Brainscape's Knowledge GenomeTM

Browse over 1 million classes created by top students, professors, publishers, and experts.


See full index

Business > Growth Series > Flashcards

Growth Series Flashcards

1
Q

Retention + Engagement

A
  1. Retention separates top 1%
  2. Retention + Engagement drive Acquisition and Monetization

Retention: breadth, were they active or not?
Engagement: depth, how active were they?

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Retention + Engagement - silent killer

A
  1. Retention gets de-prioritized due to a long time horizon
  2. Poor retention is easy to cover up and miss
  3. Breadth but no depth
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Retention definition

A

How many users or customers remained active within a defined time period after signing up?

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Retention is the output

A

Retention = Activation, Engagement and Resurrection
Activation: Establishing the habit
Engagement: Habit is established
Resurrection: Returning dormant user to an engaged state

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

How to improve retention as the output?

A

Center our approach
1. Use cases (qualitative view)
2. Measuring + analyzing (quant view)
3. Engagement strategies, explore engagement strategies to build the habit
4. Activation strategies, how to build the habit with first-time users
5. Resurrection strategies, how to regain those that once had the habit

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Retention + Engagement - define use case

A
  1. Problem, what is the definition of the problem in your users’ words
  2. Persona, who has the problem?
  3. Why, what is the core reason the user chooses your product to solve the problem?
  4. Alternative, what is their alternative to solving the problem?
  5. Frequency, how often do they encounter the problem?

NOTE: Most products evolve to serve multiple use cases

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Retention + Engagement - Frequency Spectrum

A

Habit zone: Daily, weekly and monthly
Forgettable zone: Yearly and years+

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Common mistakes for defining incorrect retention metric

A
  1. Not aligning with natural frequency
  2. Combining actions, teams would always choose the easier action to move.
  3. Optimizing just for revenue, revenue is an output of usage, and usage is the input.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Define retention metric

A

Frequency + core behavior + who

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Retention curve pattern

A
  1. Trend to 0 (bad)
  2. Flattish (ok)
  3. Flat (good)
  4. Smile (great)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Basics of building habit

A

Cue -> Routine -> Reward

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

2 reinforce organic habit loop strategies

A

Manufactured: Loops with triggers that are created and manufactured by your product or marketing
Environment: Loops with triggers that we insert into places and products our users touch when the problem occurs

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Activation’s three moments

A
  1. Signed up, The user has joined the product
  2. Setup moment, set of actions that set the user up to go through the activation experience
  3. Aha moment, the moment the user experiences the value proposition for the first time
  4. Habit moment, the moment after which the user has established a habit around the value prop
    5.established a habit and continued use within a defined time period
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Definition of activation

A

Taking a user from signup to establishing a habit around your core value prop

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Four strategies to accelerate habit creation

A
  1. Manufactured loops
  2. Environment loops
  3. Use case transition??
  4. Grand exit??
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Optimization for aha moment

A
  1. Core action
  2. Warm start
  3. Supporting actions
  4. Empty states??
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

3 common mis interception about the resurrection

A
  1. Huge pool of users
  2. A way to contact
  3. Nothing to lose
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

Resurrection probability time decay

A

The probability of being able to resurrect a user decays extremely quickly once they go dormant.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

Two types of dormant users

A
  1. Involuntary users, disengaged due to some reasons that are not conscious choice
  2. Voluntary users, who made a conscious choice to not engage
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

Involuntary users resurrection

A

Involuntary users tend to be easier to resurrect
Three categories of reasons why somebody might become an involuntary dormant user
1. Product issue
2. Leaving
3. Payment

NOTE, when think about involuntary strategies, it is critical we try to be proactive. Before during and after.??

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

Voluntary users resurrection

A
  1. The why, need to understand the reason why the person has gone dormant
  2. The message, then figure out and align the message to fit with the that why
  3. The timing, then figure out the right timing for the message
  4. The channel, then decide the right channel to deliver that message
  5. The reactivation experience. decide the reactivation experience that users will land on once they respond to our message
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

Acquisition - Three Guiding principles

A
  1. Channel Dynamics, Law of Habit Transfer: Every product is built off the back of another channel by transitioning habit. We do not control the rules of that channel
  2. Product Channel Fit, Products are built for channels, channels do not mold to products.
  3. Channel Model Fit, Models enable or disable channels and can not be thought about separately
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

Acquisition - Lifecycle of a Tactic

A
  1. Tactic discovered, effectiveness 7
  2. Tactic optimized, effectiveness 9
  3. Tactic adopted by the mass, effectiveness 6
  4. Tactic fatigues, effectiveness 2
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

Acquisition is loops not funnels

A

Key question: How does one cohort of users lead to another cohort of users?

