some causes

Candidate:
a. Alternate means available
i. Other cab service as it was cheaper
ii. Personal vehicle
iii. Public transport
b. Actual time of arrival more than ETA
c. Driver not ready to go.
d. Driver asked for a particular payment method
e. Driver couldn’t find the pickup location

Interviewer: Out of these 5, let’s solve for the 3rd one.

Candidate:
Problem: Driver not ready to go to that location.

UJ:
rider books
driver accepts
driver – rider call
driver cancels -> new driver allotted by app
rider cancels -> new booking journey begins

Pain points (assumption - driver can't deny a ride) -

1. Emergency with the driver (long route so not enough petrol)
2. Driver doesn't want to go to that location for no reason
Solutions:
1. feature to choose % of shorter to longer routes
depth – choose areas you don't want to go
risks - longer route might be priori -> put constraints in the algo
in the long term, drivers will contact & find more incentivissing config
-> limit the times a driver can change config; keep adapting the algo for incentivising diff
routes; keep min quota for all config
2. identify regions where most drivers dont want to go & incentivise more
depth – incentive = complete fewer routes
risks -
3. Pass to another driver
4. Hypothesis – cancelled due to wrong route by google maps
depth – offer the driver to change route acc to him
5. Driver ko dikha do location before accepting

Doremon Den New Session

AGENDA

1. Product Design Framework

2. A sample Product Design Question
3. Solution
4. Other Examples

PRODUCT DESIGN PROCESS

Framework Interview

1. Clarifying questions/Assumptions 1. Clarifying questions

2. Personas 2. Structure
3. Pain Points 3. Personas
4. Prod Requirements + Pitfalls 4. Pain Points
5. Prioritisation 5. Prod Requirements + Pitfalls
6. Metrics (+Roadmap, optional) 6. Prioritisation
7. Metrics (+Roadmap, optional)
8. Summary

Design An App For Children to Communicate

Clarifying questions (5W 10H)

Ask about the question, not about the solution.

1. WHO
 a. Who are kids here? Anyone below 13
b. Who are we? Not a new comer, and no tech constraint
2. WHAT
a. What problem are we trying to solve? Kids dont have a proper virtual channel to communicate.
b. What will be the mode of communication? Upto to you
c. What will be the structure for kids & parents? Upto you
3. WHY
a. Why we aretrying to solve? Because there is no in the market, and we see a market gap
4. HOW
a. How are we solving the problem? Upto you
5. WHEN
a. Whena re we planning to launch the app? No fixed timelines
6. WHERE
a. Where awe launching it (platform)? Mobiles
b. Where are we trying to launch it? India

Can we include hardware gadgets as well? No because then the reach of the product will decrease.

User Type
1. Children (age)
2. Parents/Guardians

User Personas

User Type - Children

Theme Name Demo Psycho Goal PP

0-5 Aman Tier 1,w2 Childish and Interacting - not knows how to write & communicate.
city kid, rich enjoys his day with friends - poor tech capabilities.
family, fully, lots of toys, & playing - loses interest very quickly.
recently plays a lot, games - parents dont want to promote the usage
started going watches lot of tv together of apps
to school, and plays many
parents are mobile games,
well educated studies for a some
hours during the
day

6-10 `Amani 8 yrs old, 8-2 school, Communicat - kids dont have their own device &
female, carefree, ing with parents dont give away their phones easily
school going, simpleton, cousins or - communication in chats is difficult.
lives with energetic, friends on - frequently switches devices.
parents in a curiosity, likes academic & - complex & unattractive UI.
city, mother playing with non-academi - kids may prioritise playing games over
housewife locality friends, c stuff communication
father works basic proficiency
in writing

11-13 Ankush 12 yr old, 8-3 school, Communicat - parents’prying eyes.

male, lives in evening tuition, ing with - Usage of the same app by parents &
the city with plays a lot with cousins or children makes it cluttered for the little
parents, friends outside, friends on master.
parents work. plays a lot of academic & - less attractive as compared to games.
 Looked after online games as non-academi
by well. Inquisitive. c stuff
grandparents.

Theme Name Demo Psycho Goal PP

To give - cannot monitor user activity & usage

controlled timings.
access to - Usage of the same app by parents &
kids for children makes it cluttered for the little
interacting & master.
growing - decline access to their chats

Pain Points

Not to be solved
- parents’prying eyes.

Already Solved
- Usage of the same app by parents & children makes it cluttered for the little master.
- decline access to their chats

To be Solved
- kids dont have their own device & parents dont give away their phones easily
- communication in chats is difficult.
- essential to solve for getting users engaged on the app.

- complex & unattractive UI.

- essential to solve for getting users engaged on the app.

- frequently switches devices

- user can shape his behaviour knowing he/she can only access Jingles from mother’s phone.
- high investment low return (high server usage costs for mutli-device usage)

- kids may prioritise playing games over communication

- same as unattractive UI

- cannot monitor user activity & usage timings.

- parents aere an essential stakeholder in kids using the app

Product requirements

1. communication in chats is difficult.

- Audio/Video chatting feature
- Audio to text feature.
- Text suggestion & auro-complete
- Animated stickers for communicating
- AR based message delivery assistant like Talking Tom
 2. complex & unattractive UI.
- Multi-player games
- Adding a 1bot for guiding the user
- Adding interest based groups.
- Themes & cartoon characters
- Gamifying actions on the app
- Streak

3. cannot monitor user activity

- A parent clone
- Activity dashboard, summary.
- NSFW content sharing alert
- Only parents can add contact on the kid’s phone.

4. Limitig usage
- App lock after a fixed time as set by parents, dynamic time limits
- Activating app during fixed hours of the day for use.
- Custom alerts & animations for kids.
-

Prioritisation

1. Rumercoth
2. Impact Cost
3. RICE

Metrics

Acquisition
● No. of app installs
● No. of users signed up with their mobile nos.
Engagement
● Time spent per user per day
● No. of contacts added per user
● Average messages sent per day
● Avg messages sent per chat per day
● No. of points earned/spent per user.
● No of stickers/templates/voice notes/ modulation used per day

Rest of the metrics will be based on the selection of the features.

Retention
● MAU/WAU/DAU
● 3/7/14/28 day retention of users
● Avg points scored by users.

