Antifragility in New Product Development

In the book Antifragile: Things that Gain from Disorder, Nassim Nicholas Taleb shared insights on optionality that can be expanded to include new product development environments.

This post provides an introduction to the non-linear gains associated with antifragile systems that may be realized by designing new product development environments that help individuals improve their capability to synthesize many new options continuously and enhance their proficiency to exercise options that are attractive. This post includes a comparison to concepts represented in Boyd’s OODA Loop sketch.

Fragile, Robust, Resilient, and Antifragile Development Environments

Taleb’s classification of systems as fragile, robust, resilient, and antifragile may be used to characterize development environments. Every development environment can be characterized in terms of its fragility, robustness, resilience, and antifragility.

Antifragile: Things that Gain From Disorder by Nassim Nicholas Taleb

Antifragile: Things that Gain From Disorder by Nassim Nicholas Taleb

A development environment that tends to be fragile does not welcome disorder. When uncertainty is injected, the results may be unpleasant.

In a fragile development environment, one obstacle can prevent the realization of value. Examples of harmful conditions include:

  • Incorrect, incomplete, or misleading information
  • A problematic handoff between functional groups
  • Disagreements among functional groups

Unpleasant results may include delays, cost overruns, and insufficient adoption of the product. Individuals tend to be frustrated. The more fragile the development environment, the less likely it is to thrive.

From project-to-project, a robust development environment tends to survive unchanged. Processes tend to be preserved. Individual contributors tend to retain their employment status.

From project-to-project, a resilient development environment survives changes from external factors. After a project is complete, there may be changes such as a re-arrangement of the organizational chart. New tools may be incorporated. The organization survives to serve the needs of the next project.

The word ‘antifragile’ is an adjective created by Taleb. It can be defined as the exact opposite of fragile.  According to Taleb, “Antifragile is beyond resilience or robustness.

An antifragile system thrives and grows when exposed to a moderate amount of volatility, randomness, disorder, and stressors. An antifragile system benefits from a moderate amount of adventure, risk, and uncertainty.


In Chapter 7, Taleb described the concept of iatrogenics as “damage from treatment in excess of the benefits.”

Iatrogenesis: preventable harm resulting from the treatment or advice of a healer.

The word iatrogenesis is not common in product development but harmful inputs may come from multiple sources. These include:

  • Specialists that assume that solutions to development problems relate to their area of expertise.
  • Innovation pundits, consultants, and vendors that offer their favorite tools and techniques as solutions
  • Interventionalists that believe that their contributions will improve outcomes
  • Status quo

It may be difficult to recognize the harmfulness associated with specific sources because of cognitive biases or unvalidated claims. Recognizing harmfulness is more difficult in development environments that isolate individuals of different functional specialties.

New product development efforts can suffer from iatrogenesis. Approaches to recognize potentially harmful inputs and reduce potential damage from harmful inputs include:

  • Requisite Variety
  • Disintermediation
  • Pair Development

Requisite Variety

The concept of requisite variety can be used to emphasize the importance of having a diversity of potential responses in a development environment.

Requisite Variety: For a system to be viable, only a variety in responses can force down the variety due to disturbances.

The Law of Requisite Variety was formulated by W. Ross Ashby

The Law of Requisite Variety was formulated by W. Ross Ashby

For a development environment to be successful, only a large repertoire of possible responses can address the variety presented by a complex set of development problems that emerge throughout projects.

In a new product development environment, requisite variety may be achieved by mobilizing a network of contributors with diverse specialties and multiple perspectives. To be successful, individuals may require additional training, access to individuals with unique expertise, and cooperation.

Without requisite variety, previously successful responses to familiar patterns may not be recognized as insufficient responses.

Without a variety of potential responses at the appropriate times, a development environment may be fragile.

If there is excessive variety, the agility of the development environment may be reduced. To ensure appropriate adaptability, the network determines that certain responses should be amplified. Other responses are attenuated.

Disintermediation removes layers between individual contributors and data. It removes barriers between decision makers. One way to facilitate disintermediation in new product development environments involves individual contributors experiencing the interactions of customers with prototypes (or other experiments related to the product being developed). Direct observations that promote full-fidelity interactions are preferable to mediation approaches such as presenting individuals with reports that summarize activities.

Pair Development is implemented by facilitating the interaction of individuals of different disciplines (such as a coder and a marketer). Pair development provides an opportunity for interaction through activities such as dialog and sketching. The result of pair development should be the synthesis of options, not a summary of previous activities. Typically, no slides sets are used during these interactions.

Pair development in new product development

The result of pair development is the synthesis of options that is informed by the analyses available from multiple perspectives.

The purpose of pair development is not cross-training. The purpose is to develop a self-correcting focus and direction informed by the analyses of multiple perspectives.

