Outcome Driven Innovation
Tony Ulwick has done significant work developing an approach to quantify and pinpoint underserved needs. His company, Strategyn, pioneered a method called Outcome Driven Innovation (ODI), which offers a complete process for finding strategic opportunities. They believe that if you know how the customer measures value, you can provide successful solutions in a measured, controlled way.
In ODI, needs are synonymous with desired outcomes. The formulation of desired outcome statements is specific and precise.  There are four elements in each statement:
- Direction of change: How does the job performer want to improve conditions? Each desired outcome statement starts with a verb showing the desired change of improvement. Words like minimize, decrease, or lower show a reduction of the unit of measure, while words like maximize, increase, and raise show an upward change. (Note that minimize has been found to be more precise language since people can imagine what zero looks like. Maximize is fuzzier, without a bounding limit.)
- Unit of measure: What is the metric for success? The next element in the statement shows the unit of measure the individual wants to increase or decrease. Time, effort, skill, and likelihood are a few typical examples. Note that the measure may be subjective and relative, but it should be as concrete as possible.
- Object of the need: What is the need about? Indicate the object of control that will be affected by doing a job.
- Clarifier: What else is necessary to understand the need? Include contextual clues to clarify and provide descriptions of the circumstance in which the job takes places.
Here are some examples:
- Minimize the time it takes to summarize conference insights—for example, as notes, presentations, and reports.
- Reduce the time it takes to get ingredients ready when preparing a meal.
- Minimize the time it takes to gather documents when preparing taxes—for example, pay stubs, expenses, and receipts.
The normalization of your desired outcomes allows you to quantify their relative importance. After extensive qualitative research to derive the desired outcomes, you can launch a survey to pinpoint your opportunity from the job performer’s perspective. Note that finding unmet needs using ODI is a precise process that requires rigorous investigation. I encourage you to learn more about the method as practiced by Strategyn by following the resources cited in this book. A simplified overview of the basic approach follows:
Step 1: Gather all desired outcomes.
From your qualitative research with job performers, identify a complete set of needs. You’ll know when you’ve uncovered all needs when you start hearing the same ones over and over. This may take a dozen or two interviews, or more in complex situations. In order for the process to be effective, it needs to be comprehensive and as granular as possible.
Step 2: Formulate desired outcome statements.
Check the form of each statement to ensure consistency. Have others in your team review and refine the statements to accurately reflect what job performers expressed in their interviews. Remove redundant statements and fill in the missing gaps.
Step 3: Survey job performers.
Create a survey that consists of desired outcome statements as the main items. Pair each statement with two scales: one for importance and one for satisfaction. Figure 4.2 shows a survey for quantifying underserved needs rates each statement for importance and for satisfaction, which illustrates the basic arrangement for each questionnaire item based on the example job, file taxes.
Administer the survey to job performers. The broader the domain of your main job, the more participants you’ll need. The low end of your sample size should be more than 150, extending into the thousands from there. But a good rule of thumb is to have at least twice the respondents as the desired outcome statements.
Step 4: Find opportunities.
Following ODI, calculate your opportunity score by pinpointing underserved needs. First, find the satisfaction gap, or importance minus satisfaction. Then add the satisfaction gap to the overall importance score, as shown in Figure 4.3. The resulting number will fall on a range of zero to 20. The higher the number, the bigger the opportunity will be.
You may also find it helpful to plot the scores of the needs statements on a graph like the one shown in Figure 4.3. This will show you a distribution of rankings visually.
While finding underserved needs following ODI might seem straightforward in principle, it’s difficult to execute in practice. First, ODI requires that all desired outcomes in getting a job done must be collected. Second, the survey requires a large sample size in order to get valid results. Finally, it’s challenging to get respondents to answer all of the questions with their full attention, and they must be incentivized to do so, making the process costly.
That being said, cutting corners is also problematic. If you don’t have a complete set of consistently formulated needs and don’t survey the right sample of people—namely, actual job performers—the results will be unreliable or, worse, misleading.