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

Acquisition - funnels drawbacks

A
  1. Funnels create strategic silos, think acquisition, product and monetization as silos.
  2. Funnel create functional metric silos, end up optimizing at the expense of each other
  3. Funnels push us to invest in linear activities.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
26
Q

Acquisition - why loops are vital

A
  1. If you are not compounding, you are dying. If your competitors are getting better faster, you are dying.
  2. Loops are more defensible. Loops combine how your product, channel and model work together.
  3. Loops create a more efficient cost of distribution over time. Linear strategies on the surface are easily copied, increasing cost over time.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
27
Q

Acquisition - categories of loops

A
  1. Viral loops, Word of mouth, organic, casual contact and incentivized
  2. Content loops, UGC - SEO, UGC - Social, CGC - SEO and CGC -Social.
  3. Paid loops, varies by platforms, google, facebook etc
  4. Sales loops, Inbound sales (a sales loop combined with a content lead loop), outbound sales (a sales loop combined with another human powered lead loop) and channel partnership.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
28
Q

Many fastest growing companies layer on multiple loops over time

A

But more loops does not equal a better strategy.
Every loop has a cost and a return.
1-2 high return loops is better than a lot of low return loops.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
29
Q

Acquisition - How to measure loops

A

Growth Multiplier
If I put one user into the loop how many more does it produce over time through all of the different cycles of the loops?

1 / (1 - V) where V is the ratio of new signups between two cycles

it shows how small improvements to your loop can lead to big results

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
30
Q

Acquisition - When is it a loop and when is it linear?

A
  1. Direct reinvestment of the output
  2. Time based growth multiplier, high average revenue per user (ARPU) only needs low growth multiplier. Vice versus
  3. High ceiling, every loop has a saturation point, we need to make sure it is not saturated already.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
31
Q

Acquisition - Linear channels usages

A
  1. Feed loops
  2. Activation energy. Some loops require a certain amount of activation energy to hit a sustainable point.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
32
Q

Acquisition - Viral loops

A
  1. There are 4 types, Word of mouth, organic, casual contact and incentivized
  2. The loops are not mutually exclusive
  3. The primary constraint is interaction between branching and response
  4. We measure the K factor to understand the shape and velocity. K factor means every 1 new signup will produce K new users. 100 users eventually got 15 new users from their invite, then K is 0.15
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
33
Q

Acquisition - Viral loops - Organic loop

A

Three keys??
1. Natural trigger
2. Trigger frequency
3. Response rate

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
34
Q

Acquisition - Viral loops - Casual contact loop

A

Three keys
1. Branching, need high branching number
2. Time, require time to reach escape velocity (It’s a point when the business grows exponentially, at a rate that is considerably greater than it was progressing previously) because of the low intent
3, Ripple effect, often don’t show up as a direct source. But via other channels.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
35
Q

Acquisition - Viral loops - Incentive loop

A

Three types
1. Money, coupon
2. Content, Games gem
3. Features, more storage

Three keys to incentivized loop
1. Meaningful
2. Alignment
3. Positioning, is it positioned in a compelling way? kind of cheesy though

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
36
Q

Acquisition - Content loops

A
  1. 4 types of content loops, UGCD - SEO, UGUD - Social, CGCD - SEO and CGUD -Social.
  2. They are not mutually exclusive, typically combined
  3. The primary constraints of content loops are cost/volume of content be generated and return per piece of content.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
37
Q

Acquisition - UG Content loops

A
  1. Why. What is the core motivation for a user to generate content?
  2. Volume. What is the volume of the content created with this specific content loop?
  3. Return. What is the return over time for a specific piece of content?

The return from search engine is slowly increasing then reach a cap point, while the return from social is increase fast in the beginning but drop really quickly

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
38
Q

Acquisition - CG Content loops

A

Due to the costs of company generated content, main constraint is about return.

Normally pair company distributed and user distributed to maximize return.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
39
Q

Acquisition - Paid loops

A
  1. Input costs. How much does it take to get started?
  2. Targeting. How efficiently can you target your target audience?
  3. Format/Steps. What is the format and the steps that the suer needs to go through?
  4. Scale. How much of your target audience is within that channel?

The primary constraint on paid loops is the capital to reinvest in the loop.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
40
Q

Acquisition - How to measure paid loops

A

Return on ad spend (ROAS)
The key idea of ROAS, it shows the time when we get the capital back.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
41
Q

Acquisition - Paid loops Pros vs Cons

A

Pros:
1. Quick results
2. Control
3. Targeting

Cons:
Least sustainable. So another loop should always be the core of your model. Paid loop should always accelerate other loops.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
42
Q

Acquisition - Sales loops

A

Company distributes value prop via humans in a way that attracts more customers that can be reinvested into more humans.

