Strategic Experimentation

Strategic Experimentation

• Structured experiments offer a clear path, unlike experimenting through trial and error without a hypothesis
• Resources are limited, structured experimentation is critical

Strategic Experimentation Introduction

• Experiments generate information that enables effective decision-making
• Louis Pasteur quote emphasizes the importance of experimental proof for inventions and ideas

Experimentation Types

• Hypothesis-Driven: Involves systematically testing and developing hypotheses
• Trial-and-Error: Involves learning through repeated trials and adjustments
• Thomas Edison quote highlights the value of learning from failures.

Structured Experimentation

• Helps startups save resources and learn faster
• A Better Place failed due to a lack of structured experimentation

Lean-Startup Movement

• Emphasizes experimentation in startups
• It is a business experimentation model:
• Starts with developing a vision
• Vision is translated into hypotheses
• Ensures falsifiability
• Experiments are sequenced and prioritized
• MVP test are specified
• Lessons are learned from MVP tests
• Decision is made to persevere, pivot, or perish

Business Model Experimentation Examples

• Dropbox illustrates the power of the Lean-Startup approach
• Vera Wang used high-fidelity experiments to innovate wedding gowns
• Segway highlights the importance of structured hypothesis-driven commercialization

Entrepreneurial Experimentation Principles

• Focus on testing critical hypotheses
• Ensure accuracy and precision in experiments
• Consider resources, strategic costs, and time.

Key Experimentation Takeaways

• Experimentation is essential for informed decisions and reducing uncertainty
• The Lean-Startup approach provides a structured method to develop and test hypotheses
• Strategic commitment is important for entrepreneurial strategy and experimentation

Experimentation Reasons

• Reduces uncertainty
• Determines the experiment's reliability
• Considers criticality
• Considers opportunity cost

Rent the Runway's Value Creation Hypothesis Tests

• Seeking advice from designers (although some feared brand dilution and cannibalization)
• Market Trials at Harvard and Yale
• Building a Website
• Online trials

Rent the Runway: Market Trials at Harvard and Yale

• Objective: To determine demand for dress rentals and the appeal of price points, styles, and sizes.
• Harvard Trial:
• Setup: 130 dresses from 15 designers, prices between $35-$75, plus $4 insurance
• Results: 34% of women rented dresses
  • Learning: Enjoyed trying on dresses and valued the self-confidence boost from wearing designer dresses. The rental experience is inherently social
• Yale Trial:
  • Setup: Like Harvard, but women could only view, not try on dresses.
  • Results: 75% of the women rented dresses
  • Learning: Showed there was significant demand

Rent the Runway: Online Rental Trial

• Objective: To test if women would rent dresses online without trying them on
• Setup: Sent PDFs with dress pictures to 1,000 women who wanted to be contacted when RTR launched. Rentals were made by telephone
• Results: 5% of women rented dresses
• Normal rate of 0.1%
• Validated the online rental model, rental rate was above what was needed to make a profit

Rent the Runway: Beta Launch

• Objective: To understand demand and operational challenges before the public launch
• Setup: Launched a beta version to 5,000 invited members.
• Results: Strong demand, many questions about how RTR works, and need for styling guidance.
• Learning: Stressed the importance of a waiting list to control inventory, the need to staff up and train customer service

Rent the Runway: Customer Insights Team

• Objective: Provide styling advice and gather feedback to encourage rentals and build trust
• Setup: Stylists gave advice via online chat or phone
• Results: Feedback from stylists resulted in service improvements and helped with customer concerns
• Learning: Building relationships and personalized advice were key

Rent the Runway: Post-Launch Surveys and Feedback

• Objective: To refine the value proposition and understand customer experiences
• Setup: Early customers were surveyed, and feedback was gathered from social media
• Results: Customers reported feeling special and confident wearing rented dresses, praised 12 compliment per night, up from 2 with their own wardrobe
• Learning: Emotional value was significant, shifted the company's mission to focus on delivering women's dreams and making them feel beautiful
• The experiments helped Rent the Runway understand customer needs, refine, validate the business model

Economics of AI

• As cost of math went down, strategy and decision began to adopt AI
• As with the IOWAcorn case, adopt AI early to get high returns
• Electric Production Factories take a long time to adopt (10 years +)
• Overfitting is when AI uses background to make assumptions, even if it is wrong

