Strategy Experiments in Nonexperimental Settings_ Challenges of Theory, Inference, and Persuasion in Business Strategy_.pdf

Strategy Experiments in Nonexperimental Settings_ Challenges of Theory, Inference, and Persuasion in Business Strategy_.pdf

• 1.
  p EXPERIMENTS IN NON-EXPERIMENTAL SETTINGS Strategy Experimentsin Nonexperimental Settings: Challenges of Theory, Inference, and Persuasion in Business Strategy" by Ron Adner & Daniel A. Levinthal Presenter: Rida Zaman [email protected] Instructor: Professor Francesco Rullani STRATEGY Ca’ Foscari University of Venice | Venice School of Management | PhD Course: Strategy For Innovation Management 1
• 2.
  TODAY'S AGENDA Introduction 1 Problem Statement 2 key ConceptsIntroduced 3 Proposed Framework 4 The Four Quadrants 5 Implications for Inference 6 Role of Theory 7 Practical Insights 8 Conclusion 9 2
• 3.
  This article addressesthe challenges faced by strategists when trying to learn from "experiments" in real world business settings. Unlike scientific experiments conducted in labs, strategy initiatives are often nonrepeatable and require joint action among many stakeholders. This complicates learning, inference, and persuasion three critical components of effective strategy development. Learning: Gaining understanding Inference: Drawing meaning from information Persuasion: Communicating that understanding effectively INTRODUCTION Example: The manager learned from employee behavior, inferred a better way of working, and persuaded leadership to act on it. Observation Concludes Convince 3
• 4.
  Business Strategy vs.Experiments: •Managers are told to test ideas like scientists. •But real world strategy is messy, fast changing, and unpredictable. •Actions can't be repeated the same way (non repeatability). •Success often depends on others' support (joint action). •So, it's hard to learn, prove cause-effect (inference), or gain support (persuasion). We need new ways to use theory for action, learning, and coalition-building. PROBLEM STATEMENT 4
• 5.
  Nonrepeatability: Strategic actions oftenreshape the environment in ways that prevent exact repetition. ⚬ Example: Coca-Cola revising its flagship soda formula permanently alters customer expectations and brand perception. ⚬ Example: Facebook rebranding to Meta impacts how future initiatives are perceived. KEY CONCEPTS INTRODUCED The authors introduce two important dimensions that differentiate strategic experiments from traditional scientific experiments: Joint Action: Strategies typically require alignment and coordination with multiple internal and external actors. ⚬ Example: Edison needed government permits, suppliers, and investors for the Pearl Street power plant. These dimensions highlight why strategy development must go beyond simple trial and error and rely heavily on theory. 5
• 6.
  LAB BENCH EXPERIMENTATION STRATEGIC DEBATE ACTINGWITH ENDURING CONCEQUENCES JOINT STARTEGIC ACTION The framework proposed by the authors is structured around two axes: • Repeatability: Whether the same action can be repeated with consistent outcomes. • Actor Type: Whether the strategy is driven by an individual or requires collective effort. This creates four distinct quadrants for analyzing strategic initiatives. PROPOSED FRAMEWORK INDIVIDUAL JOINT REPEATABLE NON - REPEATABLE 6
• 7.
  Individual & Repeatable THEFOUR QUADRANTS QUADRANT 2 Joint & Repeatable QUADRANT 3 Individual & Nonrepeatable QUADRANT 4 Joint & Nonrepeatable QUADRANT 1 7
• 8.
  QUADRANT 1 -INDIVIDUAL & REPEATABLE In this first quadrant, one person can try something over and over again without changing the environment. This is like working in a lab or testing a new product in its early stages. It’s easy to learn here because nothing around you changes. A great example used in the article is Edison testing 10,000 filaments for his light bulb. Each time he failed, he learned something new. But the testing space didn’t change so it was a stable environment for learning. 8
• 9.
  QUADRANT 2 -JOINT & REPEATABLE In the second quadrant, you still have a stable environment, but you need help from others. This could be internal meetings in a company where people work together to design a solution. For example, when Edison worked on the Pearl Street power station, he needed help from the government, suppliers, and investors. Even though the environment stayed mostly the same, he had to build support and understanding with others. That’s where theory helped him explain his ideas and bring everyone on board. 9
• 10.