Summary

What you did in the whole interview.

EXAMPLES
Design a pen for an astronaut
Design a phone for deaf people
Design a washer and dryer
Design a bike-based delivery service
Design an alarm clock
Design a new computer keyboard
Design Instagram for 40+ yr olds
 DOREMON DEN NEW SESSION

AGENDA

1. Favourite Product
2. Critique Product
3. Product Improvement

Favourite Product Framework

1. Product info - What’s the exact Value Proposition, How does it work
2. Business Objective - How does the product make money
3. Users types - Enlist & Prioritise
4. Pain points for each user type & how are they solved in your product.
5. Competitors - How these soln are better than competitors

1. Favourite feature/product - ‘Request’ button in GPay

Product info

Value Prop - Allowing users to request & thus remind from the same app for incomplete

payments
 How - Using Mobile No & UPI ids a user can request or pay money, a person can either pay

or decline a request received.

Business Objective

No direct monetisation, but helps in getting more transactions done.

User Types

1. Money Receivers
2. Money Senders

Pain points & how are they solved

Money Receivers (MR)

SNo Pain Points Solution

1 Not able to give reminders in the same app Request feature in the same app

3 Requesting for money on call/WA causes This feature makes it more mechanical &
hesitation appears automated

4 Hassle in tracking the received money from Tracking becomes easier

multiple acc
Money Senders (MS)

SNo Pain Point Solution

1 Non tech savvy people have to fill in a A request has all the details
lot of details

2 Payment amount is uncertain The MR can enter the

payment amount & thus MS
doesn’t have to worry

3 Does not need to set reminders for MR reminds through app

ownself

4 Tracking the fulfilment on borrowed MS can go to various

money is difficult chatheads & track

Competitors

No such evident feature on either Paytm. Phonepe allows requests from other apps but the

feature is buggy - incomplete user journey. Gpay even allows sending requests to UPI IDs on their

competitors.
 Product Critique Framework

● Product info - What’s the exact Value Prop, How does it work
● Business Objective - How does the product make money
● Users types - Enlist & Prioritise
● User Journey
● Parameters Table

Sample Problem - Critique The Request feature OF GPay

Product info

Value Prop - Allowing users to request & thus remind from the same app for incomplete

payments

How - Using Mobile No & UPI ids a user can request or pay money, a person can either pay

or decline a request received.

Business Objective

No direct monetisation, but helps in getting more transactions done.

User Types

1. Money Receivers
2. Money Senders
User Journey

Money rec/the one who sends req (MR)

Opens > searches for the money sender > opens his chat > presses request > enters amt + notes >

submits

Money sender (MS)

Sees the notif > lands on the chat > clicks pay/decline > completes payment

Parameters For Critiquing ‘Method’

● Utility (Kis hadd tak problem solve ho raha)

● Ease of Use (Kitna simple to use & understand)
Parameter Pros Cons

Utility MR MR

1. Remind in the same app 1. Can’t track the status of requests

2. Ask without any hesitation made
3. Track in the same app 2. Can’t enforce a payment on the
MS
MS 3. Can’t add a followup on the
reminder
1. Send money without worrying
about incorrect payment details MS
2. Send money without worrying
about incorrect MR’s details (UPI 1. MS cannot negotiate or pay in
id, etc.) instalments
2. Can’t ask google pay to snooze a
reminder
3. Can’t decline with a note

Ease of use MR MR

1. Can withdraw a request 1. Can’t edit a request

2. Can’t track all requests in 1 place
3. Isn’t notified about a decline
MS
4. Can’t choose the account he’she
1. The user journey is very short & wants money in
pleasant 5. Can’t set a limit to request
2. Send money without searching fulfilment time
MR’s details (UPI id, etc.)
MS

1. Might get spam requests

2. Can’t track all requests in 1 place
 Product Improvement Framework

Similar to Product Design

Product Design Product Improvement

1. Clarifying questions 1. Improvement Goal

2. Structure (AAAERRR/Specific) & Why
3. Personas 2. Structure
4. Use cases/Prod 3. Personas
Requirements 4. Use cases/Prod Requirements
5. Solutions 5. Solutions
6. Prioritisation 6. Prioritisation
7. Metrics (+Roadmap, 7. Metrics (+Roadmap, optional)
optional) 8. Summary
8. Summary

Sample Problem - Improve The Request feature OF GPay

Goal

Effectiveness of ‘Request’ button

Clarifying Questions

What do you mean by effectiveness - no of fulfilled requests

Request feature is …….Am I correct? - Yes

What do you mean by fulfilled requests? What if there are bogus requests? - Yes, requests have to be legit.

Do we have to focus on one user type related to Request button - No

User Types

Money Sender

Money Receiver

Pain Points

MR:

Can’t track all requests in 1 place

Can’t add multimedia to a request

Can’t choose the account he/she wants money in

Can’t set a limit to request fulfilment time

Can’t edit a request

Can’t track the status of requests made

​Can’t add a followup on the reminder

Can’t enforce a payment on the MS

MS

MS cannot negotiate or pay in instalments

Can’t ask google pay to snooze a reminder

Can’t decline with a note

Might get spam requests

Can’t track all requests in 1 place

Prioritised PP: MR can’t track a request & MS can’t snooze a request

Where Snooze = snooze or decline

Solutions:

a. MR can’t track the status

1. MR can be notified of the request at occasions like:
a. When MS snoozes
b. When MS declines
c. When MS has not seen despite clicking on the notification / opening the GPay / opening
the MR’s chat
i. Pitfall - clicking on the notification / opening the GPay can be misleading - a
section on the homepage to notify about requests

b. MS can’t snooze
1. Enable a functionality to snooze for x hours
a. A clock icon beside ‘Pay’ & ‘Decline’
b. 3 buttons - ‘Pay’, ‘Pay Later’, ‘Decline’
i. Remind after x hours / x days
1. Soln: 4 hours, 1 week, 2 weeks
2. Soln: MR can enter max deadline (which won’t be shown to the MS on the
1st request)
3. Soln:
ii. Remind on x date
1. Choose payment mode
a. Add note (optional)
iii. Remind later
UJ: Request received -> Pay later -> Hours/Date + Payment Mode + Add note -> Snoozed request