An alternative to prioritizing needs is to prioritize steps of the job process. Take the steps from your job map and survey them on the steps, rating each for both importance and satisfaction. Then plot each step on a simple 2 x 2 matrix based on the scores each received. From this, you can see underserved job steps instead of underserved needs. The matrix provides insight into what part of the process has the most strategic opportunity.
Dan Olsen, consultant and author of Lean Product Market Fit, independently developed a very similar approach using an importance versus satisfaction matrix.  For example, let’s say you were trying to find opportunities for Uber, the car ride service. Olsen recommended that you first find all of the service demands that people might have—for example, clean cars and timely pickups.
Similar to ODI, the next step would be to survey people who had used Uber by importance and satisfaction. But instead of a scale of 1–10 for both, Olsen suggested using a different scoring system. Importance should be placed on a scale of 1–5 and satisfaction on a scale of 1–7. Importance could best be seen as a polar scale. That is, the concept of negative importance didn’t really make sense. The scale of 1–5 went from low to high. Negative satisfaction, however, was possible. So, for satisfaction, Olsen used a Likert scale with the middle value equaling a neutral state.
Dan Olsen prioritized product needs by importance and satisfaction. To place the resulting scores on a matrix, he converted each to a percentage, as seen in Figure 4.4.
Starting with an existing solution in mind, Olsen’s technique lies within the solution space of JTBD. Importance and satisfaction were based on features or service aspects. From this perspective, the approach was closer to the voice of customer research, which considered feedback from customers about a given product or service, rather than JTBD. Still, there was a common point of attempting to find opportunity by ranking importance and satisfaction for each need.
Interestingly, Olsen suggests that the importance versus satisfaction matrix can be used with a sample size of zero. In other words, after doing research to gather the needs from job performers, a team can simply make assumptions about the relative position of each need on the graph. From that analysis, possible solutions can be conceived and then tested in experiments, tying JTBD to Lean methods. Overall, market feedback comes from reacting to a prototype rather than to a desired outcome statement.
For instance, imagine you worked for a startup that intended to help people take medication at a specific time of the day. After qualitative research uncovered the need, you could take those needs and prioritize them on the importance-satisfaction matrix relative to each other making assumptions. Then you could form a hypothesis you believed to have the most opportunity and conduct experiments to test that hypothesis with solutions. For example, you might want to see if electronic reminders on a mobile device would address the need.
In another related approach outlined in the book The Innovators Guide to Growth, Scott Anthony and his coauthors showed that marketplace opportunity comes from understanding the customer job. 
“To identify opportunities to create new growth, look first for important jobs that people can’t get done satisfactorily with current solutions.”
This doesn’t just apply to creating new offerings. JTBD can help revitalize growth in existing businesses. The authors wrote:
“Jobs-based thinking can restart growth by helping companies shake up commoditized markets and highlighting opportunities to revive even the most moribund of products.”
One specific method to evaluate opportunities from a jobs-based view of the market is the jobs scoring sheet. After doing primary research to uncover jobs, including interviewing and observations, Anthony and his colleagues recommended prioritizing jobs along several dimensions. For each job, ask these three questions and rate each on a scale of 1–5:
- Is the job important to the customer? (1 = not important, 5 = very important)
- Does the job occur relatively frequently? (1 = infrequently, 5 = very frequently)
- Is the customer frustrated by the inability to get the job done with today’s solutions? (1 = frustrated, 5 = not frustrated)
Then create a score for each job statement using the following equation: (importance) + (frequency) x (frustration). Finally, rank each job statement by its score. The higher the score, the greater the opportunity.
Table 4.1 reflects a simple worksheet for following their approach. It’s best filled out based on survey data from job performers, but can also be used as a worksheet for teams to reflect on opportunities based on their existing knowledge and understanding of customer jobs.
Finally, there are even more informal ways to use the type of thinking outlined in this play. For instance, I have used informal ways to work with unmet needs based on the importance-satisfaction matrix. During one workshop I conducted, the group was required to first generate need statements based on previous research. Then we prioritized them only by assumed importance first. From the cluster of needs that were deemed to be more important, I instructed each group to decide which they felt were most unsatisfied. Teams then selected a need and ideated solutions around it.
