A finance team asks you to approve a two-year platform rebuild. The business case looks solid, if the market behaves the way the spreadsheet assumes. It might not. You have three obvious choices: approve the whole thing, kill it, or stall and look indecisive. There is a fourth, and it is usually the best one: fund a small, time-boxed slice that tells you whether the big bet is worth making, and keep the door to it open. That fourth move has a name. It's a real option.

The quick version

  • A real option is the right, but not the obligation, to take a future action (expand, delay, switch, abandon) once you've learned more.
  • Optionality is worth most when uncertainty is high and the cost of committing is hard to reverse. Waiting has value because information arrives.
  • The move: spend a little now to keep big choices open later. Stage the spend, set a decision date, and define in advance what would make you go, grow, or kill.
  • The catch: options aren't free, and hoarding them is its own failure. At some point you must exercise, or you've just paid to procrastinate.

The idea in depth: a bet you can walk away from

The term comes from finance. In 1977 the economist Stewart Myers, then at MIT, coined "real options" in a paper on corporate borrowing, arguing that a firm is not just a pile of assets but also a set of opportunities to invest on favourable terms later (Myers, "Determinants of Corporate Borrowing," Journal of Financial Economics, 1977). A financial call option gives you the right to buy a share at a fixed price within some window; you pay a small premium for the right, and you only exercise if the price moves your way. A real option is the same shape, applied to a real-world decision: a pilot, a prototype, a six-month lease, a minority stake. You spend a little to hold open a bigger move, and you commit fully only if the world cooperates.

Why does just waiting have value? Because most consequential investments share two features that standard net-present-value maths ignores. The economists Avinash Dixit and Robert Pindyck set this out in their 1994 book Investment Under Uncertainty (Princeton University Press): real decisions are usually irreversible (you can't fully recover the spend if you're wrong) and they can be delayed (it's rarely now-or-never). Put those together and a gap opens. Acting now means committing sunk money against an uncertain future. Waiting lets you resolve some of that uncertainty first, and information that arrives before you commit is information you can act on. Dixit and Pindyck call this the option value of waiting, and they show it can be large enough that a project with a positive NPV is still better left on the shelf for now.

The practical shift is to stop treating a big decision as a single yes/no at one moment, and to treat it instead as a sequence of smaller, conditional commitments. Don't ask "should we build the platform?" Ask "what is the cheapest thing we can do that would tell us whether to build the platform, and what would we need to see to greenlight it?"

flowchart TD
    A(["Uncertain bet"]) --> B("Spend a little now: pilot, prototype, option to expand")
    B --> C{"What did we learn
by the decision date?"}
    C -->|"Signal is strong"| D(["Exercise: commit fully"])
    C -->|"Signal is weak"| E(["Abandon: walk away cheaply"])
    C -->|"Still unclear,
info still arriving"| F(["Extend the option
(if it's genuinely cheap)"])
A real option turns one big yes/no into a staged, conditional commitment. Leaders Loop

Five shapes optionality takes at work

Real options aren't only financial. The strategy literature, notably Martha Amram and Nalin Kulatilaka in Real Options: Managing Strategic Investment in an Uncertain World (Harvard Business School Press, 1999), catalogues the recurring shapes. You'll recognise all of them from ordinary management life:

  • Option to defer, wait before committing (lease, don't buy; run the trial before signing the vendor).
  • Option to expand, make a small move that buys the right to scale up (a pilot in one region you can roll out).
  • Option to abandon, design the bet so you can stop and recover something (a contract with a clean exit clause).
  • Option to switch, keep the ability to change inputs, suppliers, or direction (dual-sourcing; a flexible team).
  • Staging / growth options, fund in tranches gated on milestones, the way venture capital does.

Name which option you're actually buying before you spend. "We're paying a 15% premium for a month-to-month contract" is a defensible sentence when you can finish it with "...because it preserves the option to switch while the regulation is still unsettled." The premium is the price of optionality, and you should be able to say what you're getting for it.