Pitfall - 1. No limit on how many times he can snooze -> max 2 times

2. He will click on Pay Later more than pay ->

a. Pay button should be more highlighted
b. Snooze is at the top corner of the message
c. Better rewards if you pay through ‘Pay’ rather than ‘Pay Later’

2. Don’t let snooze from requests, but add a payments page where all pending requests will be

notified about every saturday

 DD NEW SESSION

AGENDA
1. GTM Strategy
2. Tech Part 2
a. Web Architecture - wholesome view of the srchiteture.
b. Protocols
c. Data Transfer mechanism
d. Databases

GTM Strategy

A. Positioning
● Define the problem you are solving
● Make User Types
● Note down the pain point your app solves
● Check competitors offerings & pricing

B. Launching
● Find Early adopters - Find niche users (Talk to those users to find their Pain points + Educate them)
● Where are they? - Find channels
● Decide communication & execution
■ Who
■ When
■ How
■ What

Devise a GTM strategy for Spotify Greenroom in India

Clarification Questions

What is spotify greenroom? it is an audio based social networking site - like ClubHouse
What type of users do we want to have? No barrier on the type of the user
Platform? Android & iOS
Same app as spotify or a different app/ different app
Functionality on the app? Create & join audio streams based on your interest, you can have live audio
discussions, plus an option to record audio, no recorded audio like music available
Ii it subscription based? No
I see CH/Twitter Spaces as the major competitors, do we have any other competitor? Leher (India specific)
Goal? We see a huge potential in India, huge audience, so customer acquisition is important and not
revenue.
Monetary incentive for creators? Not decided, free to decide
Room limit/ 1000

Clubhouse - group/channels - open/cosed - open/closed audio sessions - live audio group chat
Positioning
1. Problem being solved
Connecting people using live audio conversations.

2. Make user types

a. Speakers - finding audience for live audio conversations/sessions
b. Listeners - joining groups based on interest and listening to ideas from experts.

3. Pain Points Your App Solves

a. Speakers
i. Solves the problem of getting audience for live delivery & feedback.
ii. Solves the problem of trvalling to places for reaching out to audiences.
iii. Solves the problem of finding like minded collaborators
iv. Redistribution of performed content on platforms like spotify podcasts, google/amazon.

b. Listeners
i. Solves the pain point of discovering speakers ona topic of interest.
ii. Solves the pain point of finding/creating groups for discussion with like people
iii. Solves the pain point of trvalling for attending sessions

4. Competitors comparison
a. User Base - Listeners
i. Existing players like CH have a huge user base (first mover)
ii. Spotify also had a huge loyal user base of music & podcast listeners.

b. User Base - Speakers

i. Lot of creators already on CH
ii. Spotify can convert its podcast creators to greenroom Speakers.

c. Brand Recognition
i. Spotify already has a good brand image & recognition amongst a large audience.
ii. CH has a good awareness, but not as good as spotify.

d. Ecosystem
i. Greenroom-podcast integration for greater reach for speakers & greater accessibility
for listeners. Album marketing using greenroom, using spotify music data to provide
better recommendations on greenroom
ii. CH has no such ecosystem to leverage

e. Capital Availability
i. Deep pockets

Launching

1. Find Early adopters, Where are they?

How many speakers are we targeting - we want 1 Lakh listeners in 1 months. How many speakers do we need?
 a. Speakers
i. New Speakers, on social media (novelists, journalists, activists)- yt, insta, fb, linkedin,
ii. Existing Speakers (on CH, celebs)- yt, insta, fb, linkedin, CH etc
iii. Spotify Podcast artists - Spotify (new & growing artists can be a better bet)
iv. Speakers from orgs spotify sponsors - Offline Channel (Marketing Team)
v. Radio Jockeys - FM, Apple Music, gaana, jio saavn.

b. Listeners
i. Spotify podcast listeners - Spotify
ii. Spotify music listeners - Spotify
iii. Radio listeners (small villages, old people, travllers, drivers)
1. Old age - Fm Radio
2. Travllers - travels apps - ixigo, redbus, ola, uber, redtaxi, tour operators
iv. Followers of speakers on insta & linkedin - insta linkedin

3. Decide communication & execution

a. Who
b. When
c. How
d. What

Gems can be used for spotify premium

Channel Who, When, How, What

Podcast listeners Who

Avg listening time per day >30 mins
Followed creators > 5

When
1 week before launch, continued till 5 weeks.

Where (location)
Greenroom Banner On Podcast home
Greenroom Ad during podcasts

How (Communication)
1. Free subscription to spotify music on joining
greenroom.
2. Early access to podcasts on greenroom
3. Preferential access to live podcast while they
are recorded.

Other eg:
Devise a GTM strategy for Youtube shorts
Devise a GTM strategy for a Pet buying/selling app
 Web Architecture

Client-Server Model

The most essential feature of a web architecture is the client server model which is the underlying
architecture for almost all websites & web apps.

Client

The client holds our user interface. The user interface is the presentation part of the application. It’s written
in Html, JavaScript, CSS and is responsible for the look & feel of the application.

Types of Clients

● Thin Client - Holds only UI

● Thick Client - Holds Ui + Business Logic

Server

The primary task of a web server is to receive the requests from the client & provide the response after
executing the business logic based on the request parameters received from the client.
The server configuration & the type can differ depending on the use case.

● For running a backend application code written in Java - Apache Tomcat or Jetty.
● For hosting websites - Apache HTTP Server.

Protocols
HTTP

In the client-server model we discussed, where a client sends a request to the server and the server reverts
back with a response.

Keeping track of these interactions isn’t easy.

So a set of rules are needed to set the expectation on both ends.

HTTP protocol comes to the rescue here and helps in setting the rules for the interaction i.e syntax,
semantics & timings.

HTTP vs TCP/IP

● HTTP is a text based protocol & messages exchanged are text having 2 parts header & body

● HTTP is just a Application Layer Protocol & just sets the rules for communication and doesn’t
facilitate communication

● TCP/IP is the protocol responsible from transmission of data in form of small packets which
includes both client request & server response

HTTP vs HTTPS

● Requests & responses are unencrypted when HTTP is used, whereas they are encrypted when HTTPS
is used.