Sales loops are always paired with another loop that generates leads

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
43
Q

Acquisition - Sales loops constraints

A
  1. Rep productivity
  2. Time to productivity
  3. Capital
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
44
Q

Acquisition - Sales loops types

A

The follow list has efficiency going low to high and touch point from humans to product

  1. outbound sales (a sales loop combined with another human powered lead loop)
  2. Value added reseller (VAR)
  3. Inbound sales (a sales loop combined with a content
  4. Product, product usage to sales leads
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
45
Q

Acquisition - Sales loops - VAR

A
  1. Timing. Understand how to sell to customers yourself before you can train others to sell to customers.
  2. Market size. There needs to be enough VARs in the market that the loop is repeatable
  3. Value proposition. What is your value prop beyond just financial incentives?
    Value prop = Motivation + Constraint
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
46
Q

Acquisition - Sales loops - Inbound loops

A
  1. Alignment. Alignment between the value prop of the content presence and the value prop of the product.
  2. Signals. Finding the right signals of content engagement to engage at the right time.
  3. Amplification. To justify the return, the content needs to have a strong enough amplification/influence.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
47
Q

Acquisition - Sales loops - Product

A
  1. Signals. What are the right signals that indicate priority and readiness of a lead to interact with sales
  2. Cannibalization. How do you ensure the sales team doesn’t cannibalize revenue that would have happened anyway?
  3. Complexity. The product needs to follow the guidance of product channel fit/product model fit.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
48
Q

Acquisition Strategy - 3 Strategic levers

A
  1. Optimize our loops
  2. Add loops
  3. Increase linear channels.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
49
Q

Acquisition Strategy - 3 Strategic levers - Add new loops

A
  1. Adding new loops increases ceiling and makes others more efficient
  2. Order of operations matters. One loop may enable another loop, but not in the reverse order
  3. Adding new loops is extremely difficult. Normally requires new products/features.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
50
Q

Acquisition Strategy - 3 force we need to fight

A
  1. Ceiling and saturation
  2. Audience shift
  3. Product channel fit breaking.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
51
Q

Acquisition Strategy - Channel Maturity S curve

A
  1. Traction
    a. ROI: High
    b. Ceiling: Low
    c. Risk: Low
    d. Strategy: best for traction. If larger company, won’t make a dent.
    e. Example: Niche blogs
  2. Golden age
    a. ROI: High
    b. Ceiling: Low to high
    c. Risk: High
    d. Strategy: Move fast, double down, don’t just bolt on.
    e. Example: Slack, messaging apps
  3. Saturation
    a. ROI: Medium
    b. Ceiling: High
    c. Risk: Low
    d. Strategy: Focus on retention, LTV, payback periods.
    e. Example: Mobile, content marketing
  4. Decline
    a. ROI: Low
    b. Ceiling: High but declining
    c. Risk: High
    d. Strategy: Stay away if a startup
    e. Example: Print Ads, Facebook pages
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
52
Q

Monetization - 4 components

A
  1. How, how you charge (ads, transaction, subscription)
  2. When, when you charge (upfront, free trial, freemium)
  3. What, what you charge for (contacts, API calls, features)
  4. Amount, the amount you charge (price)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
53
Q

Monetization - Friction

A

Known New
$10 ARPU $100 ARPU $1000 ARPU $10000 ARPU $100,000 ARPU
Freemium Free trial Upfront
Ads Transaction Subscription
Low Friction High Friction

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
54
Q

Monetization - Balancing

A

Balancing with Acquisition: Cost & Influence (Model Channel Fit)
Balancing with Core Value Prop: Friction & Frequency (Model Product Fit)
Balancing with Segment: Size & Willing to pay (Model Market Fit)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
55
Q

Monetization - Model Channel Fit

A

Low Model Friction Viral B2C SEO Paid B2B Content/Inside Sales Enterprise Sales Hight Model Friction
Low CAC High CAC

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
56
Q

Monetization - Model Channel Fit Danger zone

A

The area between Paid and B2B Content/Inside Sales

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
57
Q

Monetization - Model Product Fit

A

Core Value Prop Fiction: If high friction, then free, freemium, free trial etc are difficult because it is hard for the customer to experience value in a free period. If low friction, those types of models become interesting

Frequency: How often does the user experience the core value prop within a certain time period? High frequency lend itself well to free, freemium or subscription type models. Low frequency means you need to capture as much value at time of use.

Best fits
Low Model Friction & Low Core Value Prop Friction
High Model Friction & High Core Value Prop Friction
High Model Friction & Low Nature Frequency
Low Model Friction & High Nature Frequency

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
58
Q

Monetization - Model Market Fit

A

Build $100M revenue/year company
$1 100,000,000 Consumer Everyone
$10 10,000,000 Mass Market Consumer
$100 1,000,000 Niche Consumers Prosumers
$1,000 100,000 Prosumers SMB
$10,000 10,000 Mild Market
$100,000 1000 Enterprise
$100,000,000 100 Fortune 500, Governments

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
59
Q

Monetization - How do we know if we have Model Market Fit

A

1 year ARPPU * # Customers in Market * % You think you can capture >= $100M

10% ————————-> 80%
Low Network Effect High Network Effect??