Play: Create Goal-based Personas
Personas are archetypal representations of users. They are used in various areas of a business to instill customer-centric decision-making—from marketing to sales to design. A persona is essentially a communication tool—a way to summarize information about customers in a way that is accessible to everyone on a team.
Some practitioners within the JTBD community have called to abandon personas. They point out the flaws in personas that are based on demographic, psychographic, behavioral, and/or attitudinal schemes. But while these aspects may be central to traditional marketing approaches, they don’t inform innovation and design.
It’s a misconception, however, that JTBD replaces the need for personas entirely. In fact, given the multiple actors in any given JTBD ecosystem, it stands to reason that personas can represent the various roles. What’s more, not all job performers are the same, and personas can be used to illustrate different types of performers.
To help understand how to create personas based on people’s goals, the best place to start is the work of Alan Cooper. His approach, introduced in his book The Inmates Are Running The Asylum,  bases personas not on demographics but rather their intended outcomes. He wrote:
“Personas are defined by their goals.”
In his follow-up book, About Face 2.0,  Cooper presents complete instructions for creating personas as part of his Goal-Directed Design (GDD) method. Although GDD and JTBD were developed independent of one another, there is overlap. Both focus on the goals that people have. However, GDD is a design method specifically created to design software user interfaces. As a result, the approach refers to users of a predetermined system and operates in the solution space.
Nonetheless, Cooper’s approach serves as a good basis for creating goal-based personas, as summarized in the following steps:
Step 1: Interview users.
Personas are based on qualitative interviewing, similar to the interview technique outlined in Chapter 3, “Discovering Value.” In fact, you can interview for both jobs and personas in the same session. Rather than asking users about their preferences or desires, focus on their intent, as well as what frustrates them and what success looks like.
A dozen hour-long interviews usually suffice. In more complex situations, it may be necessary to interview two or more dozen people. The aim is to get enough feedback from enough people to find clear patterns. You’ll know when you’ve done enough interviews when you can start predicting how people will respond.
Step 2: Map interviews to variables.
Following Cooper’s approach, the next step is to map interviews to so-called behavior variables. Think of variables as needs with two end points that create a range. Variables correspond approximately to the circumstances in JTBD, but can include other factors as well. For online shopping, for example, the variables might be frequency of shopping, degree of enjoyment, and price vs. service orientation.
Place each interview on the range you create relative to each other. After conducting research, map behavior variables of all participants. Figure 4.5 shows how a set of five interviews might be mapped to just two variables uncovered during research.
Step 3: Identify patterns in goals.
After mapping interview participants to the variables, look for clusters of similar behavior. A group of participants that cluster similarly across six to eight variables represent a significant behavior pattern.
This isn’t a statistical exercise, but rather the search patterns in causation between the variables. In the example shown in Figure 4.5, for instance, participants 1 and 4 both shop for necessity and are price conscious. It stands to reason that these goals influence each other.
Step 4: Describe the resulting persona.
For each cluster, create a separate persona similar to the example in Figure 4.6. Goal-based personas include minimal demographic information and focus instead on goals and frustrations. The basic description should include their most important goals and their relevant circumstances. You should reflect a typical workday or circumstances when reaching the goal, as well as workarounds and frustrations. Avoid adding irrelevant detail: just one or two personal aspects can make a persona feel real. Focus on the patterns of behavior instead. In the end, the personas can be used to prioritize or focus efforts to create an appropriate solution.
Overall, Cooper’s approach provides a standard for creating goal-based personas. However, because the method looks at users of a predetermined system, it also falls within the solution space. What’s more, as a specific software interface design method, GDD has had little influence outside of design communities.
Instead of differentiating based on behavioral variables, use the circumstances you uncover during your research. Frame each as a range and map interviews to them. Then find different clusters of job performers. The process is similar to the one outlined previously.
- Interview job performers. Instead of interviewing users, interview job performers. As you uncover their jobs and needs, you can also determine the different circumstances that matter most.
