Most big companies are very good at the thing they already do and quietly hostile to anything new, not out of malice, but because every process that makes execution reliable also makes invention awkward. Intrapreneurship is the deliberate fix: give a small number of people something close to a founder's freedom, and then govern all those bets as a portfolio instead of hoping one lucky project saves the year.

The quick version

  • Intrapreneurship is entrepreneurship inside an existing organisation, an employee taking ownership of a new idea with founder-like freedom and accountability. The word was coined by Gifford and Elizabeth Pinchot in a 1978 white paper.
  • An innovation portfolio spreads bets across core (improving what you sell now), adjacent (new-ish products or markets) and transformational (genuinely new) work, so you're never relying on a single throw of the dice.
  • A widely-cited benchmark from Nagji and Tuff (HBR, 2012) is roughly 70 / 20 / 10 of effort across core / adjacent / transformational, while the returns often run the other way, with the long-shot bets producing the outsized gains.
  • The move on Monday: protect new ventures from the core's metrics, fund them in small stages, and review the whole portfolio together rather than project by project.

The idea in depth

Start with the word. "Intrapreneur" was not coined by a consultant chasing a buzzword; it came from a 1978 white paper, Intra-Corporate Entrepreneurship, by Gifford Pinchot III and Elizabeth S. Pinchot. The economist Norman Macrae credited Pinchot with the term in The Economist in April 1982, and by 1992 it had a dictionary entry, The American Heritage Dictionary defined an intrapreneur as "a person within a large corporation who takes direct responsibility for turning an idea into a profitable finished product through assertive risk-taking and innovation." Pinchot's own definition is shorter and stickier: intrapreneurs are "dreamers who do."

The reason the idea endures is that it names a real structural problem. Steve Blank, who helped formalise the lean-startup method, puts it bluntly: in a startup, close to 100% of the company is focused on searching for a new business model; in a large corporation, the overwhelming majority is focused on executing the model it already has. The very things that make a company efficient, KPIs, approval gates, quarterly targets, repeatable process, are, in his phrase, the root cause of its inability to be inventive. Every new control you add to protect the core business makes the next new idea a little harder to get off the ground. The practical answer is to stop asking a new idea to survive inside the core's rules. Give it a separate operating rhythm, a separate definition of success, and a sponsor senior enough to shield it, structural protection, not just a pep talk.

The honest limitation: separation cuts both ways. The history that gets quoted most, Lockheed's Skunk Works, where Kelly Johnson's hand-picked team designed the XP-80 jet fighter in 143 days, largely outside the company's normal approval structure, is real and remarkable. But "give the team total autonomy" is also how you get an orphan: a venture so detached from the parent that it can never plug back into the firm's distribution, brand or balance sheet, which is usually the whole reason to build inside a big company rather than leave. Govindarajan and Trimble, in The Other Side of Innovation (2010), argue the harder skill isn't generating the idea, it's execution, and specifically how the new venture and the established business share people and power. Autonomy is a dial, not a switch.

Why a portfolio beats a hero project

Individual intrapreneurs make headlines; portfolios make the numbers work. Any single new venture is a coin-flip at best, so a leader who bets the year on one "moonshot" is gambling, not innovating. The portfolio view, backed by McKinsey's Three Horizons framework, set out by Baghai, Coley and White in The Alchemy of Growth (1999), says run all three time horizons at once: Horizon 1 defends and extends today's business, Horizon 2 builds the emerging ventures of the next few years, and Horizon 3 plants options for a future five to ten years out. You don't choose between them; you hold all three concurrently.

Bansi Nagji and Geoff Tuff sharpened this into something a leader can actually allocate against. In Managing Your Innovation Portfolio (Harvard Business Review, 2012) they map innovation on two axes, how new the product is, and how new the market is, into core, adjacent and transformational work, and report that better-performing firms tended to split innovation effort roughly 70% core, 20% adjacent, 10% transformational. The part leaders forget is the twist in the tail: the returns often ran in the opposite ratio, with the small slice of transformational bets generating a disproportionate share of the gains. So treat the 10% not as the budget line you cut first when the quarter is tight, but as the one with the highest expected payoff per dollar, and fund it in small, staged increments, so a failed bet costs you a probe rather than a fortune.

You invest 70/20/10. The returns can come back 10/20/70. The cheap long-shots are doing the heavy lifting.

