What Customers Want (PB): Using Outcome-Driven Innovation to Create Breakthrough Products and Services

by Anthony W. Ulwick

But after twenty years of customer-driven thinking, U.S. companies still find that 50 to 90 percent of their product and service initiatives are failures, collectively costing them more than $100 billion each year.

An agreed-on language is fundamental to success in any discipline, yet confusion has permeated product development because companies continue to define "requirements" as any kind of customer input: customer wants, needs, benefits, solutions, ideas, desires, demands, specifications, and so on. But really, those are all different types of inputs, none of which can be used predictably to ensure success.

Customers buy products and services to help them get jobs done. In our study of new and existing markets we find that customers (both people and companies) have "jobs" with functional dimensions to them that arise regularly and need to get done. When customers become aware of such a job, they look around for a product or service that will help them get the job done.

In the outcome-driven paradigm the focus is not on the customer, it is on the job: the job is the unit of analysis. When companies focus on helping the customer get a job done faster, more conveniently, and less expensively than before, they are more likely to create products and services that the customer wants.

Only after a company chooses to focus on the job, not the customer, are they capable of reliably creating customer value.

Customers use a set of metrics (performance measures) to judge how well a job is getting done and how a product performs. Just as companies use metrics to measure the output quality of a business process, customers use metrics to measure success in getting a job done. Customers have these metrics in their minds, but they seldom articulate them, and companies rarely understand them. We call these metrics the customers' desired outcomes.

For any given job, customers collectively apply 50 to 150 metrics (not just a handful) to measure how well the job is getting done.

Optimize their messaging strategy and exploit the advantages their current products have in satisfying the targeted underserved outcomes Prioritize projects in their development pipeline so they can quickly bring to market those products and services that do the best job of addressing other targeted opportunities Systematically devise ideas that address the remaining unexploited opportunities, creating valuable if not breakthrough products

In other words, companies need to figure out what jobs customers want to get done and how they measure success in getting a job done before they can determine what solutions customers want.

To formulate an effective innovation strategy, companies must correctly determine who in the value chain makes the most important judgments about value and go to them directly to understand what metrics they use to make those judgments.

To succeed at it, companies must discover which customer outcomes (the metrics used by customers to define the successful execution of a specific job) are being underserved and then devise and provide creative features in their products and services that do a better job of addressing those outcomes. Addressing one outcome may result in incremental improvement whereas a product that addresses many underserved outcomes is likely to produce a breakthrough improvement.

In contrast, with disruptive innovation, the technology exists, and the company is in search of a customer and an opportunity.

The company doesn't consider all relevant customers for innovation. Companies that are far back in the value chain (such as producers of raw materials or semiconductors that sell to OEMs) and those that sell directly to the end user (such as makers of computer hardware and software) often don't take the time to consider all relevant customers for innovation and therefore fail to capture or consider their inputs.

The company lets one customer speak for another. To formulate effective innovation growth strategies, companies must correctly determine who in the value chain makes the most important judgments about value and go to that source directly to understand what metrics they use to judge value.

Companies also like to rely on their sales forces for customer information, letting them speak for the customer—often another big mistake. Companies often believe that because the sales team is close to the customers, the sales people must know what customers want. But salespeople invariably talk about customer requirements as "solutions" and "specifications"—not as outcomes.

A critical objective of the innovation process is to discover as many opportunities for value creation as possible, so it's advantageous for businesses to consider multiple constituents and their underserved outcomes when devising their innovation strategies.

Software developers may want to consider the desired outcomes of end users, purchasers, IT administrators, and corporate executives.

For a company to innovate, it must create products and services that let consumers perform a job faster, better, more conveniently, and/or less expensively than before. To achieve this objective, companies must know what outcomes customers are trying to achieve (what metrics they use to determine how well a job is getting done)

When companies talk about customer requirements, they talk about customer needs, wants, solutions, benefits, ideas, outcomes, and specifications, and they often use these terms synonymously.

So the real question is, "What criteria do customers use to measure value?"

In a study we conducted with 270 business executives in a wide range of small, medium, and large companies across many industries, we found that 72 percent of the marketing and development managers interviewed were very satisfied with their company's ability to capture customer requirements.

customers tend to state (and managers tend to capture) four types of information during the requirements-gathering process: solutions, design specifica­tions, customer needs, and customer benefit statements.

The razor manufacturers, for instance, should realize that users who ask for a lubrication strip are really trying to "minimize the number of nicks when shaving." A lubrication strip is just one way in which this criterion could be satisfied.

Although this type of feedback may be appropriate in certain situations (among government contractors, for instance), accepting specifications as customer inputs inherently prevents engineers and designers from using their creative skills to devise breakthrough products and services.

Although these simple statements provide some indication as to what customers are looking for, they have one major drawback. They are imprecise statements open to interpretation and present designers, developers, and engineers with the impossible task of figuring out just what customers really mean by "durable" or "strong."

Good customer input must be concise, actionable, and measurable.

In one study we conducted with Motorola cell phone users, for example, we found there were twenty-one different ways in which customers defined "easy to use." Their desired outcomes included, "minimizing the time it takes to look up a needed phone number," "minimizing the likelihood of calls being made by inadvertently hitting the keypad," and "minimizing the time it takes to dial a number without looking at the keypad." In each of those situations, "easy to use" takes on a different meaning and has different implications for designers seeking to improve or extend the product.

To execute their innovation processes successfully, companies must obtain three distinct types of data. They must know which jobs their customers are trying to get done (that is, the tasks or activities customers are trying to carry out); the outcomes customers are trying to achieve (that is, the metrics customers use to define the successful execution of a job); and the constraints that may prevent customers from adopting or using a new product or service.