● Encryption is done via a security protocol called the SSL protocol.

● Once a connection is established via port 443 the server & client do a TLS handshake where they
exchange secret session keys for encryption & decryption of messages
Deep Diving Into HTTP Requests & Responses

If you have the firebug extension you can visit the network section after opening the extension and click on
any website path to see the details of the requests.

This is what you will see.

HTTP Request

Header
Sample Header

User-Agent:Mozilla/5.0 (Macintosh; Intel Mac OS X 10_10_5) AppleWebKit/537.36

(KHTML, like Gecko) Chrome/47.0.2526.73 Safari/537.36
Accept-Encoding:gzip, deflate, sdch
Accept-Language:en-US,en;q=0.8
Host:github.com
Cookie:_octo=GH1.1.491617779.1446477115; logged_in=yes; dotcom_user=iam-peekay;
_gh_sess=somethingFakesomething
FakesomethingFakesomethingFakesomethingFakesomethingFakesomethingFakesomethingFa
ke; user_session=FakesomethingFake somethingFakesomethingFakesomethingFake;
_ga=9389479283749823749; tz=America%2FLos_Angeles_

Details Of The Header

User-Agent Details about the browser and the system OS

Accept-Encoding The encoding the browser is ready to accept
Accept-Language The language of the requested page
Host The name of the host server
Cookie Data for the same server stored from past visits

HTTP Response

Header
Sample Header

Request URL:https://github.com/
Request Method:GET
Status Code:200 OK
Remote Address:192.30.252.129:443
Content-Encoding:gzip
Content-Type:text/HTML; charset=utf-8

Details Of The Header

Request URl - URL the client requested for.

Request Method The http request method
Status Code A code indicating the status of client request
Remote Address The IP address of the server + port details
Content-Encoding Name of the encoding technique used
Content-Type The type & subtype of the response body
Body

Depending on the type of request method the body of the response may vary and can be empty as
well. But for this case GET method was used, the response will contain the html code for the page
requested

HTTP Methods

Method Status Code Common Status Request

Code Idempotent

GET 200 (OK), 404 (NOT FOUND) 500 (INTERNAL Yes

SERVER ERROR)
POST (Create 201 (CREATED) along with a location header if the processing No
a resource) with the link to the newly created resource fails and the server
errors out
PUT (Update 200 (OK), 404 (NOT FOUND) Yes
a resource)

DELETE 200 (OK), 404 (NOT FOUND) Yes

THese Http methods help the server understand the type of operation to be performed on the
resource (every http request points to a particular resource. This resource can be a normal html
file a css file)
 Mechanism Of Data Transfer Between Client & Server

Ever Thought Of This?

How are scores updated dynamically on websites like cricbuzz ?

HTTP PULL
The client pulls data from the server, sends requests, receives responses, keeps on sending the response
in search for an update.

Not an ideal way to transfer data, waste of resources, repeated pulls may bring down the server.

HTTP PUSH
In this mechanism, the client sends the request for particular information to the server, just for the first time,
& after that the server keeps pushing the new updates to the client whenever they are available.

This saves a lot of network bandwidth & cuts down the load on the server by notches.

Short Polling
● Sending manual requests - HTTP GET method
● Sending automated requests at fixed time intervals - Short Polling
● Client sends a request waits for the responses (either the same old data or updated)
● Waits for a while & sends another request

Long Polling

● Clients sends a request

● Servers hold the request until an update is available.
● Clients send another request as soon as it receives the responses.

Advantages
● Lesser requests sent as compared to Short Polling
Disadvantages
● The server side implementation is slightly complex
● Sometimes holding back the requests takes a lot of memory
● Keeping a track of order of multiple requests becomes difficult
Server Sent Events
● A protocol for one way communication from server to push updates as & when they are available.

Advantages
● No need to establish connection again & again.
● Simple straightforward implementation

Disadvantages
● One unidirectional communication occurs, client cannot send any particular data,
● Limitation to no. of SSE connections over a browser.

Web Sockets
.
● Complex to implement, but faster than HTTP
● Real time simultaneous bidirectional flow of data possible.
● No need to find & establish connection after every message
 DD NEW SESSION

AGENDA

1. Market entry - Known market/geography

Known Market/Geography

eg:x wants to enter y. Do you think x should?

Clarifying questions

● Terms in the Problem Statement

● Goal
● Time frame
● Specific Geography/Market

Product/Market details

● What is the product

● How does it work
● USP
● Anything complementary with it?
● Legal Barriers

Customer of new market/geography

● Types of customers
● Why do they use this product
● Distribution channel preferred
● Pricing they like
● Payments mode preferred
● Pain points they face in existing products
● Market trends, including changing customer preferences and regulatory changes

Company fit (Relate with existing market/geography)

● Expertise
● Organisational structure
● Economies of scale
● Our Access to distribution channels
● Relationships with suppliers

Competitive environment

● Number of competitors
● Market shares
● Competitors’ resources, including financial, employees, distribution channels, suppliers and partners
 ● Competitor’s unique competencies, including differentiated products
● Penetrated market

Profit/Growth Calculations

● Market size (Using ans from Cust)

● Market share (Using ans from Competition)
● Profit margins (Using ans from Product, Company)
● Market growth (Using ans from Cust)

Profit = Size*Share*Profit Margin

Question - Should Ola enter hyperlocal delivery?

Clarify

What is hyperlocal delivery? Delivering items in a small diameter or area (5-6 km) and quickly.

Any specific geography? Anywhere in India, you deem suitable

Goal of Entry? Ola sees an opportunity to enter a big market and also an opportunity to create synergy with the
existing offering.

Are we focusing on users/revenue/profit now? Revenue

Users = users
Revenue = users * ARPU
Profit = users * ARPU * profit Margin

Time of Entry? Planning to launch in 8 months

Product/Market

What will ola deliver? Grocery & everything can be delivered, but superfast deliveries are not possible.

How will the product work? Controlled marketplace model - sellers onboarded with their inventory, users

searches for a product, ola matches it with a store based on its driver’s (Ola Bike) availability in vicinity, item

delivered and commission charged

 Marketplace model

Pros - lots of choices, inventory management unrequired (wastage losses minimised, capital

requirement reduced), warehouses not reqd, labour not reqd.

