STRATEGYWeeks to result

Try a Lot of Stuff and Keep What Works

Stimulate evolutionary progress through high levels of experimentation, opportunistic action, and purposeful selection of what proves successful

Problem it solves

unclear strategic direction

Best for

Organizations seeking to innovate and adapt in uncertain environments where the right path cannot be determined in advance through strategic analysis alone

Not ideal for

Situations where the strategic direction is already clear and what is needed is disciplined execution of a known plan, or where resources are too constrained for multiple experiments

Overview

Why this framework exists

Many of the best moves made by visionary companies came about not by detailed strategic planning but by experimentation, trial and error, opportunism, and quite literally accident. What looks in hindsight like a brilliant strategy was often the residual result of opportunistic experimentation and purposeful accidents.

This process closely resembles how organic species evolve through variation and natural selection. Collins and Porras call it branching and pruning: add enough branches to a tree (variation) and intelligently prune the deadwood (selection), and you will likely evolve into a collection of healthy branches well positioned to prosper in an ever-changing environment.

Evolutionary progress differs from BHAG progress in two key ways. First, it involves ambiguity rather than clear goals: by trying lots of different approaches, you are bound to stumble onto something that works, but you do not know ahead of time what it will be. Second, it usually begins with small incremental steps or mutations that eventually grow into major and often unanticipated strategic shifts.

Visionary companies do not leave evolution entirely to chance. They practice purposeful evolution: consciously creating environments that generate variation (through decentralization, individual initiative, experimentation budgets, and tolerance of failure) and then applying rigorous selection criteria (profitability, fit with core ideology) to keep what works and discard what does not.

The visionary companies harnessed the power of evolution to a greater degree than the comparison companies in fifteen out of eighteen cases studied. 3M became what one author called the mutation machine from Minnesota, creating organizational mechanisms specifically designed to generate a continuous stream of experiments and innovations.

Core principles

5 total
  1. What looks in hindsight like brilliant strategy is often the residual result of opportunistic experimentation and purposeful accidents.
  2. Branching and pruning: add enough branches (variation through experiments) and intelligently prune the deadwood (selection of what works), and you evolve into a healthy portfolio well positioned for change.
  3. Purposeful evolution means consciously creating environments that stimulate variation while applying rigorous selection criteria consistent with core ideology.
  4. Failure is an essential price to pay in creating a healthy branching tree. Companies must accept failed experiments as part of evolutionary progress.
  5. The evolutionary process and BHAG-driven progress are complementary: use BHAGs to define a mountain to climb, and use evolution to invent the way to the top.

Steps

5 steps
  1. Create an Environment That Generates Variation
    Build organizational structures and policies that actively encourage experimentation, individual initiative, and trying new things. Decentralize decision-making so that employees throughout the organization can pursue ideas without needing approval from the top. Allocate resources specifically for experiments. Celebrate initiative even when the specific result is unknown.
    Pro tip3M gave researchers fifteen percent of their time to pursue any project of their choosing, created an internal venture capital fund for employees with ideas, and decentralized to give business units the freedom to experiment. William McKnight's approach was captured in phrases repeated throughout the company's history: Listen to anyone with an original idea no matter how absurd it might sound at first. If you put fences around people you get sheep.
    WarningVariation without selection is chaos. Every mechanism for generating experiments must be paired with rigorous criteria for deciding what to keep and what to discard.
  2. Make Small Bets and Move Fast
    Begin experiments as small incremental steps rather than bet-the-company initiatives. Move quickly to test ideas in the real world rather than analyzing them to death. When Marriott noticed passengers at one restaurant buying food to take on planes, it did not convene a strategic planning committee. The very next day, J. Willard Marriott visited Eastern Air Transport and created a new business arrangement.
    Pro tipWal-Mart lived by the motto 'Do it. Fix it. Try it.' If you try something and it works, you keep it. If it does not work, you fix it or try something else. Speed of experimentation matters more than quality of initial analysis.
    WarningSmall bets does not mean low commitment. Once an experiment shows promise, commit resources aggressively to scale it. The initial experiment is small, but the follow-through should be vigorous.
  3. Apply Rigorous Selection Criteria
    Not all experiments should be kept. Apply clear selection criteria to determine which experiments to scale and which to prune away. The two primary criteria are: Does it work (is it profitable, effective, generating value)? And does it fit with our core ideology? Only experiments that pass both tests should be kept and invested in further.
    Pro tipJohnson and Johnson's approach was to try lots of new things, keep those that work, and quickly discard those that do not. It imposed rigorous selection criteria: only experiments that proved profitable and fit with J&J's core ideology remained in the portfolio. Despite numerous failures, J&J never posted a loss in 107 years.
    WarningBe willing to prune aggressively, including killing experiments that are moderately successful but do not fit with core ideology. Pruning is as important as branching.
  4. Embrace Failure as the Price of Progress
    Create a culture where failed experiments are treated as learning investments rather than career-ending mistakes. R. W. Johnson Jr. said 'Failure is our most important product.' Companies that punish failure will suppress the variation they need to evolve. Make it safe to try things that do not work out, as long as the experiments were conducted within the bounds of core ideology.
    Pro tip3M's near-fatal early days as a failed mining company made a permanent impression on William McKnight. He wanted enough internal variation so the company would never again have all its eggs in one basket. He created an organization designed to continuously self-mutate from within.
    WarningEmbracing failure does not mean tolerating incompetence or lack of effort. It means accepting that well-conceived experiments sometimes produce negative results, and that these negative results are valuable information.
  5. Scale What Works and Combine with BHAGs
    When an experiment proves successful and fits with core ideology, commit resources aggressively to scale it. Simultaneously, use BHAGs to provide overall direction and energy. The two approaches are complementary: BHAGs define the mountain to climb and evolution invents the way to the top. Wal-Mart simultaneously pursued bold goals and evolutionary progress throughout its history.
    Pro tipJack Welch at GE called this approach planful opportunism: set only a few clear, overarching goals, then on an ad hoc basis let people seize any opportunities they see to further those goals. This crystallized after reading about a Prussian general who argued that detailed plans usually fail because circumstances inevitably change.
    WarningDo not let the success of evolutionary progress make you complacent about setting BHAGs. And do not let the clarity of BHAGs make you dismiss the power of unplanned experimentation.

