Understanding innovation · Part IV · Organisation, impact, measurement
13 · Organising people for innovation
12 min de lectura
In brief
- Innovation is a collective cognitive activity, and research on teams indicates that the decisive variables are not demographic but functional: cognitive diversity that is actually used, contained size, and above all psychological safety, the condition in which reporting errors, doubts and immature ideas costs the person nothing.
- Traditional high-pressure incentives on measurable results, effective for executive work, damage exploratory work: the evidence shows that exploration requires tolerance for early failure and reward for long-term success.
- People whose ideas are not recognised leave, and they take the ideas with them: the history of spinouts shows that failing to recognise a contribution does not retain the value, it transfers it to the competitors that those people will found.
- The culture that sustains innovation is made of concrete conditions: autonomy over the how, a margin of unsaturated resources (slack), and the distinction, practised and not merely declared, between the experimental error (information bought) and the executive error (quality not delivered).
Section 1
Teams: what really counts
The rhetoric of the "diverse team" is so widespread that it is worth starting from what research actually supports. The diversity that helps innovation is cognitive and disciplinary: different perspectives, repertoires of solutions and domains of knowledge, which widen the space of recombination described in chapter 03. This diversity is not free, however: it increases coordination costs and the potential for conflict, and it produces its benefits only if the team has the conditions to use it, that is, if the divergent perspectives are actually expressed and integrated instead of being repressed by group dynamics. Unexpressed diversity is only friction.
The condition that decides whether diversity is used has a precise name in the literature: psychological safety, the construct formalised by Amy Edmondson (1999) in her study of hospital teams. The definition: the shared belief that the team is a safe place for interpersonal risk-taking, where reporting an error, asking a naive question, disagreeing with a senior colleague or proposing an unripe idea does not expose you to humiliation or retaliation. Edmondson's original finding is instructive: the best nursing teams reported more errors than the worst ones, not because they made more of them, but because they reported them; psychological safety does not lower standards, it raises the visibility of uncomfortable information. For innovation the relevance is direct: the raw material of the process (refuted hypotheses, observed anomalies, doubts about the dominant design, ideas that cannot yet be defended) is exactly the kind of information that unsafe contexts suppress. The best known large-scale confirmation comes from Google's internal study of its own teams (Project Aristotle, whose results were made public by the company): among the variables examined to explain the differences in effectiveness between hundreds of teams, psychological safety turned out to be the main predictor, above composition, seniority and the individual talent of the members.
On size, the organisational evidence converges on the small: teams grow in coordination costs more than in capacity, individual responsibility is diluted, and communication degenerates into ceremony. The practice of innovative organisations (the "two-pizza teams" made famous by Amazon are the folkloric formulation of a serious principle) keeps working units small enough for people to know one another, to decide quickly and for each to feel the weight of the outcome, coordinating between teams through clear interfaces rather than through plenary meetings.
Section 2
Incentives: why pressure misses the target
The standard managerial intuition (measurable objectives, reward tied closely to the result, pressure on the short term) works for executive work and turns against exploratory work. The reference theoretical result is Gustavo Manso's ("Motivating Innovation", Journal of Finance, 2011): if you want an agent to explore new routes instead of exploiting known ones, the optimal incentive scheme combines tolerance for early failure and reward for long-term success; the opposite scheme, which punishes the first setbacks and rewards immediate results, rationally pushes the agent towards exploitation, that is, towards the known route that gives measurable results straight away. Punishing early failure does not make people more prudent: it makes them conservative, which in exploratory work is the most costly form of failure.
The cleanest empirical evidence comes from the comparison between research funding regimes carried out by Azoulay, Graff Zivin and Manso (2011): scientists funded by the Howard Hughes Medical Institute (long horizons, renewal that is tolerant of early setbacks, assessment of the person rather than of the single project) compared with peers of equal quality funded by standard NIH grants (short cycles, defined deliverables, little tolerance). The researchers in the tolerant regime produced significantly more high-impact work, explored newer themes, and also produced more failures: exactly the profile that the theory predicts, more variance and richer tails, which is what exploration has to deliver (chapter 10: the value lives in the tails).