Reducing iatrogenesis is a pre-requisite to synthesizing more attractive options.

Typical Options

A typical option, such as a financial option, provides a buyer with the potential to take action by a specified date without an associated obligation to buy or sell. Typically, an individual decides to exercise an option based on their perception of value at a specific time.

According to Taleb, optionality is the property of asymmetric upside (preferably unlimited) with correspondingly limited downside (preferably small).

Optionality: a quality of state where choice or discretion is allowed.

In Chapter 12 Taleb stated that “An option is a substitute for knowledge” In Chapter 13, Taleb wrote “antifragility supersedes intelligence.

The value of a typical option depends on factors such as the negotiation skills of the individuals involved and the type of control individuals have over their decisions. Development options require additional proficiencies.


Taleb summarized the experience of Thales, an ancient philosopher. Thales acquired an option to use equipment that may be needed during next year’s harvest. His potential profits or losses were not be determined solely by the accuracy of a crop forecast. If there was an abundant harvest, he could exercise his option for the equipment and be rewarded financially. If the harvest was scarce, he could decline his option and not suffer a loss. The harvest was bountiful and Thales built a substantial fortune.

Development Options

To improve the potential for success in new product development environments, an option must be more than a negotiated agreement based on a speculation. Development options are the most valuable when associated with current capabilities or a capabilities that may be acquired within an appropriate amount of time for an appropriate cost within the project constraints.

The approach to development within a network of individual contributors includes the interplay of capabilities with analyses and synthesis.

Synthesis is a process of connection. Synthesis generates something new and different. A synthesis approach enables one to imagine how a several capabilities may work together to produce the desired result.

In a properly designed new product development environment, proficient individuals analyze situations and synthesize new options continuously.

The interplay of synthesizing options and exercising options to improve antifragility

The interplay of synthesizing options and exercising options improves antifragility

The interplay of synthesizing options and exercising options includes interaction with the environment (feedback) that enables a network of individuals to comprehend, shape, and adapt during development. This enables operation at faster tempos and rhythms and the compression of cohesive observation-orientation-decision-action time cycles.

The capability to synthesize many development options and exercise a few attractive development options rapidly and repetitively enables a properly prepared network of individual contributors to realize non-linear gains that are not possible by alternate approaches such as ones that focus on managing mandates.

Comparing Development Options to Concepts Represented in Boyd’s OODA Loop Sketch

Concepts related to development options have similarities with concepts represented in John Boyd’s OODA Loop sketch.

  • The concept of synthesizing development options is similar to the Observation and Orientation items represented in Boyd’s OODA Loop sketch.
  • The concept of exercising development options is similar to the Decision, Action, and Unfolding Interaction with Environment items represented in Boyd’s OODA Loop sketch.
  • The outcome from cycles of synthesizing a multitude of development options and exercising a multitude of development options is consistent with victory and the representation of a “series of maneuvers” in an OODA Loop context.
  • The non-linear gains associated with exercising attractive options in antifragile environments are similar to Boyd’s themes for vitality and growth.
Concepts related to development options have similarities with concepts represented in John Boyd's OODA Loop sketch.

Concepts related to development options have similarities with concepts represented in John Boyd’s OODA Loop sketch. Synthesizing development options is similar to the Observation and Orientation. Exercising development options is similar to the Decision, Action, and Unfolding Interaction with Environment items.

Enhancing Optionality throughout Development

For each development project, the goal is to change the environment to improve the capability to synthesize a multitude of development options and exercise a multitude of development options that are attractive. To increase this capability throughout development, invest to improve the following objectives:

  • Design the development environment to embrace optionality
  • Produce new repertoire based on theory and refined by practice with others seeking a high level of proficiency. This requires a sustained, deliberate effort.
  • Develop the capability for rapid cycles of observation, orientation, decision, and action which is based on the OODA Loop concept from John Boyd.
  • Improve the capability to shift rapidly between options which is similar to the concept of Fast Transients from John Boyd. This is similar to improving agility.
  • Improve the capability of individuals to synthesize options that are cohesive across the network and cohesive over the duration of the project(s).
  • Improve the proficiency to exercise attractive options. Identifying attractive options includes developing a holistic perspective and recognizing iatrogenesis.

Antifragility and Development Experience

Individual contributors invest much of their time in new product development projects. Their personal investment is what Taleb refers to as “skin in the game.” I refer to an individual contributor’s day-to-day and year-to-year set of perceptions and responses as Development Experience [DX].

There are multiple approaches to improve an individual contributor’s development experience by reducing fragility, increasing robustness, or increasing resilience of the development environment. An individual contributor’s development experience may improve dramatically by designing the development environment to improve antifragility. According to Taleb, “The option is an agent of antifragility.” Options make vitality and growth possible.