60
Q

Monetization - How to measure, 4 metrics

A
  1. Revenue Creation. How fast do we create revenue from different points of acquisition?
  2. Payback period. How fast do we make our costs of customer acquisition back?
  3. Revenue retention. How well do we retain revenue and revenue-producing users once they convert to being a paying customer?
  4. One year average revenue per paying user. How much is that paying user or customer worth to us over the course of a year?
61
Q

Monetization - How to measure - Building cohorts

A
  1. Who, who do you want to include in the population to cohort?
  2. Time period. What time intervals do we want to look at?
  3. Cell value. What do we want to include in the value of the cells?
62
Q

Monetization - How to measure - Building cohorts - Who

A

Start at the point of acquisition, not when they start paying.
1. Ad based. start with free user signup
2. Transaction. start with free user signup, whether they have transacted or not
3. Freemium. Start with user signup
4. Free trial. Start where the user signs up for their free trial
5. Subscription upfront. Start with acquired leads.

63
Q

Monetization - How to measure - Building cohorts - Time period

A

Ad, Transaction: Natural Frequency
Subscription: Frequency of billing

64
Q

Monetization - How to measure - Building cohorts - Cell value

A
  1. Revenue. This helps us understand the velocity in the pattern of revenue creation along with payback period.
  2. Paying customers. Understand the velocity and pattern of converting somebody into a paying user.
  3. Revenue per user. 1/2
65
Q

Monetization - How to measure - Revenue retention + 1 year ARPPU

A

Same method Who + time period + cell value
Who: Start with paying customer
Time period: same
Cell Value:
1. Revenue
2. Payment customers
3. Revenue per user (this help us estimate the 1 year ARPPU which in turns answers market fit)

66
Q

Monetization - Pricing - 3 parts

A
  1. Price metric. What do we charge for?
  2. Price structure. What are the different buckets and segmentations?
  3. Price amount. How much do we charge?
67
Q

Monetization - Pricing - Price metric

A

Criteria for metrics
1. They are easy to understand
2. They grow with the value of your customers. As your customers invest more in your product and they grow their business, they value that you get to capture also grows.
3. They are very well aligned with customer needs

Max diff is useful here. What do you value the most and the least?

68
Q

Monetization - Pricing - Price structure

A

Still use max diff to figure out what features customers like the most

69
Q

Monetization - Pricing - Price amount

A

Use Van Westendorp

We give description about the product and different tiers with features includes but NO price, then we ask
1. At what price would you consider the product to be so expensive that you would not consider buying it? (Too expensive)
2. what price would you consider the product to be so low that you would feel the quality would not be very good? (too cheap)
3. what price would you consider the product to get too expensive so that it is not out of the question, but you would have to give it some thought about buying it? (Not a Bargain)
4. what price would you consider the product to be a bargain, a great buy for the money? (Not expensive)

Then convert the answers to a percentage curve

The range of acceptable prices is the intersection of the too cheap and not a bargain and the intersection of the not expensive and too expensive

70
Q

Monetization - Optimization

A

How do we convert more customers without changing the actual price?
Perceived Value > Perceived Price + Friction

71
Q

Monetization - Optimization - Perceived Value

A
  1. Emotional (More relevant in B2C) What feeling am I gaining from purchasing this?
  2. Functional (More relevant in B2B). How much money/revenue or cost reduction will I gain?
  3. Accrued. The longer they get to experience the product the more perceived value.
72
Q

Monetization - Optimization

A

Difficulty
High Perceived Value
Perceived Price
Low Friction
Low High
Impact

73
Q

Marketplace

A

For a marketplace to be successful, it has to build supply, build demand and match supply and demand to create value.

74
Q

Marketplace Liquidity

A

The point at which supply can reliably sell what they offer and demand can reliably find what they seek
In other words, they marketplace is reliably delivering value to both supply and demand.

To identify when a market is at liquidity, we focus on demand. This is because demand is the ultimate decision maker in any marketplace.

75
Q

Marketplace Liquidity - Finding the tipping point??

A
  1. Aggregate demand decision drivers and value metrics.
  2. Identify which decision drivers are best correlated with value
  3. Determine the tipping point of demand value.
76
Q

Marketplace Liquidity - Finding the tipping point - demand decision drivers and value metrics

A

Demand decision drivers: Any data point that influences demand’s decision to transact on the marketplace. Inputs to the value for demand
Demand value metrics: Metrics that measure the value that demand gained from the marketplace. Outputs of the decision drivers. E.g conversion rate, net promoter score(NPS), transaction repeat rate and long term retention rate.