  QUADRANT 3 -INDIVIDUAL & NONREPEATABLE Now we move into trickier territory. In the third quadrant, a single decision maker makes a big move but the move changes the environment so much that it can’t be repeated. Take Apple, for example. When they launched the first iPhone, it completely changed the smartphone market, what customers expect, and how competitors behave. Even if they tried to “undo” that move, they’d never go back to how things were. That’s non repeatability in action. 10
• 11.
  QUADRANT 4 JOINT& NONREPEATABLE Finally, we have the most complex case when the strategy needs many people working together, and it changes the environment so much that it can’t be done again the same way. A few examples from the article include Pfizer’s failed inhalable insulin, which changed how the pharmaceutical industry viewed similar projects. Or Facebook’s move to Meta once they did that, it set a new standard for tech platforms. Another case is Better Place, an electric car company with a battery-switching model that failed and scared off other companies from trying something similar. 11
• 12.
  So, what doesthis mean for how we learn, or what we call inference? In science labs, things are predictable because conditions don’t change. You can assume that the same input will give you the same result. But in business, one strategic move can completely change the game. The past might not help you understand the future. An example is A/B testing on platforms like Airbnb. If users in one group interact with those in another, the test results get mixed up and it becomes hard to know what really worked. IMPLICATIONS FOR INFERENCE 12
• 13.
  That’s why theorybecomes so important. It does two things: • First, it helps you predict what might happen. • Second, it helps you persuade others your team, your investors, your partners. Inside the company, theory helps people agree on a plan. Outside, it helps you win support and trust. ROLE OF THEORY 13
• 14.
  LEARNINGS FROM EACH CASE DISCUSSEDIN ARTICLE 14 Company -Case Key Lesson Scenario Discussed Coca-Cola – New Coke Nonrepeatability + Loss of brand trust Coca-Cola changed the formula of its iconic drink. The public responded negatively, and even after reverting to the original, trust and perception had already shifted. Facebook – Meta Rebranding Irreversible identity shift Facebook rebranded as Meta to signal a focus on the metaverse. However, the name change impacted how the world viewed the company in an irreversible way. Edison – Light Bulb Individual + Repeatable experimentation Edison independently tested thousands of filaments in a lab-like setting. He failed repeatedly but learned with each test without changing the external world. Edison – Pearl Street Station Power Plant Joint + Repeatable experimentation To launch the first commercial power plant, Edison needed collaboration from governments, suppliers, and investors. The challenge was persuading others to support a joint vision. Microsoft / SAP – Buggy Software Update Big brand = Big reputational risk When small companies release buggy updates, people are more forgiving. But for major firms like Microsoft or SAP, such failures can severely damage reputation. Pfizer – Inhalable Insulin Failure discourages future ecosystem initiatives Pfizer attempted to launch inhalable insulin. After it failed, other companies became hesitant to pursue similar innovations, fearing the same fate. Project Better Place – Battery Swapping One failure can kill an entire model The company introduced a battery-swapping system for electric cars. It failed to scale, and its failure discouraged others from pursuing similar infrastructure-heavy ideas. Michelin – Repairable Run-Flat Tires Innovation alone ≠ Market success Michelin launched innovative repairable tires. Despite the technology being sound, the market didn’t accept it, showing that innovation needs market validation too. Pfizer / COVID-19 Vaccine Scientific theory ≠ Public persuasion Even though the COVID vaccine passed scientific tests, public mistrust and conspiracy theories led to resistance, showing that success also depends on effective communication and trust- building.
• 15.
  TAKE AWAYS WHAT ARETHE KEY TAKE AWAYS FROM THIS PAPERS: 1. First, know that big strategic actions are often irreversible you can’t just try again if they go wrong. 2. Second, learning from experience is limited so using theory is key. 3. And third, persuasion is just as important as planning. In the article, the authors talk about how difficult it is to restart belief and support after a failed strategy. When companies “pivot,” they don’t just need a new idea they need to rebuild shared purpose. 15
• 16.
  CONCLUSION Strategic decision makingin real life is very different from scientific experiments. In business: • You can’t always repeat actions. • You often need others to make your plan work. So we need theory to: • Help us navigate (like a compass) • Help us convince (like a megaphone) Because strategists don’t just watch the world they help shape it. 16
• 17.
  THANK YOU Rida Zaman [email protected] foryour time and attention 17