Development options are agents of development experience. Development options drive non-linear gains in antifragile development networks. Designing to improve development options stimulates better performance from individual contributors even when there is volatility. This inspires better performance from others. This creates virtuous circles, beneficial cycles of development efforts. This inspires greater commitments to project success.

Additional Information

1. Taleb includes Post-Traumatic Growth as a characteristic of antifragility in Table 1 in his book. In “Beyond Surviving New Product Development” I defined:

Post Development Growth: the positive changes experienced by individuals that result from enhanced new product development capabilities. Post Development Growth includes reflection to achieve cognitive clarity. It goes beyond reflection to action.

An antifragile development environment is more likely to produce Post Development Growth. This tends to enable better outcomes in future projects.

2. A fragile development environment is consistent with the model introduced in “The Devastating Zero Model of New Product Development.”

3. Too many inputs may be harmful because it may be difficult to discern the valuable from the harmful or signal from the noise. Too many inputs reduce a network’s agility. A requisite variety approach must include ways to evaluate potential contributions to project goals. One approach is the development of “continuously correcting, network-informed schwerpunkt” described in my Reimagining How New Product Development Artifacts Impact What We Should Be Doing Today post.

4. This post included extracts from my book “Developing Winners: Assimilating the Insights Encapsulated in Boyd’s OODA Loop

[Requires iTunes or Apple Quicktime. Duration 15 minutes]

Reimagining How New Product Development Artifacts Impact What We Should Be Doing Today

In this post, I will share ways to categorize new product development artifacts. I will clarify several memes. Then, I will offer a concept that individual contributors can use as they determine what they should be doing on a particular day.

Two Types of Artifacts in New Product Development

During new product development, many artifacts are produced. The word artifact is from the Latin phrase arte factum, skill + to make.

Typically, the product is a valuable project artifact. Other artifacts include items such as “design documents, data models, workflow diagrams, test matrices and plans, setup scripts, …” (from “What does artifact mean” on Stack Exchange).

In the context of new product development, deliverables are a subset of artifacts.

A product may be characterized as a set of external deliverables. Other items may be characterized as internal deliverables. Some of the internal deliverables may be maintained as documents. These are not the only artifacts.

In the context of new product development, deliverables are a subset of artifacts.

In the context of new product development, deliverables are a subset of artifacts.

Other Artifacts in New Product Development

In addition to external deliverables and internal deliverables, artifacts include:

  • Items used to produce deliverables such as tools
  • Secondary items produced during development. These may be unintended items.
  • Items that are not incorporated into the current project but may be incorporated into future development efforts. This includes training.
  • Incomplete, unfinished, or abandoned items
  • Intangibles such as development strategies, tactics, and culture

These items are a subset of artifacts.

Why it is Difficult to Determine What is Important Today

Individual contributors ask “What should I be doing now?” and “Why?” There may be questions such as ’Should a specific artifact be created?’ and ‘How much effort should be expended creating it?’ Resolving priorities may be difficult because of situations such as:

  • The number of items that could be explored during a project may be greater than the network’s (1) development capacity.
  • Predictions about the future (including how much effort will be required to develop a particular item such as a user story or feature) are estimates about an emerging set of conditions.
  • Some items are not on the list of considerations. Because of a lack of experience and insight, these items are not known.
  • Some priorities may be specified explicitly. Some issues seem to require immediate resolutions. Some priorities evolve.
  • Interruptions impact flow (2). This includes flow within functional groups and flow between functional groups and the system.
  • There are dependencies. For example, a coder’s efforts may precede a technical writer’s efforts during development.

Perspectives Influence the Perception of Value

The value of any artifact is subject to the perspectives of the stakeholders. Coders tend to value working code. Some individuals may stress the importance of prototypes they created. Copywriters tend to value persuasive messages. Other individuals favor spreadsheets.

Even though the word artifact has noble origins, it may have positive or negative connotations. Sometimes, there is an implication that certain types of artifacts have less value than a product delivered to the customer. For example, the Agile Manifesto includes the phrase “working software over comprehensive documentation.”

Perceptions about the value of artifacts and the attention that they should receive are driven by factors that include:

  • The status quo. A bias to repeat what was done previously. Value is attributed to what was delivered previously.
  • The loudest voice. The person that has the most authority. HiPPO, which is an acronym for “Highest Paid Person’s Opinion”
  • Curated information that may of may not be validated
  • Expectations from estimates and milestones.
  • Feedback from well-crafted experiments


When a multitude of individuals with diverse specialties develop artifacts, there may be a tendency to sub-optimize efforts.

Sub-optimization: A situation characterized by an individual (or a group of practitioners with a specific function) that tends to work within their specialty (sometimes referred to as silos) without regard for the impact on the output of the entire development effort (also known as the system).