Customers are often trying to perform multiple tasks simultaneously. But companies tend to focus their products and services on performing a single task; to address ancillary as well as primary jobs would require them to develop new or different competencies or to cross organizational boundaries—something they may not believe they are prepared to do, financially or culturally.

We have found that there are three different types of jobs that customers are often trying to get done in a given circumstance: functional jobs and personal and social jobs (two types of emotional jobs).

Functional jobs define the tasks people seek to accomplish, personal jobs explain the way people want to feel in a given circumstance, and social jobs clarify how people want to be perceived by others.

These are just a few of the eighty or so outcomes they captured from customers. For most jobs, even those that may seem somewhat trivial, there are typically 50 to 150 or more desired outcomes—not just a handful.

In fact, in a typical one-hour interview, it is possible to capture between twenty and thirty desired outcomes from a single customer.

Years ago, we tried to be more colorful and used a combination of words to denote "minimize"—for instance, "reduce" and "eliminate" and "prevent." But when we asked customers to rate the importance and satisfaction levels of specific outcomes, we found that the word used to introduce the statement affected the way the statement was rated, introducing unwanted variability into the process.

As we said earlier, the method used for gathering customer requirements is not as important as knowing what type of information you want from customers (jobs, outcomes, and constraints) and working to obtain them.

The interviewers asked the agents, "What makes one insurance company's offering better or worse than another, and why? And what characteristics describe the ideal service offering?" By forcing this type of discus­sion, the interviewers were able to prompt the agents to reveal their desired outcomes.

In the outcome-driven paradigm, an opportunity for growth is defined as an outcome, job, or constraint that is underserved.

In any of those scenarios, companies must rely on opinion, intuition, and guesswork, and as a consequence they often make three common missteps: making unnecessary improvements by targeting outcomes that are already well satisfied, making improvements that satisfy unimportant outcomes, and making improvements that, while satisfying certain outcomes, inadvertently have a negative effect on other, more important outcomes.

The concept of a low-end disruptive innovation, as described in The Innovator's Solution, is only possible when the customer population, or a segment of that population, is overserved.1

remember that the customers' outcomes are stable over time. People who shave, for example, have always wanted to minimize the number of nicks, minimize shaving time, and minimize the number of passes that must be made. These and other shaving-related desired outcomes will remain the same for years to come. What does change, however, is the degree to which these outcomes are satisfied by new technologies and product and service features.

In addition, an outcome-driven competitive analysis makes it possible for a company to: Identify competitors' strengths and weaknesses Know which of its competitors' features to emulate Know when competitors are headed in the wrong direction and should be left to go it alone.

Once a company learns which competitor does best at addressing a key opportunity in the market, it can attempt to emulate that competitor.

In effect, Nortel's market and product strategy, resource selection, and capabilities were being dictated by a vertical-segment classification it had initially chosen for sales, marketing, and accounting purposes.

an effective segmentation scheme must create a population that: Has a unique set of underserved or overserved outcomes Represents a sizable portion of the population Is homogeneous—meaning that the population agrees on which outcomes are underserved or overserved and responds in the same manner to appropriately targeted products and services Makes an attractive strategic target—one that fits with the philosophy and competencies of the firm Can be reached through marketing and sales efforts

In most markets there exists a group of customers who are more demanding than the rest. They are underserved along many dimensions of value; they want more and are willing to pay for it. This segment may be less than 5 percent of the total market or it may be 20 percent or greater. A company benefits from knowing if this segment exists and how large it is.

An effective targeting strategy will enable a company to add function and performance (but not necessarily cost) in areas that are underserved and to reduce cost and function in areas that are overserved. Product and service offerings must deliver all the performance that can be absorbed but no more, so that customers are not paying for function they do not need. That is, companies must find that fine balance between delivering too much and too little function.

Our experiences have led us to identify five types of broad-market opportunities, each of which represents a unique avenue for growth and innovation: Related opportunities that form a theme Unrelated opportunities that represent growth avenues A single opportunity that can be addressed with a new, ancillary product Overserved outcomes that add unnecessarily to product cost Opportunities for technology development and long-term growth

Many software applications suffer from the problem of overserving the outcomes for which they were originally designed. In the software industry, however, the overserved outcomes rarely get addressed because the unneeded function does not add to the cost of the product. Because software is unique, companies in this industry need to consider whether or not the unneeded function is negatively impacting the user's ability to get a job done. If it is, the developer may be able to charge more for a simpler version of the same application.

When AIG used outcome-based segmentation to segment their agents, for example, they found one very demanding segment (segment 3) and two other underserved but less demanding segments. AIG targeted the least-demanding segment first (segment 1) because they were able to satisfy all the underserved outcomes in that segment with only a handful of new service features. Along the way, those same features helped AIG make some progress in addressing the underserved outcomes in segment 3.

We would argue that acting on the data is the only logical option, but we have on rare occasion worked with managers that have spent the company's time and money gathering outcome-driven information only to refute and ignore the data in the end. Why? Based on our subsequent undercover work in this area, we have found that managers choose to invalidate data when: They do not get the "right" answer back. We find that managers will sometimes commission research to help them validate an idea they have been backing and want to pursue. If the research fails to validate their hypothesis, they may decide to refute or ignore the data. They feel the results threaten their jobs. In a couple of situations we have seen mangers refute the results because accepting them would place the managers in a bad light and might result in the loss of their jobs. This is most likely to happen when the research reveals opportunities in areas that others have suspected all along, but that the managers themselves have failed to identify. They naturally want to protect their reputations, but doing so may be at the company's expense. Addressing the opportunities would require the development of a new competency. We also find that managers will refute the results when accepting them would force the managers to develop new competencies within the organization, and the managers are not interested or are unwilling to take on that added responsibility.