The third element, complementary to economic incentives, is recognition of the contribution: the literature on the motivation of creative work (starting from the work of Teresa Amabile) documents that intrinsic motivation (interest in the problem, the sense of progress, recognised authorship of the result) is the main engine of creative performance, and that it is eroded both by pressing control and by the unrecognised appropriation of the contribution. Hence the principle that connects this chapter to those on capital: in an activity in which value is created by identifiable individual contributions (ideas, technologies, competencies, relationships: the non-monetary capital of chapter 09), formal recognition of authorship and of the share, with declared criteria and a third-party body that applies them, is not only fairness: it is the structural incentive that retains people and their contributions. The following section shows what happens when it is missing.
Section 3
Case study: the Fairchildren, or where unrecognised value goes
In 1957 eight young researchers leave the Shockley Semiconductor Laboratory together. William Shockley, a Nobel laureate for the transistor, had gathered the best talent in the sector and was dispersing it with a paranoid and humiliating style of leadership (the descriptive opposite of the psychological safety of section 1) and with technical choices imposed from above. The eight ("the traitorous eight" in the vernacular of the time) found Fairchild Semiconductor, which within a few years develops the planar process and the commercial integrated circuit, becoming the progenitor of the industry.
The second half of the story is the part that concerns this chapter. Fairchild was controlled by an east coast parent company that drained the profits of the semiconductor division and granted its creators neither a significant shareholding nor autonomy. The result is documented by the industrial genealogy of Silicon Valley: out of Fairchild came, in successive waves, the founders of dozens of companies (the "Fairchildren"), among them Noyce and Moore, who founded Intel in 1968, and Sanders, who founded AMD. Steven Klepper's research on spinout dynasties has generalised the mechanism across several sectors and periods: firms generate spinouts above all when employees with valuable knowledge and ideas meet an organisation that does not recognise them (in decisions, in shareholdings, in authorship), and the spinouts thus generated tend to become the best competitors, because they take with them the parent company's most valuable knowledge. The lesson, in one line: the value created by unrecognised people does not stay in the company waiting for better times; it leaves with the people, and comes back to the market as competition. Recognition of contribution is therefore not a personnel policy: it is a value retention strategy, and its absence is the most costly of implicit policies.
Section 4
Culture as a set of concrete conditions
"Culture of innovation" risks being an empty formula; it is worth reducing it to the operational conditions that research and practice indicate as active.
Autonomy over the how. Exploratory work requires objectives to be defined and methods to be left to those who explore: micro-control over the how destroys exactly the variation on which exploration depends, as well as the intrinsic motivation of section 2. The mature form is the pairing of clear direction and real delegation: where we are going is decided and communicated, how to get there belongs to the team.
Slack. An organisation saturated one hundred per cent with operational work does not explore: exploration lives off the margin of time, attention and resources not allocated to the urgent. Slack is not inefficiency but the capacity to respond and to recombine; its total elimination is a portfolio choice (all horizon 1, in the terms of chapter 08) that presents itself as rigour.
The right to make mistakes, with the distinction that makes it serious. No healthy organisation celebrates error as such; the innovative one distinguishes two species. The experimental error is the well constructed hypothesis that the evidence refutes: it is information bought at the foreseen price, the normal product of exploration (chapter 07), and punishing it means teaching people not to explore or, worse, to hide the outcomes. The executive error is negligence about what you knew how to do: standards missed, controls skipped, and treating it indulgently in the name of a culture of error corrupts quality. The distinction has to be practised in facts (who is promoted, who is defended after a setback, which post-mortems look for causes and which for culprits), because it is from facts, not from declared values, that people deduce what is really safe to do. It is the point at which this chapter joins up with the psychological safety from which it started: culture is what happens to the first person who brings bad news.