Cons - lower margins,dependency on sellers (cartelisation, quality assurance difficult), cannot control

seller discounts, cannot control delivery time.

Dark Store model-

Pros - quality control, delivery time can be optimised, higher margins, control on discounts, returns

can be easily processed, better control on availability of items.

Cons - high investment (warehouse, inventory, labour, wastage), centralisers

How will we manage inventory? Onboarding sellers

Will we have separate riders to deliver items? No we will map riders based on their route for dropping or picking

someone up

Will we have a different App? Same app as users are already there

Will we also deliver parcels? No as of now

Customer

Why will customers use our app?

Width of items - marketplace - good width

Cost - lower cost (as compared to competitors)
Delivery availability - limited (marketplace)
Speed of delivery - weak
Quality - moderate

Reason for ordering - lower delivery costs & wider availability of products

Who will be our customers? B2C orders only, usually lower middle & middle class people preferring cost over
speed

Age of customers? 15-40

What demography shall we consider? In areas where ola bikes operate, mostly Tier 1 & 2 cities

Payment method? All methods UPI/ Wallets?cards

 Company

Are we building from scratch or acquiring someone? Open to acquisition (should be avoided)

Will the current brand position hurt our new entry? There some chances if the matching is not done properly

Competitors

I suppose Dunzo, Swiggy & Blinkit are the main competitors? Yes

What is the market share? Dunzo 50% | Blinkit 40 % | Swiggy 5% | Other 5%

Competitors
Dunzo
Model - Marketplace
Geography - Tier 1 & 2
Product delivered - groceries 7 everyday products, parcels, medicines, food
Delivery Fee - 10-15 rs for half od the stores & free for half
Blinkit
Model - Dark Store
Geography - Tier 1 & 2
Product delivered - groceries, veggies, everyday items
Delivery Fee - store dependent like dunzo | freeabove 500
Swiggy
Model - Dark Store
Geography - Tier 1 & 2
Product delivered - groceries, veggies, & everyday items (limited choices) & parcels
Delivery Fee - free above 100 | 15/35 rs based on delivery speed

Market penetration - 40%

Calculations:

Market Size

Population tier1 - 100M (50 cities * 20 lakhs)

BPL neglect - 70M (bpl population = 30% assumed)

Households - 18M (avg household size = 4)

Types of households:

Low - 40% * 5K monthly * 10% * 18 = 3.6 B

Middle - 50% * 10K monthly * 50% * 18 = 45 B

High - 10% - 20K monthly * 25% * 18 = 10 B

Total = 58 B (monthly revenue of online groceries) * 12 months = 700 B spent on grocery & daily needs online

Rev for hyperlocal = 20% = 140 B

Penetrated market = 40% of 140 B = 56 B

Total revenue per year in online hyperlocal market is 56 B

Suppose 1% in 1 year = 560 M

1 B rev ina year - we should not enter

300 M rev in a year - yes we should enter
 DD NEW SESSION

AGENDA
1. Viewing v/s Hosting
2. Hosting a site Online
a. Web Hosting Providers
b. Types of Web Hosting
c. Architecture of Remotely Hosted Sites
3. Databases

We now know how users can fetch data from any site hosted on a server connected to the internet

URL to DNS via ISP to IP Address to TCP/IP connection to GET HTTP Request to Response with Header & Body
having HTML CSS & JS files to being rendered by Web Browser

Viewing vs Hosting

As a User you need As a Provider you need

1. Device 1. Computing & Storage

2. Web Browser 1. Website code & files.
3. URL 2. Domain Name
4. Internet 3. Internet

As a User you need As a Provider you need

Device mobile | Laptop | Desktop Computing & Storage Owned | AWS, GCP| GoDaddy
Web Browser Chrome | Firefox | Safari Website code & files Likho Khud Se!
URL Type | Paste | Search Engine Domain Name GoDaddy | Hostinger
Internet Jio | VI | Airtel Internet Required if 1 is self owned
 Hosting a Site Online

1. Building a website
Developing the website either from scratch/template or via No/low code tools

2. Purchasing a Domain Name

Purchasing a domain name from any web hosting provider.

3. Acquiring Resources for Hosting the Site

Subscribing to web hosting providers or cloud providers based on website requirements &
cost

4. Connecting Domain to Name Server

To map your domain to where your files lie, linking the purchased domain to web host’s name
server.

5. Uploading files on the Server

Transferring files from local to the web host’s server

Web Hosting Service Providers

Companies which help you host your website on remote servers connected to the internet & provide tools
& services to manage it efficiently.

Famous Providers

Services Provided

1. Providing Processor (CPU), Memory (RAM), Storage (Disk Space).

2. No/low code Website Builders
3. Dashboard for managing files.
4. Firewall Protection
5. SSL Certification
6. Data Backup
 Web Hosting Types

1. Shared Hosting
A setup wherein multiple sites are hosted on a single machine and all of them use the same
shared resources.

2. VPS Hosting
A setup where multiple sites are hosted on the same physical machine, but each of them run
on a separate VM and have their own OS & resources.

3. Dedicated Server Hosting

A setup where a single site is hosted on a physical machine.

4. Cloud Hosting
Cloud Hosting is similar to PVs but the site is hosted on not just a single physical machine
but ona cluster of such machines.

Another Analogy
Shared Hosting - Multiple User Profiles on the same OS
VPS Hosting - Multiple OSs/VMs on the same machine.
Dedicated Hosting - Single OS with Single User Profile on a machine
 Architecture of Remotely Hosted Sites

Layers
Layers in a software architecture refers to logical separation of codes, where each piece performs a
specific function.

1. Presentation Layer - This layer is responsible for displaying the user interface and managing user
interaction.

2. Application Layer - It has all the business logic. It's the bridge between the Presentation Layer &
the Data Layer.

3. Data Layer - This layer is responsible for storing the data.

Tiers
Separation based on distinct physical hardware

1. 1-Tier - The entire product (UI + Business Logic + Database) lies on a single physical system.

2. 2-Tier - With 2 physical devices, the client holds the UI & Business Logic & server has the database

3. 3-Tier - The UI, Business Logic & the Database all lie on different machines.