- Map interviews to circumstances. Map the interviews to the circumstances. Rather than a range, as in GDD, simply take inventory of which interviews correspond to each circumstance.
- Find patterns. Look for logical clusters of circumstances that show causality.
- Describe the persona. Use a similar format to the persona shown in the previous example.
In addition to circumstances, Steven Wunker, Jessica Wattman, and David Farber recommend including two other variables they call drivers in their book Jobs to Be Done.  They recommend considering attitudes, or the personality traits relevant to getting a job done, as well as people’s backgrounds, or long-term context that affects decision-making. Together, these drivers—circumstances, attitudes, and backgrounds—form the basis of segmenting individuals into different segments based on their goals.
It’s also possible to create personas for roles other than the job performer. Chief among these is a buyer persona. You might also find that personas are needed for approvers, technicians, or people who are beneficiaries of job outcomes.
To speed up the process, try creating proto-personas, a quick format for creating personas that can be used in a group exercise. Also referred to as assumption personas, these are depictions of job performers based on what you know today.
Proto-personas can be created on a simple grid with four squares, as shown in Figure 4.7. There are four main elements:
- Name and sketch: Pick a name that is memorable and include a sketch
- Demographic details: Include factors that are relevant to the main job.
- Behaviors: Indicate the key actions the person takes when performing the job.
- Needs: Indicate what the person needs and what difficulties he or she might encounter.
Use the proto-persona format in a collaborative group session.
Overall, the success of using personas can vary. The key is not only to base them on goals, but also to make them visible and interactive. Involve your team in their creation for greater buy-in. Otherwise, it’s not uncommon for personas to get shelved and have limited utility across time. Don’t fall into this trap, and make your personas relevant and actionable.
Play: Compare Competing Solutions
Traditional competitive analysis almost always involves a technical comparison of product specifications and features, yet the analysis is conducted without knowing how customers measure value or how much value that competing features deliver to the customer. This is the problem and the myth that misleads: companies are not competing against other companies or their products. They are competing for the customers, and their one goal is to create value for them. And there is only one way to do that: by offering a product or service that is better than any other at helping them get their jobs done.
Part of the problem, too, is a myopic view of the competition. Just consider this quote from Scott Cook, the founder of Intuit: 
“The greatest competitor [to tax software] we saw was not in the industry. It was the pencil. The pencil is a tough and resilient substitute. Yet the entire industry had overlooked it.”
Think about it: tax software doesn’t just compete with other tax software solutions. It competes with everything a customer uses to get a job done. When preparing taxes, there are many subcalculations and figures you have to bring together. Grabbing for a pencil is often the easiest solution. So from a JTBD perspective, tax software competes with pencils.
Since JTBD is solution agnostic, it enables a new way of viewing competition. Rather than considering providers in a defined industry only, it’s possible to compare across product categories.
Companies typically conduct competitive analysis to make sure their products and services are better than those offered by competitors. But reacting to feature-to-feature comparisons is no guarantee of success. Competitive analysis, when seen through a jobs-to-be-done lens, is not about head-to-head comparisons. Instead, it’s about assessing how much better or worse a product is at helping the customer get a job done.
For instance, in his famous milkshake story, Clayton Christensen discusses different forms of breakfast that people might get while commuting to work. He shows that milkshakes perform better where others fail, such as a bagel, banana, and Snickers.
Typically, milkshake providers would compare their products to other milkshake providers. They’d look at the milkshake itself and make theirs thicker or thinner to differentiate it. But thickness doesn’t apply to bananas, Snickers bars, and bagels.
With JTBD, you can compare products and services to user needs. This gives a different view of the competition, and it steps outside of defined product categories. Looking at just a small sample of possible needs for getting breakfast on the go, it’s easy to see how well each one performs, as shown in Table 4.2.
While this approach may seem straightforward on the surface of things, it’s more difficult to complete in reality. First, deciding what needs to compare is tricky. Ideally, you will compare all of the needs you’ve uncovered. Alternatively, focusing on underserved needs only will help narrow the list. Second, in many cases there may be an endless number of solutions to compare. Deciding which to include requires a strategic selection.
