The Experimentation Field Book, by Jeanne Liedtka et al. (chapter one)

Introduction to the Experimentation Process

DOES THE WORD experimentation conjure images of test tubes, safety goggles, and possibly a dash of fire or smoke? Our early exposure to experimentation is often in science class as part of the scientific discovery process. Accompanying those beakers and lab coats is the scientific method, a guide to running experiments.

But many people run experiments every day outside of science labs. Long before you were learning about experiments in school, your toddler brain was experimenting as a primary method of learning: If I stack blocks like this, they don’t fall over. If I put my fingers near the door, I can get hurt when it shuts.

Many organizations use experimentation to explore ideas for new products, services, programs, and strategies. In this field book, we’ll share stories from different organizations. Nike wanted to experiment

Organizations in Brief

Nike is one of the largest and most successful athletic apparel companies in the world.

Whiteriver Hospital, located on the Fort Apache Indian Reservation, has both inpatient beds and an emergency room where most of the population’s medical needs have to be met.

South Western Railway (SWR), a joint venture between two of the world’s leading rail companies, operates some of the busiest train routes in the United Kingdom, with 235 million passenger journeys a year.

The Project Management Institute (PMI) is the premier global project management association, with nearly 500,000 members.

with a shoe subscription service for kids. Whiteriver Hospital wanted to reduce patient wait time in their emergency room. The Project Management Institute (PMI), a professional learning organization, was interested in testing different ways to serve its members. South Western Railway (SWR), one of the UK’s biggest commuter railways, hoped to improve passenger experiences. People at each of these organizations had good ideas on paper but wondered if they would succeed in reality. They worried about being wrong in their hunches. So, they used experimentation to test their hunches before they built and launched their new ideas and before spending a lot of time and money on them. Throughout this field book, we will return to these four stories for inspiration and lessons learned as we build your experimentation skills.

Entrepreneurs, teachers, caregivers, marketers, graphic designers, social workers, even people doing their everyday routines all experiment. Have you ever caught yourself saying, “I wonder if taking this route home is faster? Let me try it and time it.” You’re experimenting! Or think of a team leader considering, “If we send the agenda in advance, it might make our meetings more effective. Let’s try it for a few weeks and see if that’s true.” Experimenting! This field book focuses on helping you to systematically design and execute your own high-quality experiments on any idea you have.

Experimentation is much talked about in the abstract, but poorly understood in practice. Yet it is the crucial connector between imagining an idea and making it happen in reality.

Why Experiment?

Here are a few reasons putting in the time to experiment is better than just building a new idea:

• Experimentation protects you from overspending on a solution that won’t work for the people you designed it for. By placing small bets and learning at a fast pace, c n e n e e c nce e f the unmet needs of your users—whether patients, passengers, customers, or members—and if the idea is really desirable to them.

• It encourages you to test a portfolio containing multiple ideas rather than converging prematurely on a single idea. Then it lets your users highlight the one that works best for them.

• Through experiments, you learn how to scale your idea e ec ive nd cce e ing ec e idea at smaller scales reduces the risk of early service delivery friction and is less expensive.

• You’ll get to know your early adopters, and their feedback will inform future features and help accelerate the development of your new product or service.

• Experimentation is fun and engaging! It gets you out of the conference room and into the real world and allows you to invite those who will be part of the implementation of a solution into the testing process, building buy-in and adding energy and momentum to a project.

Experiments exist to collect the data to build an evidence base, to support whether and how to launch new solutions, but they are often underutilized or badly designed. At one extreme, they end up looking like “soft launches” with little appetite for pivoting. In other fields, only elaborate, formal experiments like randomized controlled trials, which often last years rather than days or weeks, count. In this book, we will show you how to take advantage of the oftenunexplored territory in between, where well-designed, learning-oriented field experiments can add tremendous learning and risk reduction but take only hours instead of weeks and cost next to nothing.