4. N-Tier - Products where the UI, Business Logic & the Database lie on more than 3 machines.
 Databases

A database also known as a DBMS is an application that manages access to a physical data store. Uses of
database.
1. To store data
2. To update or delete data.
3. To fetch data when required

Types Of Data

Structured Data
● Has a defined structure. (5 fields, name, email, phone, class, division)
● No need to perform any data preparation operations.
● SQl is used to deal with structured data.

Unstructured Data
● Has no defined structure.
● Data is usually heterogeneous (all possible formats)
● Data preparation needs to be done.
● The data is usually picked from multiple sources.

Semi-structured Data
● Combination of structured & unstructured data.
● Usually stored in data transport formats like XML 0or JSON.

Relational Database

● Data stored in various tables are interrelated.

● Relation types
○ One to one
○ One to many
○ Many to many
○ Many to one
● Normalised data saving (at one single place & in the simplest form)
● Data consistency (single point to be modified no inconsistency)
● Relational Database Transactions are ACID in nature.
○ Atomicity - Each transaction is all or nothing, no transaction is partially executed.
○ Consistency - Any transaction will bring the database from one valid state to another
○ Isolation - Executing transactions concurrently has the same results as if the transactions
were executed serially
○ Durability - Once a transaction has been committed, it will remain so

Sample Relational Database

Let’s take an example of digitising the medical records of patients in a hospital

Details you need to hold

1. Patients personal info
2. All his visits & checkup data
3. Doctor’s info

Tables you to need to make

1. Patients
2. Doctors
3. Visits

Querying Data From Database

Select <column name>

from <table name 1>
join/union <table name 2>
on <column name >
Where <condition>
Group by <condition>
Order by <condition>

Case - Suppose we want to see the no of unique patients that came in FEB
Solution -Select count(distinct(Patient)) from Visits where date between ‘2022-02-01’ and ‘2022-02-28’

Case - identify the doctor with most no of patients attended

Solution - select a.id, count(distinct b.MRN) as patient_count
from doctors a left join patients b on a.id = b.primary
Group by a.id
Order by count(distinct b.MRN) desc;
When Should We Use a Relational Database

● Where there are many relations between the data - social media sites
● Where data needs to be highly consistent & accurate - banking, finance, trading, medical records
● Where the data needs to follow a predetermined schema

Disadvantages of Relational Database

● Difficult to scale horizontally ,due to cross node relationships
● For data that varies a lot and doesn’t conform to a fixed schema, optimisation becomes difficult

Popular Relational Databases

● MySQL
● PostgreSQL
● Microsoft SQL Server
● MariaDB
● Amazon Aurora
● Google Cloud SQL

NoSQL Database

● Built for high-frequency read and writes

● Are developer friendly and easier to design
● No strict schema required to keep data atomic.
● Increased chances of data inconsistency.
● Do not support ACID transactions

When To Use NoSQL Database & Why?

● High frequency read & write - Social applications like Twitter, Real-time sports apps , Online massive
multiplayer games

Common types of Non-relational databases

● Key value Datastore
○ The data stored is usually in key-value pairs, where key == question & value == answer.
○ Usually used for storing data which involves high frequency read & write operations.
○ The data is stored in Primary memory as compared to secondary as in case of RDs.
○ The data is usually textual as the primary storage is lesser.

● Graph Database
○ Similar to Relational Database
○ Can hold highly interlinked data
○ Queries need no joins as the data is not stored in tables.
○ Social media - friends Relations

● Document Store
○ JSON files stored with a key identifier.
○ Stores all information about a relevant topic.
○ The schema can vary making changes easier.
 ○ New Agencies storing news articles.

● Search Engine Database

○ Text search option on a very large amount of unstructured data.
○ Data can come from various sources
○ Fuzzy search also possible

● Time Series Database

○ Data needs to be strictly structured by date.
○ Can be easily partitioned by time range

Popular NoSQL Databases

● MongoDB
● Redis
● Cassandra

Polyglot Persistence

Polyglot persistence means using several different persistence technologies to fulfil different persistence
requirements in an application.

Multi-Model Databases

With the advent of multi-model databases, we have the ability to use different data models in a single database
system.
Popular Multi-Model Databases

● ArangoDB
● CosmosDB
● OrientDB
● Couchbase

Scalability

Scalability means the ability of the application to handle & withstand increased workload without sacrificing
the latency.

Latency
Latency is the amount of time a system takes to respond to a user request. The time taken between the
action you perform on a website and the moment you see the response of the website on your device.

Network Latency - Network Latency is the amount of time that the network takes for sending a data
packet from point A to point B.

Application Latency - Application Latency is the amount of time the application takes to process a user
request.
How to decrease latency in a resto

1. Simultaneous cooking.
2. Division of tasks between cooks.
3. Pre cook items
4. Pre-ordering food.
5. Hiring efficient cooks & waiters.

Ways To Decrease Latency

Network Latency
Using better internet connection

Application Latency
Delaying the requests for unpaid users
Using better hardware (more efficient cooks ya more no. of cooks)
Using distributed systems (division of tasks)
Caching on your device (pre ordering)
Caching servers (pre cooking)

Network Latency - Using CDNs (Content Delivery Networks) deploying servers all across the globe.
Application Latency - running load & stress tests to identify bottlenecks in the system & fixing them

Types Of Scalability
● Vertical Scaling
● Horizontal Scaling

Vertical Scaling
Increasing the power of the hardware to handle more load, replacing the system with an alternate with a
higher computational power.

● Requires no code refactoring or any complex configuration.

● But there is a limit to how powerful a device can be.
● Scaling up gets increasingly costlier with increasing size.
● Issue of a single point of failure exists.
● Data consistency is high.

Horizontal Scaling

Horizontal scaling means adding more no of hardware to the existing hardware pool.

● Requires code refactoring or any complex configuration.

● No limit on the no of devices can be added.
● System is resilient, stays active even if one or more nodes goes down.
● Data consistency is lower as compared to vertically scaled systems

Why Do We need to Change The Code For Horizontal Scaling

The code should be stateless, none of the instances of any class should be static, as static instances hold
application data and if the server goes down the system will be in a state of inconsistency.