Artificial Intelligence (AI) Introduction

• AI is the ability of a machine to copy intelligent human behavior
• Machine Learning is an AI application with a computer learning without direct instruction

Historical Context: AI

• Internet: Revolutionized communication and commerce starting from 1969 with ARPAnet
• Semiconductors / Transistors: Key technological advances with Moore's law, introduced in the late 1700s

AI and Prediction

• Prediction uses current information to generate information that is not available
• Accurate prediction is essential for decision-making and is ubiquitous in applications like recommendation engines, stock trading, speech recognition, medical, insurance, and autonomous vehicles

Decision Process Framework

• Data: The types include training data, feedback data, and input data
• Data collection is costly but critical to enhance accuracy
• Actions: Machines (automation) or humans (augmenting decision-making) can take action
• Judgment: Assigns a reward determined by humans

AI Canvas Example

• MBA Admission: AI is used to predict successful students as well as redesign organizational workflows

AI in Organizational Workflows

• Goldman Sachs reduced its equity traders from 600 to 2 through AI automation.
• AI can change workflows and strategies, this requires organizations to adapt

AI and Strategy

• Adoption is being led by large companies like Google, Amazon, and Alibaba by levraging AI across multiple products
• Start-ups can exploit incumbents' inability to quickly adapt to AI creating disruption potential

AI Prediction Accuracy

• Strategy is impacted and accuracy reshapes roles, training and hiring practices

AI-Related Risks

• Problems can arise related to fraudulent purposes
  • Applications must be monitored for misuse
• Discrimination and biases
• AI might reflect existing biases in data
• Overfitting can occur
• Optimize for specific data, this leads to fragile accuracy
• Distributional Shift and Adversarial Attacks could occur
  • Data training may become outdated and AI can be tricked

Decision-Making Impact of AI

• Includes prediction, judgment, outcome, action, and data
• Complementary assets are human skills in judgment, data collection, and actions

AI Wrap-Up

• AI affects strategy, workflows, hiring, HR, and business processes
• Innovations like AI take time to become productive, like electricity

Platforms

• Platforms enable matches between two or more parties, thereby creating value
• Helps lower transaction and search costs
• Have a low cost structure

Platforms: Strategic Challenges

• Chicken and egg dilemma

Platforms: Network Effects

• Positive network effects create lock-in for users to switch to another platform. Example: Social network platform

Types of Network Effects

• Indirect: Users only gain value if the other side has more users
• For example, users need to have more drivers
• Direct: The value of a platform goes up on the same side
• The more users on Facebook, the more value

Platform Money and Subsidy Sides

• Subsidy side - Uber rider
• Money side - Uber driver

Platform Network Effect

• Strong network effects mean that the chicken and egg dilemma is a challenge to invite new users

Platform Multi-Homing

• Low Multi-homing: Ease of using multiple apps in the same industry(Lyft/Uber)
• High Multi-homing cost: Hard to use multiple apps in the same industry (Apple/ Music streaming)

Infrastructure for Platforms

• There needs to be infrastructure for platforms to increase
• Amazon was not a platform at first, follow the rabbit

Platform Producer Evangelism

• Producer evangelism for teachers is one method to get students
• Teacher is attracted to create course inline which increases the value of the platform, the student is invited to use

Platforms Introduction

• Match different market sides and facilitate the exchange of goods and services to create value.
• Reduce communication fees, find each other, gather information and service fees

Platforms Advantages

• Platforms don't take possession of the services and products leading to high gross margins For example, 70% for eBay and 60% for Etsy.
• Many companies are platform based (e.g., Facebook). 65% of startups w/ $1 billion valuation are platforms

Strategic Platforms Challenges

• Must make a selection for number of sides
• More sides can lead to more diversification in revenue streams while also leading to complexity and conflicts of interest
• Examples Uber (drivers/passengers), Airbnb (hosts/guests), BlackBerry (developers/users).
• The chicken-or-egg dilemma has to be solved
• One side must come onboard first to be able to attract others
• Airbnb, Uber have had successful examples
• The BB10 OS is an unsuccessful example
• positive network effects need to be considered
• Indirect: A new user benefits from a different group joining (example: Lyft)
• Direct: A user benefits from the same group joining (example: Facebook).
• Deciding pricing structure
• Subsidy side-attract users
• Money side charge
• Subsidize the price sensitive sides while deriving more value from the other sides: For example, Uber charges riders vs drivers.
• Must consider governance rules for platform establishement
• Who has access?
• How does interaction occur?
• There is quantity v quality trade-offs( IOS vs Android)