77
Q

Marketplace Liquidity - Finding the tipping point - Identify decision drivers

A

A line chart with demand decision metrics on the x-axis and the demand value metric on the y axis

4 curves
Positive S-curve
Negative S-curve
Linear curve, has some correlation, it doesn’t create a step change in value for demand. It’s the tipping point for liquidity.
Random walk. No correlation

78
Q

Marketplace Liquidity - Finding the tipping point - Reasons for lack of correlation

A
  1. Incorrect demand decision drivers or value metrics
  2. Metrics selected were not instrumented correctly
  3. Marketplace has insufficient scale of transactions to conduct this analysis
  4. Marketplace doesn’t have product market fit.
79
Q

Marketplace Liquidity - Finding the tipping point

A

This is the numerical value of the demand decision driver where value to demand spikes.

80
Q

Marketplace Liquidity - Measure

A
  1. Define your market. The market is the most granular unit of supply which is not replaceable for demand.
  2. Visualize Liquidity for those markets, the proportion of demand for which the tipping point of demand value was met.
  3. Determine the optimal range of liquidity. Use a scatterplot of ll markets, On X axis is liquidity while Y alix is a growth metric, like month over month increase in demand or month over month in transaction.
81
Q

Marketplace Liquidity - Measure - Define market

A

Example of markets Definition Dimensions
Plumers in Denver Service + Geography Service&Geography
Studio for 2 min Manhattan Dates + Size + Location Travel Dates&Size&Location

82
Q

Marketplace Liquidity - Strategies to drive liquidity

A

Key is to narrow scope
Other levers
1. Volume of supply
a. Invest in low ROI paid marketing
b. Build single player mode for supply
c. Tap into existing sources of supply
d. Acquire tentpole supply with favorable terms
2. Quality of supply
a. Demand changes needs in pre-liquidity and post-liquidity
b. Set rules or guidelines
c. Incentivize quality
3. Matching supply and demand
4. Transaction risk
a. Behavioral. Is this behavior acceptable?
b. Financial. Am I going to lose money?
c. Safety/Security. Will me and my loved ones be safe?
the solution
a. Acceptability signals
b. Insurance
c. Free returns and cash back policies
d. Two way feedback mechanism
e. Security feature
f. Branding and PR

83
Q

Growth Model??

A

A qualitative or quantitative representation of how your product grows.

84
Q

The purpose of a growth model

A

To answer a key question

How does your product grow?

85
Q

Growth Model - 3 steps to build

A
  1. Qualitative model
  2. Single loop quantitative model
  3. End to end model
    Each is built on the previous one
86
Q

Growth Model - 3 Steps - Pros and Cons

A

Qualitative model
Pros: Great for broader communication, brainstorming, strategic thinking and planning
Cons: Lacks accurate quantitative measures so doesn’t help with prioritizing, goal setting, and understanding outcomes
Single loop quantitative model
Pros: Helps validate and prioritize ideas around a single loop without dealing with the end to end model
Cons: Doesn’t help understand how multiple loops work together and influence each other
End to end model
Pros: Good for goal setting, building confidence on strategic investments, and directionally correct predictions.
Cons: Typically complicated to understand fully unless you live in it. Terrible for communications.

87
Q

Growth Model - how to build a qualitative model

A
  1. Define output (the retention metric)
  2. Add habit loops
  3. Add acquisition loops
  4. Add linear activities
88
Q

Growth Model - qualitative model - 3 variations

A
  1. Double loops, represent immediate areas of opportunity, as they have double the effect on your product.
  2. Split Model, supply and demand
  3. Current –> future
89
Q

Growth Model - how to build a single loop quant model

A
  1. Write out the steps of the loop
    1.a Steps should describe the core actions happening
    1.b End where the first one begins
    1.c Have an output
  2. Define the output metric for each step
    2.a Good output metrics are often total or new numbers
    2.b Not conversion numbers or percentages
  3. Define the conversions formulas between each step
    3.a Previous step output
    3.b Conversion variables
    3.c External resources
    3.4 Defining these is our hypothesis about the causal relationship between each step
  4. Insert the baseline
    4.a Choose a consistent time period for each one of these variables
  5. Assemble and evaluate our quantitative model
90
Q

Growth Model - a single loop quant model - three big questions

A
  1. Is the loop viable? Does it produce up into the right curve? Is it self sustaining?
  2. What is the largest constraint on the loop? Which variable impacts the loop the most?
  3. How does this loop compare to another loop?
91
Q

Growth Model - a single loop quant model - Ways to make the model more detailed

A
  1. Detail, adding additional detail to the conversion formulas to come close to reality
  2. Conversion variables, Layer on how version variables may evolve/change over time.
  3. Segmentations, Add in segmentations of different types of users and how they perform differently
  4. Cohorts, key off cohorts rather than average numbers.
92
Q

Growth Model - End to end quant model - common trap

A

To make your model too complicated in an effort to make it exact

93
Q

Growth Model - how to build an end to end quant model

A
  1. Identify the key output, what is the key output that indicates we are growing value for our users and the company? Common key output metrics are Retention metric and Liquidity Metric.
  2. Break it down into its inputs
  3. Insert its baselines
  4. Project the output over time
94
Q

