1. Introduction: The Billion-Dollar Flop
In 1984, Tony Ulwick was a product engineer at IBM working on the PCjr—a machine the Washington Post predicted would become “the standard by which all other home computers are measured.” Instead, the day after its release, the Wall Street Journal branded it a “flop.” It was a billion-dollar embarrassment that proved even the world’s most sophisticated R&D engines were effectively guessing.
This failure revealed a structural misalignment that still plagues the industry today. Research from Simon-Kucher & Partners shows that 72% of new products fail to meet expectations. Even more damning, data from Frost & Sullivan suggests that a staggering 99% of new products fail to even recoup their development costs. These are not merely bad streaks of luck; they are the result of a fundamental misunderstanding of the innovation process. For decades, innovation has been treated as a hit-or-miss craft—an art form fueled by “epiphanies.” To achieve predictability, we must transition to a science of understanding “Jobs.”
2. The Definition Crisis: 95% of Teams are Lost
The industry is paralyzed by a crisis of definition. An estimated 95% of product teams cannot agree on what a “customer need” actually is. Marketing, R&D, and strategy teams operate in internal chaos, speaking different languages while pursuing a singular goal.
While some attempt a “Needs-First” approach, it is often flawed in execution because it is treated as an interpretative art. Firms rely on anthropologists or focus groups to seek out “epiphanies” using the customer’s own words. This results in the wrong inputs because it lacks a rigorous syntax. Without a shared vocabulary and a strict linguistic structure for needs, companies fall back on the “Ideas-First” trap: brainstorming solutions and hoping they find a problem to solve.
“People don’t want to buy a quarter-inch drill. They want a quarter-inch hole!” — Theodore Levitt
As Levitt’s classic adage suggests, the product is an ephemeral solution to a stable problem. Jobs-to-be-Done (JTBD) Theory provides the framework to define these problems with scientific precision before a single idea is ever generated.
3. The Math of Failure: Why Brainstorming is a Lottery
The popular “Ideas-First” mentality treats innovation as a numbers game, assuming that more ideas increase the probability of a breakthrough. Mathematically, this is a fallacy.
In any given market, a customer has between 50 and 150 desired outcomes. If we assume a market has only 15 unmet needs and we generate just three competing ideas for each, the probability of an idea addressing all 15 needs by chance is 1 in 14 million. Expecting a team to “brainstorm” its way to a hit is like expecting a sharpshooter to hit a target they cannot see, or a doctor to prescribe a cure without knowing the symptoms.
Without knowing the specific metrics (Desired Outcomes), adding more ideas simply increases the noise and ensures the filtering process will fail. The “failing fast” mantra—testing countless concepts to weed out the bad ones—is often just a justification for “doing something bad faster.” If your metrics are unknown, your filters are broken.
4. Needs are Metrics, Not Solutions
To move from guesswork to a science, we must adopt a new unit of analysis: the Desired Outcome. These are the metrics customers use to measure success when getting a job done. A properly structured outcome statement is devoid of technology and includes:
- A direction of improvement (e.g., minimize)
- A performance metric (e.g., time or likelihood)
- An object of control
- A contextual clarifier
Example: “Minimize the likelihood that the music sounds distorted when played at high volume.”
While solutions change, jobs and their 50–150 associated outcomes remain stable. The “job” of listening to music has not changed in fifty years, despite the shift from vinyl records to streaming services.
“Innovation can be far more predictable—and far more profitable—if you start by identifying the jobs that customers are struggling to get done.” — Clayton Christensen
“Tony has turned innovation into a science. I call him the Deming of Innovation.” — Philip Kotler
5. Stop Asking Customers for Solutions
A fundamental tenet of the Outcome-Driven Innovation (ODI) process is that it is not the customer’s job to be the engineer or the scientist. Asking a customer “What do you want?” is an abdication of responsibility.
Customers are not equipped to articulate “latent needs” or envision breakthrough technologies. However, they are experts in their own metrics for success. Customers should be “hired” only to provide the requirements—the desired outcomes—while the company remains responsible for the solution. When innovators stop looking for “unvoiced” technology requirements and start looking for articulated metrics, the myth of the unarticulated need disappears.
6. Demographics are Dead: The Case for Needs-Based Segmentation
Traditional demographics (age, geography, gender) are the primary creators of “phantom targets.” Marketing to a 28-year-old in Montana and a 55-year-old in Florida as different segments is a mistake if they both share the same unmet outcomes regarding their internet connectivity.
Real segments are defined by “added complexities” that make a job harder for one group than another. For example, in research for Bosch, ODI did not find segments based on age or job title. Instead, they discovered a segment of tradesmen who struggled specifically with the frequency of blade height and angle adjustments when making finish cuts. This functional complexity—not a demographic profile—defined the segment and allowed Bosch to target a high-profit, underserved group with surgical precision.
7. Conclusion: From Guesswork to Predictability
The shift from an “Ideas-First” model to the “Jobs-to-be-Done Needs Framework” represents the professionalization of innovation. We no longer have to settle for the 17% average success rate that has characterized the industry for decades. By making the job the unit of analysis and using the Outcome-Driven Innovation (ODI) process, that success rate climbs to 86%.
The JTBD framework brings order to a historically chaotic practice by identifying stable jobs, capturing a complete set of 50–150 outcomes, and segmenting the market based on where customers actually struggle.
If your team can’t agree on what a ‘need’ is today, what are the odds your next product launch is actually solving one?