For most of us, building an evidence base is the best way to manage the inevitable risk of trying new things in today’s dynamic environments. In fact, skills in the design and execution of experiments are fundamental and critical competencies for success in an increasingly uncertain world where there are known unknowns (things we know we don’t know) and unknown unknowns (things we don’t know we don’t know). Experimentation helps us to better address the first category and discover the second one. This valuable tool is embedded into many problemsolving approaches (like Lean Startup, Agile Software Development, Design Thinking, Kaizen, and Process Improvement). Experiments also go by many names; you might have heard about a new Google feature that is “in beta” or about organizations running pilots or testing a minimum viable product (MVP). Yes, the terms used for experimentation can be confusing and full of jargon—but the terms we use are less important than being precise about what the experiment seeks to

test and how it will accomplish that. Experimentation is a powerful and effective learning tool for individuals or teams who are improving a product or service that exists or creating something new to them, their organization, or the world—all require deliberate learning through action. Though the need for experimentation may seem obvious, many organizations skip it. Why? Because the urge to “just build it” (and hope they will come) is strong.

A hypothesis is the essential starting point for experimentation. Yes, back to science class! A hypothesis is a best guess about what you expect to be true or assume is true. Nike employees could have just assumed parents would want a shoe subscription service and built it, but instead they experimented to test the assumption that it met a real customer need.

Hypotheses are conjectures that can be true or false. We find out if they’re true by collecting data to confirm or disconfirm them. For example, you can test if that new route you suspect to be faster truly is by timing it and then comparing that to your old route time. Hypothesis-driven thinking is the anchor of experimentation—and it is fundamentally different from just using existing data to make your decision. Hypothesis-driven thinking flips the usual process that starts with the data you’ve already got. Instead, you start with the hypothesis and then identify the data you need to test it rigorously (which is often not the data you’ve already got). For your data to be most useful, your data collection must be done with intention. To do that, you need to be hypothesis driven.

The Five Steps

Experimentation is not a “one and done” activity. It is a process that evolves over time. You will run multiple tests, which will increase in complexity as your hypothesis is further developed and refined. This de-risks your solution development. Rather than just building the real thing right away, experimenting helps you manage risk (you spend time and money incrementally as risk moves from higher to lower), gather evidence to test your assumptions from a range of actual users (not only from conversations among colleagues or early adopters), and improve your odds of success. The iteration that occurs with experimentation creates a stronger solution by making sure that people really want what you’ve designed and demonstrating that you can deliver it in a viable way. Experimentation can also be used to understand whether any given problem is even worth solving. Selecting Your Test 4 3

We hope that we’ve convinced you of the why of experimentation and the need to be hypothesis driven. In the rest of this field book, we will focus on the how by providing a hands-on, structured process to guide you through the design and execution of your very own experiments. We will outline how to design and run tests to address a variety of challenges—improving an existing solution, fixing something that’s broken, making a new product, or launching a new service for your existing customers. To do this, we provide a stepby-step process, using the templates we have created. Together we will explore five steps.

Let’s look briefly at each step. Steps 1 and 2 focus on the what of experimenting. Step 1, framing a testable idea, begins the process. Because we will treat our ideas as hypotheses, we need to consider what makes any given

idea truly testable. We will explore this as we look at the specifics of your idea and who it serves—information that is needed to design a rigorous test. Next, in Step 2, we define what constitutes evidence—what kind of data will tell us whether our hypothesis (and the assumptions behind it) is true or false? Where will we locate such data? Having specified what we are looking for, we then move on to the question of how to conduct our experiment. In Step 3 we sort through a variety of options to zero in on the best type of test to collect the desired data. What type of test best suits the particulars of our idea and the evidence we need to gather? Once we have selected the test type, in Step 4 we develop the stimulus we will use to provoke feedback by building the prototype, the simplest one that will do the job.