In the Volcano method
This chapter gives the foundation of the core area that the model calls talent and organisation, with the declared formula "a company is people before it is processes". The recognition of contribution of section 2 is institutionalised in the central mechanism of value sharing: the shareholding in each startup is distributed among all the contributors (entrepreneurs, technical staff, investors, Volcano itself) in proportion to the value contributed, recognised by a committee with transparent metrics, with the declared principle that "those who create value own it; Volcano's role is to bring it out and remunerate it, not to retain it". It is the structural answer to the Fairchild mechanism of section 3: where authorship and share are recognised by protocol, value does not need to leave in order to be remunerated, and the competencies door (a paid role plus a shareholding in the startups) applies Manso's scheme, income in the present plus reward tied to long-term success. The network of independent companies under a common manifesto, with Volcano as coordinator and guarantor that "neither commands nor owns the startups", is the alternative to hierarchy that section 4 describes: autonomy over the how for the venture teams, common direction and standards from the method. And the distinction between experimental and executive error is built into the path itself: the stops decided at the seven decisions with a public protocol are honoured outcomes of the process, communicated transparently and with the rights of the participants fixed in writing from day one.
Readings
Further reading
- A.C. Edmondson, "Psychological Safety and Learning Behavior in Work Teams" (Administrative Science Quarterly, 1999): the founding study; for the extended and operational treatment, The Fearless Organization (2018) by the same author.
- G. Manso, "Motivating Innovation" (Journal of Finance, 2011): the theory of incentives for exploration; to be read together with P. Azoulay, J. Graff Zivin and G. Manso, "Incentives and Creativity: Evidence from the Academic Life Sciences" (RAND Journal of Economics, 2011), which is its empirical verification.
- T.M. Amabile, "How to Kill Creativity" (Harvard Business Review, 1998): the accessible synthesis of decades of research on the motivation of creative work.
- S. Klepper, "Employee Startups in High-Tech Industries" (Industrial and Corporate Change, 2001): the spinout mechanism and the Fairchild genealogy generalised.
Frequently asked questions
Frequently asked questions
Does psychological safety not risk becoming comfort and an absence of standards?
No, if you understand what it measures: safety in taking interpersonal risks (reporting, disagreeing, proposing), not an absence of demands. Edmondson's data arise precisely from this: safe teams report more errors because they see them and say so, not because they make more of them. The strong configuration combines high standards and high safety: a great deal is demanded of the work and whoever brings the uncomfortable information is not punished. The degradation into comfort happens when safety is confused with giving up frankness, which is instead its product.
Does rewarding those who fail not create an incentive to fail?
Manso's scheme does not reward failure: it tolerates early failure and rewards long-term success. They are two joint levers, and the second prevents abuse of the first: those who explore know that early setbacks do not condemn them and that the final value will be recognised as theirs, so they have an interest in exploring seriously, not in failing. What the scheme discourages is the simulation of short-term results and defensive conservatism, which are the typical products of the opposite scheme.
Why recognise shares and authorship instead of simply paying well?
Because current pay remunerates time, not the long-term contribution, and exploratory work produces value that matures over years: without a participation in that value, the person who created it has a rational incentive to take it where it is recognised, and that is exactly what the history of spinouts documents. Formal recognition (declared criteria, a third-party body, written shares) aligns the horizons: those who create stay because staying is the best way of harvesting what they have sown.
Does a bit of internal competitive pressure not help performance?
It helps executive performance and damages exploratory performance, and that is the distinction that runs through the whole chapter. Internal competition between ideas (alternatives compared, explicit criteria: chapter 03) is healthy; internal competition between people (rankings, punishment of the last, resources contested as a zero-sum game) produces concealment of information, conservatism and hoarding, that is, the exact opposite of the conditions that innovation requires. The right pressure comes from the problem and from the standards, not from colleagues.
John F. Kennedy, 1962