Growth Model - End to end quant model - 3 areas to make your model more accurate

A
  1. Key action cohorts, What key actions drive the model?
  2. State segmentation, Do my user base segments interact differently with my growth loops?
  3. Retention loops, How are users returning to my product?
95
Q

Growth Model - Apply your models

A
  1. Predict + Project, Peering into possible futures
  2. Set goals, What are possible paths to X goal?
  3. Prioritize, deciding between X and Y initiatives
  4. Communicate, explaining why X is so important
  5. Continually update, your model is an evolving hypothesis.
96
Q

The job of growth

A

Our job as growth professionals is to help your users/target audience make decisions to take an action that leads to growth

97
Q

Growth model and user psychology

A

Growth model -> What
User psychology -> Why

98
Q

The heart of user psychology

A

Technology changes, but people don’t.

99
Q

User psychology - ELMR framework

A

Decision hill
1. Emotion, start with here!
2. Logic
3. Motivation, ability + motivation
4. Reward

100
Q

User psychology - ELMR framework - Emotion

A

Emotions stem from the gain or loss of our desires

Desire: Money/Economic
+: Secured Powerful, Superior, Accomplished
-: Stressed, Ripped off, Poor, Ashamed, Powerless, Worry
Desire: Knowledge
+: Curious, Understanding, Confident, Certain, Easy, Liberated, Trust, Clever
-: Inferior, Incapable, Useless, Pathetic, Unsure, Confusion, Indecisiveness, Apprehension
Desire: Social Approval
+: Important, Reassured, Admired, Superior, Accomplished, Achievement
-: Disapproval, Dejected, Rejected, Humiliated, Anxious, Depressed, Isolated, Inferior
Desire: Companionship/Sex
+: Confidence, Love, Passion, Sympathy, Comfort, Devoted, Frisky, Accepting, Happy, Affectionate
-: Loneliness, Sadness, Crushed, Hurt, Heartbroken, Inadequate, Alone
Desire: Entertainment
+: Surprise, Interest, Anticipation, Amazed, Fun, Happy, Delighted
-: Boredom, Disinterested, Annoyed
Desire: Purpose
+: Satisfied, Confidence, Optimism, Thrilled, Spirited, Eager, Free, Tenacious, Achievement
-: Empty, Distressed, Frustrated
Desire: Health
+: Energetic, Alive, Playful, Confident, Courageous, Strong
-: Fatigued, Pain, Unpleasant, Uncomfortable
Desire: Understanding
+: Comfortable, At ease, Calm, Devoted
-: Nervous, Uncomfortable, Lonely
Desire: Physical Feeling
+: Secure, Reassured, Calm, Peaceful, Relaxed, Trust
-: Fear, Scared, Anxious, Aggressive, Unstable

101
Q

User psychology - ELMR framework - Logic

A

We use logic to justify the emotions we have used to make a decision

4 Logical Appeals
1. Features, The specifics of what the product does
2. Statistics, Facts and figures
3. Reliability, Can I rely on what this product says it can do?
4. Price, If we introduce price before emotion, the user can’t justify the purchase.

102
Q

User psychology - ELMR framework - Challenges With Logic

A
  1. Our logical brain is very lazy
  2. Our logical brain is easily depleted
  3. Our logical brain gets overwhelmed
103
Q

User psychology - ELMR framework - Motivation

A

Motivation + Capability

The bigger capability required the steeper the hill is.

To get someone over the decision hill, we can either increase motivation or decrease size of the hill.

The motivation fuel is the first emotion part. But we can use motivational boosts to drive our motivational fuel even further.

104
Q

User psychology - ELMR framework - Motivational Boosts

A

Trust/Credibility/Authority: There is a greater chance we will do it if we feel it is from a trusted source
Urgency: We will do more if we feel there is limited time to do it
Scarcity: We want something more if we think there isn’t a lot of it
Bargain: We will more likely do something if we think we are getting a bargain
Belonging: We will do more if we fell that everyone else is doing it.
Liking: We will do more for people/things we like
Reciprocation: If something is given to us, we feel the need to return the gesture.
Consistency: We look to stay consistent with previous actions
Completion: We have an urge to complete things we started

105
Q

User psychology - ELMR framework - Things that make the decision hill bigger

A

Time: How long does it take to complete an action?
Money: How much is the financial cost to complete the action?
Social: Is this an accepted behavior by others?
Routine: Is this an action users are used to taking?
Thought: How much thought is required to complete the action? i.e. Confusion

106
Q

User psychology - ELMR framework - Reward

A

The reward is the confirmation that we made a good decision

The effect of a reward decreases with repeated exposure

We need to reward enough until organic habit takes hold

107
Q

User psychology - ELMR framework - 3 Types of Rewards

A
  1. Extrinsic, Time, Money and Information
  2. Intrinsic, Completion, Mastery and Joy
  3. Social, Recognition, Confirmation and Competition
108
Q

User psychology - ELMR framework - Ways to maintain reward effects

A
  1. Increase with exposure, more money each time
  2. Add new rewards
  3. Decrease the time between completing the action and receiving the reward
  4. Add variability.
  5. Multiply the effect of one reward, by reminding them of the reward in multiple places.
109
Q

User psychology - Psych framework

A

Psych = A unit of measurement of user motivation
Can have negative or positive psych

We assign positive or negative psych to each element, then we add them up which would be the final psych for that page.