Prototypes and Pyramids

Prototyping has been around a long time! The Great Pyramid at Giza, constructed in 2528 BC, is the oldest of the Seven Wonders of the World and the only one that remains largely intact. Scholars believe that its construction, estimated to take decades—or even centuries—almost certainly relied on an iteraive ce ing fde i e inc ding drawings and schematics. “The wonders of ancient Greece and Egypt required engineering far in advance of their time, and marvels of engineering such as the Pyramids and the Parthenon didn’t sprout up out of the ground whole,” explains engineering expert David Walsh.*

* “The Top Four Ancient Design Prototypes” by David Walsh, 8/19/15, https://www.asme.org/topics-resources/content/ top-4-ancient-design-prototypes

Finally, with the key ingredients in hand—our testable idea (our hypothesis), our test type (our intervention), and our prototype (our stimulus)—we move on to the question of what we have learned in Step 5, as we execute our test plan, analyze the results it produces, and iterate our way to an improved solution.

But these steps are rarely as linear as our model suggests. Frequently we loop back to earlier steps. The selection of a test type, Step 3, and prototype format, Step 4, are usually highly interactive, and it is not uncommon to move back and forth in an iterative way to complete them. Sometimes our ideas fail in Step 5 and we loop back to Step 1 to start all over again with a new idea. Or maybe they succeed, so we loop back to Step 2 to design a more sophisticated test with a higherfidelity prototype and different evidence—or to test a different set of assumptions. And, of course, things happen before our Step 1—the process of generating the idea, for instance—while the essential activities around scaling an idea in the real world occur after our Step 5, when the period of experimentation ends and implementation begins.

As we move through these five steps and apply them to your idea, we will illustrate them in action with examples from our own experiences and research working with innovators at Nike, Whiteriver Hospital, the Project Management Institute, and South Western Railway. Because the ideas you will want to test come in different forms—products, services, processes, software, or a combination of these—our stories encompass this variety. Let’s preview these four stories, which we will return to time and time again.

Our Four Stories in Detail

Context

Nike is one of the largest and most successful athletic apparel companies in the world. Like many consumer c ged g d f e wanted to explore expanding into services that would allow them to diversify in ways that would develop deeper customer relationships.

Challenge

i e d iden ifed e e addressing: kids grew fast and wrecked shoes, so parents constantly needed to buy new ones to accommodate their kids’ growing feet. Shopping at a store with kids wasn’t typically enjoyable. In addition, though some old shoes were donated, many old pairs lay abandoned in musty closets, and Nike, with a commitment to sustainability, had invested in technology to recycle old shoe materials. So, they wanted to explore the idea of a shoe subscription service, Easykicks, for young athletes, a population they wanted to get to know and serve better. The new c i i n e vice e ed nnity to address both the hassle of shoe shopping and the need for recycling.

Process

Nike knew they needed help in testing the business model of services, like Easykicks, that had caused challenges for them in the past, so they partnered with ee n ig f n n c en in market experimentation. Peer Insight led a process using the steps outlined in i fe d de ign nd n e on the new subscription service over an 18-month period.

Whiteriver Hospital

Context

Whiteriver Hospital is located on the Fort Apache Indian Reservation, which covers more than a million acres and serves a population upward of 15,000, mostly Apache. It falls under the jurisdiction of the U.S. Department of Health and Human Services (HHS) and has both inpatient beds and an emergency room where most of the population’s medical needs have to be met, including e c i i n ef

Challenge

The hospital faced a serious situation: close to 25 percent of emergency department (ED) visitors were leaving without being seen, a problem attributed to long wait times. Nonemergency patients consistently got delayed as dd e ed e e e gencie i arrivals sometimes waiting as long as six hours before being seen. When potential patients left the emergency room (which they did at a rate twenty times the U.S. national average), their medical problems worsened. Often these eventually became true emergencies, and patients needed e e ic e ed e e e v i n more extensive and expensive care. The Whiteriver Performance Improvement Team wanted to explore the idea of adding an Electronic Kiosk, similar to one at Johns Hopkins Hospital in Baltimore, where a patient electronically signed in upon arrival, and the electronic system informed other parts of the hospital of that patient’s potential needs, saving administrative time and speeding up the intake process for patients so that more patients could be seen more quickly in the ED.