Take a step back from your solution and product category, and compare the competition as your market sees it: by getting a job done. Here’s how it works:
Step 1: Determine alternatives to compare.
Start with your main job and job map. Then list all of the means by which people get the job done at each stage in the process. From this list, select the solutions that are most relevant to compare and place them across the top of the row of the table as column headers. Note that you can do this before you even have a product or service of your own on the market.
Step 2: Determine the needs to compare.
It’s possible to use the full set of outcomes to do a competitor comparison, but typically you’ll focus on a subset, such as the top unmet needs. List these on the left column of the table.
Step 3: Rank how well each solution meets those needs.
Ideally, you’d survey job performers about each of the needs you selected to compare. Set up a questionnaire similar to the one described previously for targeting underserved needs: for each need statement, provide scales for importance and for satisfaction. Then calculate the opportunity score for each and enter that into the table.
If surveying job performers isn’t possible, estimate how well each one meets the needs in a team exercise. Rank the needs with low, medium, or high, and negotiate a response for each together. Just be aware that without feedback from job performers, your rankings are speculative.
Step 4: Find your sweet spot in the competitive landscape.
Determine the needs that are missed by others or where you perform better. The aim isn’t to find a feature that hasn’t been built, but rather what needs are being underserved. When designing a new solution or when improving an existing one, use this insight into which opportunities to tackle first.
Instead of comparing need statements, compare steps in the job process. The process is identical to the comparison outlined previously, using the steps in the job process as the basis for comparison.
Table 4.3 illustrates what a comparison might look like with a subset of all steps for the job— prepare a meal. It compares three services: manually preparing a meal from scratch, using precut ingredients—for example, chopped vegetables—and full meal kit services with portioned ingredients. The relative level of satisfying each step is indicated with a simple low, medium, or high.
From just this simple comparison, it’s clear that both preprepared ingredients and meal kit services help customers in the earlier steps of getting the job done—more so than preparing a meal from scratch.
It’s also possible to graph your comparison to show advantages and disadvantages clearly. For instance, I previously worked for a provider of online content. At the time, we were trying to understand why people preferred print resources over digital resources. From interviewing dozens of people, we found a key set of needs that reflected the difference.
On a graph, we showed how the needs compared in terms of being met or not met by each solution. Figure 4.8 shows a modified version of that graph that can help compare advantages based on needs. We then hypothesized about what would need to be true for people to use our online content solutions more, labeled “New Online Experience” in the diagram. It became clear that we needed to cover more of the needs that people had reading and annotating documents, as well as comparing sources and navigating them, seen on the right side of Figure 4.8.
Play: Define a Jobs-based Value Proposition
A value proposition is a promise an organization makes to its customers. Although often associated with a specific statement, a value proposition is much more than a marketing device. It ultimately explains the benefit provided to the customer.
Value in this context is a perceived attribute. What a company thinks is very valuable may not be seen as such by consumers of its products or services. JTBD helps pinpoint those aspects that people value most, namely, meeting their needs.
Value Proposition Canvas
Alexander Osterwalder and his team at Strategyzer have developed a systematic approach for arriving at a value proposition. The Value Proposition Canvas (VPC) is designed to facilitate discussion around value propositions and align them to customer goals. Overall, the intent is to match the value you intend to create to the value that customers perceive as most beneficial. Accordingly, the VPC has two parts, reflected in the diagram in Figure 4.9. 
On the right is the customer profile with three components: the customer jobs, their pains, and their gains. Together, these elements represent the customer value profile.
The left half of the canvas details the features of your value proposition. There are three elements: your products and services, pain relievers, and gain creators. Together, these aspects reflect an understanding of the offering you’re providing.
Note that Osterwalder doesn’t provide specific guidance on how to uncover and formulate jobs to be done, nor on how to prioritize them. His process is much more intuitive. Using some of the techniques in this book, however, you can come to the VPC with key jobs and needs in hand.