The limitation worth stating plainly: 70/20/10 is a benchmark, not a law. Nagji and Tuff themselves say the right mix differs by industry, company and moment, a mature utility and an early-stage biotech should not carry the same split, and a firm under genuine disruption may need to skew far more aggressive. Treat the numbers as a starting hypothesis to argue with, not a target to hit for its own sake.

flowchart LR
  A(["Innovation effort"]) --> H1(["Core (~70%)
improve what we sell now"])
  A --> H2(["Adjacent (~20%)
new product or new market"])
  A --> H3(["Transformational (~10%)
genuinely new bets"])
  H1 --> R(["Returns can invert:
the 10% often pays the most"])
  H2 --> R
  H3 --> R
Effort is weighted toward the core; outsized returns often come from the small transformational slice. Ratios after Nagji & Tuff (HBR, 2012). Leaders Loop

What good governance actually looks like

The failure mode isn't a shortage of ideas; it's funding every promising one a little, starving all of them, then killing the survivors at the first bad quarter. The fix borrows from venture capital: stage-gate the money, not just the schedule. Fund a new venture in tranches tied to evidence, does the problem exist, will anyone pay, can we deliver it, and be willing to stop a bet cheaply and without shame. This is where the lean-startup discipline of build-measure-learn earns its keep: each tranche buys a test, not a promise.

Equally important is who keeps score, and how. If a Horizon 2 venture is judged on Horizon 1 metrics, this quarter's revenue, this quarter's margin, it will lose every budget fight to the established business, every time, because the established business is supposed to win on those numbers. So the move is to give each horizon its own success measure: validated learning and pipeline for the new bets, profit and efficiency for the core. And review them together, in one portfolio conversation, so trade-offs are made on purpose rather than by neglect.

flowchart TD
  I(["Idea / problem worth solving"]) --> G1{"Real problem?
(small probe)"}
  G1 -- no --> X(["Stop cheaply,
capture the lesson"])
  G1 -- yes --> G2{"Anyone will pay?
(small pilot)"}
  G2 -- no --> X
  G2 -- yes --> G3{"Can we deliver
& scale it?"}
  G3 -- no --> X
  G3 -- yes --> S(["Fund the next stage,
integrate with the core"])
Stage-gated funding: each tranche buys a test, and stopping early is a win, not a failure. Leaders Loop

A worked example

The figures here are illustrative, but the shape is true to how this plays out. Picture a mid-sized logistics firm, call it Meridian Freight, whose core business is regional trucking. Margins are fine but flat, and a board member keeps asking what happens when freight goes electric and software-routed.

Meridian's leadership sets a deliberate split. Roughly 70% of its modest innovation budget stays in the core: better load-matching software, fuel savings, depot efficiency, unglamorous, reliable, defending today's revenue. About 20% goes adjacent: offering its routing tools to other small carriers, a new-ish product for an existing market. And 10%, a single full-time intrapreneur, two engineers, and a small staged budget, goes transformational: a pilot for an electric last-mile delivery service in one city.

The transformational team is run differently on purpose. It does not report into regional operations, whose manager would (rightly) judge it on this quarter's loads moved and kill it by March. It reports to the COO, gets eight weeks and a small tranche to answer one question, will city retailers actually pay a premium for zero-emission last-mile?, and is told a clean "no" returned on time is a success, not a black mark. Meanwhile the adjacent bet becomes the year's surprise: other carriers pay for the routing tools, and within eighteen months that "20%" line is a real revenue stream, exactly the inversion Nagji and Tuff describe. None of it depended on a genius. It depended on three things: separating the new bet from the core's scoreboard, funding it in stages, and reviewing the whole set together.

Frequently asked questions

Isn't intrapreneurship just "let people work on side projects"?

No, and that's the common misread. A famous "20% time"-style policy gives people permission to tinker; intrapreneurship gives a person ownership, a sponsor, and accountability for turning an idea into something the business ships. Permission without protection and resources tends to produce what Steve Blank calls innovation theatre: lots of visible activity, very little that survives contact with the core's budget process.

How is this different from disruptive innovation or the innovator's dilemma?

Those describe the threat, why successful firms get unseated by cheaper, worse-at-first entrants. Intrapreneurship and innovation portfolios are part of the response: a structured way to fund and protect the new bets that might otherwise never get oxygen inside an organisation optimised for its current business. See the related links below.

We're small. Do we even need a portfolio?

The percentages matter less than the principle. Even a ten-person company can decide that most of its energy defends the core, some explores adjacencies, and a sliver is reserved for a genuinely new bet that won't pay off this year. The discipline is making that split on purpose, and not letting the urgent core quietly consume the 10% every single time.

What's the single biggest reason these efforts fail?

Judging new ventures by the core's metrics. A young bet measured on this quarter's profit will always lose to the mature business that's designed to deliver this quarter's profit. Give each horizon its own scoreboard, or the new work dies in the budget meeting.

Should the new venture be fully separate from the parent?

Usually not fully. Total separation buys autonomy but risks an orphan that can't tap the parent's customers, brand or scale, which was the point of building it inside. Govindarajan and Trimble frame this as a shared-staff, shared-power problem to design deliberately. Treat autonomy as a dial you set per venture, not a switch.

Related in the Toolkit

Where to go next