Process

HHS operated the Ignite Accelerator, a program of their IDEA Lab aimed at bringing new innovation approaches to employees across the United States. gni e e ed ed c i n c c ing and a small funding stipend, to boost ec e ed e e addressing agency problems. Whiteriver’s Performance Improvement Team was invited to test the electronic kiosk idea as part of the Ignite program, using a process similar to that outlined here.

Our Four Stories in Detail

South Western Railway

Context

SWR, a joint venture between two of the world’s leading rail companies, operates some of the busiest train routes in the United Kingdom, with 235 million passenger journeys a year. SWR faced challenges with industrial relations, major ne e i e and the passenger experience.

Challenge

SWR leadership saw an opportunity to improve the passenger experience. Unlike Nike and Whiteriver, SWR did not start with a solution in mind; they wanted to f e e e e enge faced in more detail. They commissioned the consultancy David Kester and ci e D e ding de ign f e e fnd ic in nd long-term strategy to improve the SWR customer experience. They captured the challenge: to learn fast from customers, id de ive c nfdence i ding ic at SWR stations, and together shape the future SWR experience.

Process

Using ethnographic research tools, DK&A ne ed i ee cen rail stations to better understand the problems and needs of their passengers. They synthesized this research, identifying insights and creating journey maps and personas. Based on these, they facilitated a collective brainstorming process iden ifed i e ide ee concepts emerged as particularly attractive: the Concierge i ed e ing warm welcome and providing the right infrastructure to better support customers in the ticketing process), Cleaner and Cleaner (a hygiene awareness and nudge campaign to provide reassurance to the public), and the Wayfnding Audit (addressing poor signage and visual clutter with smart customer-led tools). DK&A then embarked on an experimentation phase, using a process similar to the steps outlined here.

Project Management Institute

Context

PMI is the premier global project management association, with nearly 500,000 members. Like many membership associations, they struggle with how to evolve and continue to add value for their members beyond their core product, the Project Management Professional (PMP) ce ifc i n

Challenge

PMI had already done extensive research to better understand the problems and needs of their members, and had idenifed ig eve ide e g d ignifc n en i c e e v e for them:

1. Snippets: A microlearning platform that curated quick, bite-sized training nd e ce f e e day-to-day learning needs.

2. Career Navigator: A self-assessment tool that showed potential career paths based on a member’s experience, skills, and interests.

3. Hive: A peer-to-peer connection platform that allowed members to get answers to their toughest project management questions by connecting directly to experienced peers.

4. Spot: An experiential learning opportunity that matched members with real-world, low-risk opportunities to help them hone needed skills.

Process

Each idea was intriguing but untested, and with an important board meeting approaching, PMI leadership needed to make choices about which ideas to invest in. They partnered with Peer Insight to run a series of experiments over three n n e e ing n down which they should move forward. From there, Peer Insight ran a series of deeper market experiments over six months on the best performing concepts, g in ing fve e ce

Each of these organizations spent time to first understand the problems their customers faced, to ensure that they invested in solutions that solved problems that mattered. We will pick up their stories as each enters the experimentation process with particular ideas that they want to test. Though experimentation can be used to identify problems as well as to test solutions, in this field book we will focus on how to test a given solution—like Easykicks, the Electronic Kiosk, the Concierge, or Career Navigator—to see whether it does, in fact, meet an identified need.