Starting psych is influenced by where they came from and emotional fuel.

110
Q

User psychology - Tapping into emotion

A
  1. Identify emotional strategy
  2. Identify core emotion
  3. Improve product with emotion
111
Q

User psychology - Tapping into emotion - Identify emotional strategy

A

Product need
low candy -> medium vitamin -> high pain killer??
Lower emotional starting point will require more fuel to overcome friction

112
Q

User psychology - Tapping into emotion - Identify core emotion

A
  1. Identify the benefit/why of your product
  2. Moments your product solves
    2.a For Candy and Vitamin products, the moments around benefit
    2/b For Pain killer products, moments around problem
  3. How does your target audience feel in those moments?
  4. How would they feel if that was eliminated/recurring?
113
Q

User psychology - Tapping into emotion - Improve product with emotion

A

4 S elements
1. Selfishness, Make it about them
2. Sensory, More visualize
3. Specific
4. Simple

114
Q

Experiments - Typical Human Biases

A
  1. Overestimate probability
  2. Overestimate potential
  3. Inflate in hindsight
115
Q

Experiments - Possible sources of outside influence

A
  1. Seasonality
  2. Bugs
  3. Acquisition changes
  4. Product changes
  5. Competitive movements

Experiments isolate impact of changes while controlling for outside influences

116
Q

Experiments

A

Experiments help us identify the truth and learn about the truth.

117
Q

Experiments - Common misunderstandings about experiments

A
  1. The purpose of experiments is to optimize a metric
    1.a Instead we view experiments as a way to improve the product
  2. Hypotheses are guesses of how things will perform
    2.a Hypotheses are statement, a testable answer to a problem.
  3. Tests should always be the easiest/fastest thing to implement
  4. Experiments are cheap so we should always run them
  5. Analysis ends with success/failure
    5.a success/failure -> accuracy -> why
118
Q

Experiments - 4 steps

A
  1. Defining well structured hypotheses around key problems
  2. Define minimum viable tests (MVT) that are both efficient and valid
  3. Prioritizing MVTs
  4. Analyze experiment results
119
Q

Experiments - Define a hypothesis

A

Hypothesis is a testable answer to a problem or question

  1. Problem, what is the problem you are trying to solve or the question you are trying to answer?
  2. Hypothesis, What do you believe to be true that will solve the problem?
  3. Evidence, What evidence do you have to back up your hypothesis?
  4. Prediction, What do you believe will happen if your hypothesis is true? if, then …
120
Q

Experiments - Define a hypothesis - Problem

A
  1. Problems should be tied to high impact areas in your growth model
  2. Get specific on what you need to happen
  3. The Goldilocks rule, not too wide nor too narrow
121
Q

Experiments - Define a hypothesis - Hypothesis

A

While problems are rooted from our growth model, the hypothesis is rooted within our user psychology.

122
Q

Experiments - Define a hypothesis - Evidence

A
  1. Qualitative data, user research, surveys, sales conversations, support tickets etc
  2. Quantitative data, analysis of metrics, data science projects
  3. Prior experiments
  4. Observations, observations of competitors or similar products and how they do it
123
Q

Experiments - MVT

A

Minimum Viable Test is the most efficient way we can test the hypothesis in a valid way

124
Q

Experiments - Define MVT

A
  1. Scope, does this hypothesis contain one or multiple questions?
  2. Design, what is the design of our experiment?
  3. One success metric and tradeoff metrics, What will indicate our hypothesis is true? What tradeoff might we see???
125
Q

Experiments - Define MVT - Design

A
  1. What is the most aggressive thing that we could do to test the hypothesis?
  2. What is the simplest minimum viable test to test the hypothesis?
  3. Is there a smaller experiment to run towards an ideal experience?
126
Q

Experiments - Prioritizing MVT

A

Key questions: is this worth trying?
1. Effort, t-shirt size
2. Probability that this experiment will be a success, high medium or low
3. Upside, How much do we think our success metric will move? Upside = Reach * Impact
4. Sample Size + runtime, www.experimentcalculator.com
5. Scorecards, priority based on 1 dimensional and 2 dimensional scorecards?

127
Q

Experiments - Prioritizing MVT - Upside Reach

A

It’s better off thinking outside of your current user base??