Functions are stateless and hence functional programming is becoming famous with the advent of
distributed systems.
Bottlenecks To Scaling Up

● Having a monolith structure for the database.

● Using Sequential processes where asynchronous processes are possible.
● Using too less or too many load balancers.
● Not using the caching service properly.
● Adding business logic to the database.
● Not picking the right database.
 DD NEW SESSION

AGENDA

1. Market entry - Unknown market/geography

Unknown market/geography - eg: where should x enter?

New Market/product or service

● Clarifying questions
a. Terms in the PS
b. Ask Goals
c. Limitations / Restrictions
● List existing products
● List Already successful products/markets
● List existing company strengths
● List new adjacent markets
● Analyse each adjacent market based on PCC
Product/Market details
○ What is the product
○ How does it work
○ USP
○ Anything complementary with it?
○ Legal Barriers

Customer of new market/geography

○ Types of customers
○ Why do they use this product
○ Distribution channel preferred
○ Pricing they like
○ Payments mode preferred
○ Pain points they face in existing products
○ Market trends, including changing customer preferences and regulatory changes

Competitive environment
● Number of competitors
● Market shares
● Competitors’ resources, including financial, employees, distribution channels, suppliers and partners
● Competitor’s unique competencies, including differentiated products
● Penetrated market

● Choose 1 based on Goal

Question) A market Amazon should enter next.

Clarifying Question
Are we targetig any particular geography? Not as such, but India is preferable.
Are we focusing on any particular market/product? Nothing
Amount of funding available for the project? Bahut hai
Timeline? 12 months, flexible as per requirements
Mode of entry? Haven’t planned it yet, depends on the product chosen and existing players for the same
Goal of market entry? Entering new markets with high growth opportunities in future
Ultimate goal? its user acquisition & retention
Competitors? Product not yet decided
Any specific age group? No

Current products
1. AWS - cloud provider
2. Amazon Fire OS - OS
3. Amazon App Store - App Marketplace
4. Alexa - Voice Assistant
5. Amazon Pay - Payments
6. Prime - OTT
7. Music - music streaming
8. Amazon Shopping - Ecommerce
9. Fresh - Grocery Delivery
10. Pharmacy - medicine delivery
11. Audible & Kindle - ebooks & audiobooks
12. Fire TV Stick, Alexa Home - IoT Devices
13. Kndle device - Hardware
14. Twitch - Game streaming
15. IMDB & Goodreads - books & movie reviews

Successful Markets (market share)

1. AWS - cloud provider (more profitable + Azure/GCP not that good)
2. Amazon Shopping - Ecommerce (a good fight by flipkart, myntra etc)
3. Alexa - Voice Assistant
4. Prime - OTT
5. Audible & Kindle - ebooks & audiobooks
6. Kndle device - Hardware
7. IMDB & Goodreads - books & movie reviews
Company
Strengths
1. Deep pockets
2. Brand recognition & loyalty.
3. Logistics expertise
4. Technology
5. Customer Data
6. Exposure to multiple industries

Adjacent markets - India

1. AWS - cloud provider - cloud Storage in B2C, Photos App, CMS like Wordpress, No code
(Honeycode), cyber security
2. Alexa - Voice Assistant - fully automated homes
3. Prime - OTT - sports/news
4. Amazon Shopping - Ecommerce - analytics management software, b2b ecommerce, fashion
ecommerce, Amazon consulting, neo bank, investment arm
5. Audible & Kindle - ebooks & audiobooks - e-papers
6. IMDB & Goodreads - books & movie reviews -

Possibilities to enter (P = product C = competitor C= consumer)

1. Cloud Storage in B2C (large audience)
Pros
1. large audience - almost all users have cloud storage accounts.
2. Data - stored data can be leveraged

Cons
1. Established Competitors - players like Google & Microsoft already there.
2. Lower margins - lower margins in b2c as compared to b2b
2. Photos App (same as above)

3. fully automated homes

Product - possibility of new products | High potential in future
Consumer - the existing base is very low though | but they have a high capacity to pay
Competition - no established player

4. sports/news
 Product - not much differentiation
Consumer - huge user base
Competition - established players like Hotstar, Sony lIv | (OTT + TV experience not there)

5. Amazon consulting
Product - data driven consulting on product development, product positioning & marketing
Consumer - B2B clients - All D2C brands | high margin business
Competition - Payment processors like VISA Mastercard
6. E-papers
7. analytics management software
8. b2b ecommerce
9. fashion ecommerce
10. neo bank
11. investment arm
12. CMS like Wordpress
13. Cyber security

Markets To Enter
1 Smart Homes 2 Consulting 3 Neo Bank

New Geography

1. Clarifying questions
a. Ask Goals
b. Limitations./ Restrictions
2. Existing Geographies
3. Existing company Strengths
4. List Potential Geographies
5. Analyse each potential geography based on GCC
a. Geography - Legal Barriers, Supplier Power, Market Size & Growth potential
b. Customer - propensity, willingness, pricing, payment modes, pain points
c. Competitors - number, size, employee, finances, distribution channels
6. Choose 1 based on Goal
Question - Swiggy wants to go international.
Clarifying questions
What product line are we expanding with? Food delivery only
Any specific geography? Asian Markets preferable
Goal? Why to enter? EMerging economies are a good way to expand their user base & possibility of revenue &
profits in future
Mode of Entry? Acquisition will be preferred
Budget Constraints? Out of scope of the problem
Changes in model, if any? Geogra[hy dependent, but we’ll try to use the same model.

