The Design of Business

A shampoo brand gave design thinking somewhere to operate

In 2004, a Procter & Gamble team stepped away from its ordinary jobs and into a refurbished building on Clay Street in Cincinnati. Herbal Essences had lost momentum, and the assignment was to revive it. For roughly ten to twelve weeks, people from different functions worked together on what the brand might mean to a younger generation. A young intern's interpretation of “organic,” recalled in a later participant interview, helped the group see that its inherited category did not match the consumers it hoped to reach.¹

The idea did not travel from workshop to shelf by inspiration alone. The 18-month restage joined consumer visits, marketing, chemistry, package engineering, in-house designers, and outside firms. Bottles, colors, formulas, names, and the relationship between shampoo and conditioner were designed as one offer. Packaging World reported P&G's figures of a nine-percent increase in dollar share and a twenty-percent rise in purchase intent. The Clay Street participant account said a year-over-year sales index moved from 76 to 124.² Both reports describe a successful case; neither independently estimates what design thinking alone caused. They do show the organizational work hidden inside the phrase.

Roger Martin uses P&G as the book's major corporate transformation. The case gives The Design of Business a practical problem: established companies know how to repeat an answer, but the controls that make repetition dependable can also make a declining answer hard to question. Herbal Essences needed both a new hypothesis about value and the manufacturing, distribution, and brand systems capable of delivering it repeatedly.³

The knowledge funnel turns discovery into dependable work

Martin names the governing tension reliability versus validity. Reliability seeks consistency, prediction, and proof from known data. Validity asks whether an intervention produces an outcome people actually want. A reliable company can reproduce yesterday's answer after its meaning has changed. A valid insight that cannot be repeated, taught, or supplied at scale remains fragile. Martin's claim is not that one should defeat the other, but that an organization has to move between them.⁴

His knowledge funnel is the book's model of that movement. Inquiry begins with a mystery, a question whose important variables are unsettled. Work on the mystery produces a heuristic, a rule that narrows attention while still requiring judgment. Repeated use can turn a heuristic into an algorithm, a specified sequence that delivers a result without rediscovering it each time. The Herbal Essences team's interpretation of a generation was a heuristic; the formulas, package specifications, production steps, and launch routines made the offer repeatable.⁵

That reconstruction applies Martin's model to the documented episode. It is not a process trace supplied by an independent evaluator, and the funnel itself is an authored framework rather than an empirically established law of knowledge. The book illustrates it with corporate histories—including P&G, Research in Motion, Cirque du Soleil, and McDonald's—and makes a normative case for organizations to keep moving in both directions. Algorithms should free capacity for new mysteries; anomalies should be able to reopen an algorithm that has stopped producing value.⁶

The movement also redistributes authority. Turning expert judgment into a checklist or production routine can make quality widely available. It can also leave workers operating a system whose categories they cannot contest. The histories of knowledge, expertise, and professional autonomy and work design, productivity, and automation expose that second side of the funnel.⁷

“What would have to be true?” organizes a different kind of proof

Martin contrasts analytical reasoning from established evidence with intuitive leaps that may never submit to evidence. His design thinking combines the two through abductive reasoning: imagine a possible explanation or future, ask what would have to be true for it to work, then make and test. In an organization accustomed to defending complete proposals, the question turns disagreement into assumptions that prototypes and observations can examine.⁸

Abduction generates a hypothesis; it does not validate one. Nor does a protected creative space make budget, status, and career incentives disappear. Operations teams are often accountable for variance while innovation teams are rewarded for novelty. The Clay Street arrangement changed that local distribution of time and permission: a cross-functional group received an unusual interval in which to question the established brand, then had to reconnect its proposal to executives, technical specialists, suppliers, and routine operations.⁹

Later design scholarship disputes the idea that this progression is a universal canon. Lucy Kimbell argues that managerial design-thinking accounts can divide thinking from situated action, flatten diverse design practices, and privilege the designer as the main agent. Ulla Johansson-Sköldberg, Jill Woodilla, and Mehves Çetinkaya distinguish management's popular “design thinking” discourse from longer and more contested traditions of designerly thinking.¹⁰ A 2024 systematic review found 24 empirical studies of design thinking's effects on teams, using varied definitions, measures, and conditions. That developing evidence concerns team effects; it does not prove Martin's funnel or isolate the cause of P&G's sales result.¹¹

A later bottle changed the meaning of validity

The 2006 package was designed around visual and shelf cues for a “Gen Y” consumer: shape, color, and shelf presence carried much of its value. More than a decade later, Herbal Essences added raised stripes to shampoo bottles and raised circles to conditioner bottles so blind and low-vision users could distinguish them by touch.¹² An industry case history traces the work to Sam Latif's 2015 appointment to develop inclusive design at P&G and reports an expansion around World Sight Day 2019. P&G's current account describes the markings as part of a wider accessibility practice.¹³

The tactile package is a later comparison, not evidence available to Martin in 2009. It reveals what “validity” leaves open: valid for whom, in which situation, and according to whose evidence? The first redesign used observation of a chosen market segment to revive a brand. The second began from an experience that the segment and its visual design could miss. A method can invite empathy and still exclude anyone absent from the research coalition.¹⁴

The same boundary applies beyond the buyer. Reported share and purchase intent say nothing by themselves about manufacturing work, supplier bargaining, chemical exposure, water systems, or the material life of a discarded bottle. No lifecycle or labor comparison accompanies the celebrated case, so those effects remain unanswered rather than neutral. The institutional history of P&G brand management shows how the firm joined consumer research, product development, and portfolio authority at scale.¹⁵

The Design of Business is most useful as a language for a recurring organizational conflict: exploit what is known without closing the institution to what no longer fits. The Herbal Essences episodes show that the funnel is not an automatic march from mystery to truth. It is a succession of choices about whose experience becomes a clue, whose judgment becomes a rule, and who retains the power to open the question again.

Structured reading paths and evidence limits

The Sciences of the Artificial supplies a foundational account of design as changing existing situations into preferred ones; it is an intellectual predecessor, not evidence for Martin's funnel. Innovation, entrepreneurship, and renewal places the book's design practices beside other mechanisms for generating and testing new value. The existing paths to knowledge, expertise, and professional autonomy, work design, productivity, and automation, and P&G brand management expose the authority, labor, and institutional questions that a smooth knowledge-funnel diagram can hide.

Organizational intelligence connects the reliability–validity tension to a wider capacity for sensing, learning, and revision. Benefit for all life sets the moral boundary: commercial uptake and consumer preference do not establish benefit for workers, suppliers, communities, or ecosystems. No structured impacts, typed relations, or formal reading dependencies are asserted for this work. The seven related paths are editorial navigation and do not establish empirical support or direct intellectual influence. The evidence package combines Martin's own framework, publisher records, participant and trade accounts, company and industry accessibility histories, peer-reviewed conceptual critiques, and a systematic review of later team research. It contains no independent causal evaluation of the Herbal Essences restage and no labor, lifecycle, community, or ecological outcome comparison.¹⁵