  1. Core users
  2. Active users
  3. Registered users
  4. Uniques
  5. Channel
  6. All channels combined
128
Q

Experiments - Prioritizing MVT - Upside Impact

A

We need to look at impact over time from the model (loop specific or end to end quant model)

129
Q

Experiments - Prioritizing MVT - Upside - Points of leverage

A

3 Common Leverages
1. Loops, especially compound loops
2. Retention + engagement
3. Constraints

130
Q

Experiments - Methods to call experiments

A

Standard Statistical Significance
Pros: Easiest to implement and most well understood
Cons: Risk calling something significant when there is still large volatility in p-value; based on size of treatment effect, which is har to know.

Continuous Monitoring
Pros: Avoids calling a test too early when p-value volatility is high. Let’s people peek
Cons: The math is no different than a standard p-value test

Sequential Sampling
Pros: Can reduce the number of conversion observations needed for a successful experiment by 50% or more, Easy calculation.
Cons: Less understood. Works better for small conversion rates

Dynamic Decision Boundaries
Pros: May enable ability to call experiments earlier on an automated basis
Cons: Difficult to implement. Need tons of historical data to model decision boundary correctly

131
Q

Experiments - Analyzing

A

Typically less than 25% of experiments succeed. An experiment is only a failure if you fail to learn.
1. When to end
2. Analyzing the results
3. Applying + communicating experiment learnings

132
Q

Experiments - Analyzing - Sequential Sampling

A
  1. Choose sample size N assign 50/50 control vs experiment
  2. Track experiment successes E
  3. Track control successes C
  4. If E - C >- 2√N standard deviation, declare variation the winner
  5. If E + C > N, declare control the winner
133
Q

Experiments - Levels of Analysis

A
  1. Why, ask why the experiment was a success/failure. What are the potential reasons?
  2. Accuracy, How accurate were the results to my hypothesis? Close or really far off?
  3. Success/Failure, Was the experiment a success or a failure? Did it improve the metric in the hypothesis?
134
Q

Experiments - Levels of Analysis - Why

A
  1. Break apart the pieces
  2. Segment
  3. Qualitative data
  4. Additional experiments
135
Q

Experiments - Applying learning

A

Success
How do we double down? New Experiment
How do we apply to other parts of our growth model? Experiment / Backlog
Does it inform a current backlog item? Re-prioritize

Failure
Alternative hypothesis Experiment / Backlog
Does it inform a current backlog item? Re-prioritize
How does it inform my growth model and user psych map? review growth model / user psych

136
Q

Experiments - Applying learning - Communicating

A

Eteam:
Charter -> N/A
Roadmap -> Monthly Update
Process -> Management Meetings + Wiki
Learnings/Wins -> Monthly Update

Entire Org:
Charter -> N/A
Roadmap -> Company Wiki
Process -> N/A
Learnings/Wins -> Company Showcase

Growth team:
Charter -> Onboarding & Offsite
Roadmap -> Team Sync
Process -> Onboarding & Offsite
Learnings/Wins -> Team Sync

Growth Pod:
Charter -> Onboarding & Offsite
Roadmap -> Pod Sync and 1:1s
Process -> Onboarding and 1:1s
Learnings/Wins -> Pod Sync

137
Q

Experiments - Applying learning - Communicating Matrix

A
  1. What, What are we communicating?
  2. Who, Who are the key groups you need to be communicating to?
  3. How, How are you communicating these items?
    3.a Cadence
    3.b Channel
    3.c Contents
    3.d Format
  4. Evaluate, Where are you strong and week?
138
Q

Defensibility - Forms of defensibility??

A
  1. Network Effects
  2. Brand
  3. Embedding
  4. Scale
139
Q

Defensibility - Network Effects

A

More adoption leads to more product value then leads back to more adoption.

Virality != Network Effects

140
Q

Defensibility - Understanding Network Effects

A
  1. Types of Network Effects
  2. How network effects impact growth
  3. Build network effects
141
Q

Defensibility - Types Network Effects

A
  1. Direct NFX, More users increase the product value
  2. Data NFX, More data increase the product value
  3. Cross-side NFX, marketplace
  4. Platform NFX, More users leads to more developers then leads back to users.

Value of network effects
1. Linear
2. Asymptotic
3. Exponential

142
Q

Defensibility - How to Develop Network Effects

A
  1. Constraint, How do you constrain the initial audience to make density easier?
  2. Hook, What is the hook/value you offer to your initial audience?
  3. Time to Density, What is the minimal density you need for your target audience? How do you get there ASAP?
  4. Leverage, How do you use your initial audience to build demand for the next audience?
143
Q

Defensibility - Brand

A

Brand reinforces the key emotional constraint around the product, making acquisition and retention more efficient

Reinforcing loop around the primary emotional constraint

Every product needs to tap into a core emotion.

144
Q

Defensibility - Embedding

A

As a user adopts the product, switching costs increase, leading to more adoption.

145
Q

Defensibility - Scale

A

As volume grows, value increases to end customers

Business (11 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2024 Bold Learning Solutions. Terms and Conditions