Indicators of sub-optimization may include:

  • Efforts expended to improve a component do not improve the system performance
  • The success-limiting component does not receive the appropriate resources
  • Rewards are silos-centric instead of the system-centric

For the individuals that comprise the development network, producing artifacts quickly is not the main objective of new product development. That would be sub-optimizing based on a proxy for value. However, the capability to produce artifacts quickly contributes to achieving the main objectives.

Besides making decisions about which artifacts to develop, choices are made about developing the capabilities to produce artifacts. These choices depend on the proficiencies of the individuals that are mobilized for the development network and additional training that they receive.

Preparing to Make Better Choices

Suggestions that may help individual contributors make better hour-by-hour and day-to-day choices include:

  • Better choices are enabled when individuals improve their proficiency. According to John Boyd,  “It is advantageous to possess a variety of responses that can be applied rapidly to gain sustenance, avoid danger, and diminish adversary’s capacity for independent action.” (Boyd, Patterns of Conflict, #12)
  • Some specialists tend not to believe that solutions exist outside of their area of expertise. Although a cross-functional team strategy may surface a few perspectives, employing the concept of Requisite Variety provides multiple perspectives to inform better choices about the system.
  • Better choices require considering the immediate needs and the entire timeline of the current project and the impact of the timelines related to future products.
  • The interplay of functional specialties accelerates the development of implicit coordination. It improves the resiliency of the network.

Suggestions that may help development networks (which are also be known as the system) improve their capabilities for cooperation, collaboration, and harmony include:

  • Invest in the development of Einheit, the short-term alignment of individual efforts.
  • Invest to improve schwerpunkt, a concept that provides focus and direction for the long term. It provides actionable guidance in situations where there are no explicit directions. It reinforces mutual trust. According to John Boyd, Schwerpunkt “acts as a center or axis or harmonizing agent.“ (Boyd, Patterns of Conflict #78) It contributes to a focus on the results.
  • Invest in the development of continuously self-correcting, network-informed schwerpunkt. This is the capability to adapt the concept that provides focus and direction over time so that the network can detect and correct mismatches from factors such as accumulated learning, evolving market conditions, and new boundary conditions.
NPD networks and the product

A continuously self-correcting NPD network has the capability to detect and correct mismatches from factors such as accumulated learning, evolving market conditions, and new boundary conditions.

The phrase “working software over comprehensive documentation” from the Agile Manifesto should not be over-simplified to “eliminate documentation.” The improvement kata, a systematic, scientific routine of thinking and acting from the Toyota Production System should not be over-simplified to “eliminate artifacts by reducing the seven wastes (muda).”

Functional Specialists Embracing the System Perspective

To be more successful in a development network, find better ways to produce and integrate artifacts that contribute to the goals of development.

To make better hour-by-hour and day-by-day decisions, embrace a system perspective when writing more lines of code or producing more pages of documentation. Move beyond a perspective that is limited by reductionism or procrustean solutions. Encourage deliberative subtraction (deciding what not to develop) from the perspective of the system. Embrace new product development as more than a collection of diverse artifacts.

Embrace new product development as more than a job characterized by obvious answers where choices are framed as ‘OR’ selections. Embrace “AND” selections that meet the needs of the present and anticipate development in the future. Facilitate set-based design over point-based design. Strive to be proficient problem solvers that also invest in future capabilities. Incorporate artifacts that contribute to the goals and minimize distractions.

First rate individual contributors have the ability to hold a requisite variety of ideas about artifacts in their mind at the same time and still retain the ability to function. Refactoring F Scott Fitzgerald from The Crack-Up, April 1936 (3)

With that capability, you can improve your Development Experience [DX] which is another artifact that should be made skillfully.


This post included extracts from my book “Developing Winners: Assimilating the Insights Encapsulated in Boyd’s OODA Loop
1. The definition of New Product Development Network from the Glossary for New Product Development

A New Product Development (NPD) Network is a temporary, dynamic, adaptive system designed to evolve a product vision and compare that to the reality of their current version. In an NPD network, individuals may report to multiple managers in multiple companies and have multiple priorities. Many individuals engage and disengage during the project.

Often, the word “teams” implies that most of the individual contributors are employed by one organization and assigned to a relatively small number of functional areas (such as engineering and marketing). Often, an individual is assigned to a project for most of the duration of the project. In contrast, NPD work groups are assigned specific tasks.

2. Mihály Csíkszentmihályi proposed this concept of flow.

3. The original quote was “The test of a first rate individual contributor is the ability to hold two opposed ideas in the mind at the same time, and still retain the ability to function.” by F Scott Fitzgerald

[Requires iTunes or Apple Quicktime. Duration 12:18]