As we get started, a note to you, our reader: we wrote this book under the assumption that you have an idea of your own that you want to test. Great! This is the most effective way to work with this book—to learn while doing—whether it's with an actual idea or one you want to use just to learn with. We will use your ideas to practice the learning-by-doing process in Steps 1 through 5. In each step, we will give you an assignment to apply what we’ve talked about to your own idea—so you’ll be both learning and doing as we go. We will ask you to note the milestones in your journey through the five steps using our Progress Tracker [TEMPLATE 1, PAGE 82] .

By the time we conclude, you will have followed your idea through one full cycle of testing.

Let’s get started exploring the world of experimentation!

Progress Tracker

Step 1

2

Ideas 1 Defining Evidence

Selecting Y T WHAT are we testing?

1 Defining Evidence Selecting Y T WHAT are we testing?

Frame a testable idea Prioritize concepts for testing

Complete a Concept Snapshot

Step 2

Surface critical assumptions Prioritize assumptions for testing

Establish what constitutes evidence and identify targets

Defining Evidence

Defining Evidence

he Prototype

Selecting Your Test 4 3 HOW are we testing?

Selecting Your Test 4 3 HOW are we testing?

the Prototype

are we learning? testing? 2 Executing Analyzing Iterating 5

the Prototype Selecting Your Test 4 3 HOW are we testing?

learning?

Step 3

Choose between Say and Do tests

Select a test type

Design your test

Step 4

Assess the level of fde i needed

Select prototype format

Construct your prototype

Step 5

Audit and pretest your design

Run your test e fnding

Iterate your concept

“This is one of those rare books that everyone should read. We are in the era of smart technology that will require every person and organization to learn, unlearn, and relearn at the pace of technological change. This book will help you become a faster, better learner.”

EDWARD D. HESS, PROFESSOR EMERITUS OF BUSINESS ADMINISTRATION, DARDEN SCHOOL OF BUSINESS, AND AUTHOR OF FIFTEEN BOOKS INCLUDING OWN YOUR OWN WORK JOURNEY! THE PATH TO MEANINGFUL WORK AND HAPPINESS IN THE AGE OF SMART TECHNOLOGY AND RADICAL CHANGE

“Liedtka, Chen, Foley, and Kester present the most detailed and structured articulation of the experimentation process I’ve seen outside of the sciences. The Experimentation Field Book allows readers to put these methods to practice immediately.”

GRAHAM HENSHAW, DIRECTOR OF THE ALAN B. MILLER ENTREPRENEURSHIP CENTER, WILLIAM AND MARY

“It used to be that we would tinker, try things out, prototype, and ask around and see how our vision of a new service formed. Forget how it used to be. Experimentation has grown up. The ideas and templates in

EDDIE OBENG, PROFESSOR OF INNOVATION AND ENTREPRENEURSHIP, HENLEY BUSINESS SCHOOL, AND INNOVATION LEAD AND EDUCATOR, PENTACLE: THE VIRTUAL BUSINESS SCHOOL

“Experiments are conducted every day in business to test if new ideas work before investing heavily in them. the way through, from framing the experiment and gathering the evidence for it to building the prototype, analyzing the results, and making the right adaptations. It’s sharp, succinct, and actionable, with four great case studies driving the narrative. And you won’t need to go near a science lab to learn the process!”

JEREMY MYERSON, DIRECTOR, WORKTECH ACADEMY, AND PROFESSOR EMERITUS, ROYAL COLLEGE OF ART

“With software and AI consuming more and more economic value, it’s not an exaggeration to say experiments

ALEX COWAN, AUTHOR OF

and organizations alike. Yet navigating the practicalities of adopting new technologies and facilitating successful change can be daunting. Enter The Experimentation Field Book tackling transformational goals, whether they are small-scale initiatives or larger endeavors. With this invaluable

STEPHEN JAMIESON, GLOBAL HEAD OF CIRCULAR ECONOMY SOLUTIONS, SAP

“The Experimentation Field Book covers the end-to-end process of experimentation in a very accessible way. An excellent and logical work.”

E E

UNIVERSITY OF CAMBRIDGE JUDGE BUSINESS SCHOOL