By mapping the left side to the right side—the solution space understanding—you can make explicit how you are creating value for your customers. When the pain relievers and gain creators correlate to the pains and gains of your customers, you have a potentially strong fit. The process is collaborative and fairly straightforward.
Step 1: Understand the customer profile.
Start by discussing your customers and the jobs they want to get done. List the main job along with crucial social and emotional aspects. In the end, you should have a limited list of jobs here, about two to five on average.
List and discuss more important pains, or the negative aspects that people have in relation to getting the job done. These are the frustrations, challenges, risks, and struggles you uncover in your research. Avoid talking about specific solutions and discuss the general pains that people have. Focus on the most severe pains, typically the top dozen or so.
Gains are not simply the opposite of pains. They are the positive benefits people want to have from completing a job, including positive emotions, surprises, and even ambitions. List the most relevant desired outcomes that people have when performing the job independent of a given solution.
Typically, you can fill out the right side of the VPC intuitively, based on your prior research. However, your team may find more interviews are necessary to get more granular on the gains and pains. Fill in any gaps in knowledge with additional investigation.
Step 2: Discuss the solution profile.
Start with the products and services you offer or plan to offer. Then make it explicit how your products and services might alleviate the pains you identified. Consider the entire job process here—before, during, and after getting a job done. Next, consider how your products and services create customer gains and list them under Gain Creators.
Step 3: Ensure fit between the customer and the solution.
Map the pain relievers and gain creators to the pains and gains respectively. This shows which of the pains you’re addressing, and which gains you are building. Draw lines connecting them as your team discusses each point.
According to Osterwalder, you have problem-solution fit when the features of your value proposition—on the left—match the characteristics of your customer profile. In other words, mapping left to right on the VPC shows problem-solution fit.
If you’ve done Switch interviews, use your Forces analysis to determine the pains and gains. The factors that push customers toward a new solution translate well to gains, and the forces pulling customers back represent pains.
Looking at the other half of the canvas, you can then gauge potential product-market fit. Validate this match with the market. Often, your assumptions aren’t true, and real market feedback shows weaknesses in the value you propose to create. To ensure product-market fit, conduct experiments to validate your value hypothesis.
Step 4: Form a value proposition statement.
After validating the fit of your value with the market, encapsulate its essence in a single statement. There are many formats for doing this, usually similar to the following pattern:
- For [target job performers]
- Who are dissatisfied with [the current alternative]
- Our solution is a [product or service]
- That provides [key problem-solving capability]
- Unlike [the product alternative]
Overall, the VPC clarifies the value that an organization provides with JTBD as a starting point. Ultimately, people value getting their job done. Aligning to the job from the beginning helps ensure that it is the value that people really want.
JTBD provides a framework and common language for translating insights from job performers into a model for action. The process begins with field research, but then moves into more precise modeling, both qualitative and quantitative. Before creating solutions, define the value you’re going after. There are several ways you can do that.
 Anthony Ulwick and Lance Bettencourt. “Giving Customers a Fair Hearing.” MIT Sloan Management Review, April 2008. Note that Ulwick has refined his approach since publishing this article. For instance, within the ODI method, desired outcome statements now show only a downward or minimizing direction to provide more accurate responses.
 Dan Olsen. The Lean Product Playbook. Hoboken: Wiley, 2015.
 Scott Anthony, Mark Johnson, Joseph Sinfield, and Elizabeth Altman. The Innovator’s Guide to Growth. Boston: Harvard Business Press, 2008.
 Alan Cooper. The Inmates Are Running the Asylum. Indianapolis: SAMS, 1999.
 Alan Cooper and Robert Reimann, About Face 2.0: The Essentials of Interaction Design. Indianapolis: Wiley, 2003.
 Stephen Wunker, Jessica Wattman, and David Farber. Jobs to Be Done: A Roadmap for Customer-Centered Innovation. New York: AMACON, 2016.
 Quoted in Scott Berkun’s book The Myths of Innovation. Sebastopol, CA: O’Reilly, 2007.
 The Value Proposition Canvas is reprinted here with permission from Strategyzer. Learn more on their Web site.