Existing geographies
1. Geography - India Tier 1 & 2
2. Model - AN app as marketplace, restaurants partners as entities on marketplace & delivery
personnels as last mile connectors
3. Revenue Model - commission Based
4. Penetration - 70%
5. Market Share - 50%

Current strengths
1. Tech infra for making online food ordering & delivery seamless is already there.
2. Expertise in handling stakeholders (consumers, delivery partners, restaurant owners)
3. Have hyperlocal verticals as well
4. Good Brand Image
5. Experience of working in aprice sensitive market

Potential Geographies

1. Bnagladesh
2. Indonesia
3. Taiwan
4. Singapore
5. Thailand
6. UAE

Country Populati Internet speed Incom Economi Competit Penetrati Rank of Rank of
on (in (qualitative) e c Growth ors on (%) Populatio entry
M) i (rank) n*penetra
tion

Indonesi 270 50 - 3 3 3 60 *270 = 1 Okayis

a 160 internet,
income &
Its a
highly
 competitiv
e market

Banglade 170 32 - poor 4 4 1 (local) 20 * = 2

sh 34M

Malaysia 30 92 - 2 2 2 40 * 30 = 3
12M

Nepal 30 25 - 5 5 1 Lower
income &
unsuitable
terrain

UAE 10 123 1 1 2 50 *10 = 4 Good

5M income,
internet
low
population

Bhutan 1 15 6 6 0 0
 DOREMON DEN NEW SESSION
AGENDA
1. Pricing Model
2. Pricing Question

Pricing Strategies

Models Product Examples Why

Free Trial Prime/Netflix To lure or tempt customers

Freemium Spotify, Canva Not ad-based, because not all services are free

Pricing Models

Models Product Examples Why & plan examples

Ad based Free Facebook, Youtube Keeping it free & accessible

Razor Blade Gillette, Xbox, JioPhone

A la carte Gaming, LinkedIn, Tinder Remain in spotlight/See who’s

interested/Passport

Tiered Netflix, Spotify, Bumble, Based on device, video quality, etc.

Commission Airbnb, Ola, Zomato

based

Subscription Prime

Tiered vs A la carte
Family plan (tier 1)- 6 (a,b,c,d,e,f)
Tier 2 - a,b,c,
Tier 3 - a, b
Tier 4 - b

Linkedin plans:
Learning
Career
Business
Sales
Recruiter
So, it is a la carte.
Eg - A new OTT platform is coming from YRF productions. Assume - subscription model

How to solve the question?

● Clarifying questions
● Different Pricing Techniques
1. Competitor based (direct competitors, having similar offering & same TG, cannot take competitors
that differ highly)
2. Cost based (pricing based on the cost that you incur on the product & profit margins you wish to
have.)
3. Value based (pricing of current offering + extra price user is willing to pay for the additional offering
you provide)
● Compare & Suggest a Range

Value based pricing

1 Make customer personas:
You can have multiple buyer personas for a single market segment.

Answer these:
1. What is the single most valuable feature that the buyer loves?
2. What is the most appropriate next best alternative that the buyer prefers?
3. Which product(s) the buyer loves and why?
4. The biggest goal in terms of the product- you’re offering?
5. What are the biggest fears in terms of product and price?
6. The average monthly spending?
7. Is the buyer cost-sensitive?
8. What is the most realistic price the buyer is willing to pay?

2. Talk to your customers:

You have to collect data on the following variables:

The product that they buy now. This is your next best alternative. Your price should be based on this product.
The feature that they miss the most.
Perceived value of the feature that you’re offering in your product.
The maximum price they’re willing to pay for the ‘feature’.

3. Evaluate the gap:

Evaluate the gap in terms of biggest challenges that the customers are facing — relating to price, product, features
offered currently.

4. Evaluate the differentiated value:

What does your proposed feature mean to the buyers?

Are they really interested in the feature?
Do they really need it in the first place?

5. Assign a Rupee value to the differentiated value

Eg - TV brand - LG
Rn Samsung sells a 50 inch TV at 40K
LG launching 50 inch TV + screen that lasts 3 yrs extra - 5K

Value based price = Price of best alternative + Value of differentiated feature

40 + 5K = 45K

Cost based - x - 30k

Compet based - y - 40k
Value based - z - 45k
Final 42k

Price confirmation - Demand vs Price curve

Q: How will you price YRF (yash raj films) OTT? Assume - Subscription model.

- Content that will be available? All new yrf produced & distributed movies & all old ones as well.
- Are the movies on other OTT platforms? Yes. But won’t be available post our product launch
- Will the movies be pulled from other platforms? Yes
Any other production partners? No
- Early release? YRF OTT only release within 10 days of theatrical release
- No. of movies released? 8-10/yr
- Are we open to other OTT partnerships - No. Only yrf films/shows
- Device specific models - No. Planning to launch 1 plan only. Will launch this product on Mobiles to start
with
- Model can include ads? - No (we can give a free trial if you want)
Target theatre or non-theatre user base - Both. Majorly YRF fans (be it from any platform)
- Will any content be released only on OTT - No
- Is there an option to share subscription - No
- Is there a jukebox (other features) - Not right now
- Are we planning to price higher than competitors - What do you think?
- Can we partner with movie theatres (based on choices to swap) - Not right now
- Subscription duration - Annual
- Will it be avail only for Indians - Worldwide
Will users be pre informed about no of movies to be released per year - Yes
- What age group watches YRF movies - YRF is an old publication house & have been creating content
catering from kids to elderlies. So all.
- Will old movies be available in 4K quality - HD yes (if we already have for those specific movies)
- What languages will we be providing - All (90% Hindi)

- Why are we launching - OTT industry is booming & we feel we are well positioned to enter this now.
Planning to get 1cr users (at least product acquired) by the end of year
Competitor Based Pricing

Prime video - 1000

Netflix - 2500
Youtube premium - 1500
SonyLive - 400
Hotstar - 400
Eros now - 800
Zee5 - 500
Voot - 300
Altbalaji - 300

Hotstar & SonyLiv - people buy majorly for sports content

Zee5 & Eros Now our closest competitors

Competitor = (500+800)/2 = 650 per year

Cost Based Pricing

Assume - Movie making cost not involved here
Tech cost (maintaining servers for 1yr, employees & operational) - 60cr
Marketing cost (including all types of acquisition) - 50 cr
Extra costs - 10cr
Commission & royalties (opportunity cost) - 2.5cr per movie * 200 movie = 500cr
Total cost = 620cr
Effective users for yr 1 = 50Lakhs
Cost per subscription = 1240 Rs per year

Value based Pricing

Value = price from best alternative + differentiated feature price
= 650 + x

X = No of movies launched per year * ticket price * % movies watched per year * Vibe reduction factor
= 10 * 150 * 0.33 * 0.45
= 222.75

Value = 650 + 222.75

= 872.75

Comparison

Comp - 650
Cost - 1240
Value - 875

Price bracket - 650 to 875 per year

 😍
Thank You!