Checklist

Saved in your browser

Examples

3 cases
3M: The mutation machine from Minnesota

3M began as a failed mining venture. Under William McKnight's leadership, the company built organizational mechanisms specifically designed to generate continuous experimentation: researchers got fifteen percent of their time for self-directed projects, an internal venture capital fund supported employee ideas, each division was required to generate twenty-five percent of revenue from products introduced in the prior five years, and the culture celebrated phrases like 'Listen to anyone with an original idea' and 'If you put fences around people you get sheep.' McKnight's curiosity about a random letter from a Philadelphia ink maker led to the acquisition of waterproof sandpaper technology and, more importantly, the hiring of its inventor.

Outcome3M evolved from a failed mining company into one of the most innovative companies in history, producing products from Scotch tape to Post-it notes. Bill Hewlett of HP said if he had to bet his life on the continued success of any single company over the next fifty to one hundred years, he would bet on 3M, precisely because of its evolutionary capacity.
American Express evolving from freight to financial services

American Express began in 1850 as a regional freight express business. Through a series of unplanned steps, it evolved into something entirely different. In 1882, facing declining demand for cash-shipping, it created a money order that became an unexpected success. In 1892, the president's difficulty cashing letters of credit in Europe led to the invention of the traveler's check, which accidentally created a multi-billion-dollar float. An entrepreneurial employee in Paris defied the CEO's explicit order against entering the travel business by incrementally experimenting with ticket windows and travel services.

OutcomeThrough opportunistic experimentation, American Express transformed from a freight company into a financial and travel services company. None of this was planned. The company's evolution looked like brilliant strategic foresight in retrospect, but was actually the result of purposeful accidents, incremental experiments, and seizing unexpected opportunities.
Wal-Mart's people greeters

A store manager in Crowley, Louisiana, was having trouble with shoplifting. He tried an experiment: putting a friendly older gentleman by the front door to greet people on their way in and out. No one at Wal-Mart, including Sam Walton, had conceived of anything like this before. The greeter made honest people feel welcome while sending a message to potential shoplifters. The odd experiment proved effective and eventually became standard practice across the company.

OutcomeThe people greeter became a famous competitive advantage for Wal-Mart, but it was never designed at headquarters. It emerged from a local manager's experiment. Wal-Mart's culture of 'Do it. Fix it. Try it.' enabled such experiments to emerge, be tested, and spread organically when they worked.

Common mistakes

4 traps
Viewing the company's success as the result of brilliant foresight
It is tempting to look at well-adapted visionary companies and conclude they were the result of a master plan. But in many cases their strategic positions evolved through trial and error. American Express did not plan to become a financial services company. Wal-Mart's system came into being largely by evolutionary processes, not by strategic genius. Viewing success as planned rather than evolved prevents organizations from building the experimental capacity they need.
Killing experiments too quickly
Some experiments need time to prove their worth. Band-Aid products had a slow start before becoming J&J's biggest selling category. 3M's Wetodry sandpaper emerged from a seemingly irrelevant inquiry. Organizations that demand immediate returns from every experiment will prune promising branches before they have a chance to grow.
Running experiments without selection criteria
Variation without selection is chaos, not evolution. Organizations that try lots of things but never rigorously assess what to keep and what to discard will dissipate their resources. The pruning is as important as the branching.
Requiring approval from the top for every experiment
American Express's evolution into travel services happened despite the CEO explicitly declaring the company would not enter that business. An entrepreneurial employee in the Paris office moved incrementally and experimentally. Organizations that require top-down approval for every initiative suppress the variation they need.

Origin story

How this framework came to be

Collins and Porras were struck by how many key strategic moves at visionary companies were unplanned. Johnson and Johnson got into consumer products quite by accident when customers began requesting the talcum powder included as a packaging insert. Marriott stumbled into airport food services when passengers at a shop near Hoover Airport started buying meals to take on planes. American Express evolved from a freight company into financial and travel services through a series of opportunistic, unplanned steps, even against the explicit orders of its president who declared the company was not going into the travel business. These were not aberrations but a pattern found across most of the visionary companies studied.

Source

Traced to primary
Source · BOOK
Built to Last
Jim Collins & Jerry I. Porras · 1994
Open source →

Related frameworks

Browse all Strategy →