Chicken-or-Egg Dilemma: Bad Industry Conditions

• Strong Network Effects need critical scale of users.
• High Mulit-homing Costs: Costs for users to participate in multiple platforms
• Differentiation needs to be low
• there needs to be uniform user preferences
• High Barriers to Entry with economies of scale and data-related advantages

Solving Chicken-or-Egg Dilemma Strategies

• Follow the rabbit: Utilize demonstration (Example, Amazon Marketplace)
• Single side: For one side, you have to create a standalone value (e.g., OpenTable ,Trulia.)
• Seeding: Provide or simulate to attract users (e.g., Android, Apple iOS, Quora)
• Piggyback: Connect with user base on user platform for example: (MySpace, Youtube, Paypal on ebay)
• Big-Bang; high attention (South by southwest, Twitter)
• Micro market Target localized markets for example ( GrubHub, Facebook)
• producer evangelism; pull in consumers (SkillShare or Udemy)
• attract influential: (gaming consoles)

PayPal Case Study

• Faced challenges related to online payment platforms.
• Had these solutions
• Simplicity;
• Incentive:
• Financial Piggybacking: eBay
• Created bot, simulate: 4. Simulate Demand: Created a bot to bid on eBay items, insisting on PayPal payments.

Platform Wrap Up

• A key to competitive advantage and exponential growth is through network effects
• Strategy is depedentent on platforms
• You want value focused to be effective

CreatX Guest Speaker

• The assistant program is from Gatech
• Helps customer discover entrepreneur

Entrepreneurship: Evidence based

• Customer must face-to-face to get customer points of view

Miscellaneous

• You don't realize/ know that the ease of use/packaging that customers worry about bandages more size/absorption

Location and Ecosystems

• Turn out large effect thatAmenities and things will Impact OJ doing businesses
• Drivers made silicon valley today -WWII, California. military action was in California was localized to Japan, population in government

Ecosystems and Clusters

• Start-up location needs to local to monitor and closely follow up on Silicon Valley
  • A cluster is categorized from particular geographic concentration of interconnected companies
  • An ecosystem is a network of companies including, capital, firms, and other catalysts that allows an exchange
• Ecosystem Silicon Valley
  • Startups are in other company and support despite the costs -natural resource are important
  • governments and military are supported -universities early firms semiconductor
• luck hometown

Startup location choices

-Access of the consumers Importance of customers base quality of customers for better feedback understanding to move EU startups for access

• Access the suppliers cost
• innovation interaction for face interaction
• labor many skilled players/employees
• cultural stigma
• entrepreneurship culture cultural attitudes cultural low low rates stigmas- Japan

Startup Location Considerations

• Knowledge spillovers through encounters and recombination
• e prohibited innovation closed
• Capital you can get
• activity you can start-up locally by following start-ups
• Capital VC availability is influences
• VC pension for Government in USA
• regulations
• break
• Regulatory sandbox in UK
• networks
• Importance of social capital an entrepreneur
• Entrepreneurs may have social connections or may to to building network
• city location varies
• networking and proximity

Location Wrap Up

• Venture economies and venture economies on clusters and ventures
• silicon valley successful
• opportunities for resources exploitation by start-ups
• Founder’s stage resource depends decision in- location In industry specifics in IT

Platform Economics and Strategies

• Platforms are entities enabling matches between different market sides; platforms reduce transaction costs.
• They have strategic challenges, like the chicken-and-egg dilemma.
• Network effects can be direct (like Facebook) or indirect (like Uber).
• Key considerations for platforms include pricing subsidy and money sides.
• Bad Industry Conditions include strong network effects, high multi-homing costs, and low differentiation.
• Strategies for solving the chicken-or-egg dilemma include the follow-the-rabbit strategy and single-side strategy.