From valuation tool to leadership habit

The most useful translation of real options for leaders came from the strategist Rita Gunther McGrath. Most managers will never run an options-pricing model, and McGrath's contribution was to turn the maths into a way of reasoning. With Ian MacMillan she introduced discovery-driven planning (Harvard Business Review, 1995): for genuinely uncertain ventures, don't write a plan that assumes you know the numbers, write down your assumptions, then spend money to test the riskiest ones cheaply, at milestones where you can change course. (HBR's 2017 refresher by Amy Gallo is a clean summary if you want the short version.) This is real-options thinking in plain operational dress: each milestone is a chance to expand, persevere, or abandon, and you've capped your downside before you've earned the right to the upside.

McGrath pushed the idea further in a paper that reframes failure itself. In "Falling Forward: Real Options Reasoning and Entrepreneurial Failure" (Academy of Management Review, 1999, an award-winning paper), she argues that when you pursue high-variance bets but only invest more if conditions stay favourable, a string of small, cheap failures isn't waste. It's the price of buying information, and it's exactly how optionality is supposed to work. The discipline isn't avoiding failure; it's making each failure small, fast, and informative, and not throwing good money after a dying bet out of pride.

The skill isn't avoiding failure. It's making each failure small, cheap, and quick to read, then actually acting on what it tells you.

This connects to a broader analytical posture. Preserving optionality is a sensible default because we live in a world of stochastic, not deterministic, outcomes, and it pairs naturally with Bayesian updating: an option is precisely a structure that lets you change your mind cheaply as the evidence updates your priors.

Where it breaks down (the honest limit)

Optionality has a failure mode, and it's the opposite of the one you'd expect. The danger isn't usually committing too soon, it's never committing at all. Options cost money to hold. Run too many pilots, keep too many doors open, and you bleed cash and attention while competitors who picked a direction pull ahead. An option only has value if you're genuinely willing to exercise it; a portfolio of bets you'll never scale is just expensive indecision dressed up as strategy.

The practitioner literature is candid about a second limit: the formal maths is easy to misuse. In "A Real-World Way to Manage Real Options" (Harvard Business Review, 2004), Tom Copeland and Peter Tufano note that companies often borrow the Black-Scholes option-pricing formula wholesale, a model built for liquid, tradable financial options, and apply it to messy strategic bets it was never designed for, producing confident numbers built on sand. Their advice, and ours: use real options as a way of framing the decision, stage it, price the flexibility, set the kill criteria, and be very cautious about precise valuations. The framing is where almost all the value is.

flowchart LR
    A(["High uncertainty"]) --> B{"Can we delay,
and is committing
hard to reverse?"}
    B -->|"Yes to both"| C(["Optionality pays:
stage it, set a kill date"])
    B -->|"No, cheap to reverse
or info won't improve"| D(["Just decide:
options are dead weight here"])
    C --> E{"Decision date reached?"}
    E -->|"Yes"| F(["Exercise or abandon,
don't roll it again by default"])
Optionality is a tool, not a virtue. When delay buys nothing, holding the option just costs you. Leaders Loop

A worked example

Your team wants to enter a new overseas market. The full play, a local entity, a hired country manager, a localised product, is roughly a year of work and (illustrative figure) about £600k. The upside is real but hinges on demand you can't yet measure. The instinct is to model an NPV, get a positive number from optimistic assumptions, and commit.

A real-options framing changes the shape of the decision. Instead of one £600k yes, you stage it:

  • Tranche 1 (the option premium, ~£40k illustrative): a three-month landing test, a localised page, paid acquisition, a few partner conversations, with one explicit question: does demand convert at a cost that could ever pay back?
  • The decision rule, written in advance: if cost-per-qualified-lead lands below your threshold, you exercise the option to expand and fund Tranche 2. If it's wildly above, you exercise the option to abandon, and the £40k bought you a cheap "no" instead of a £600k mistake.
  • The discipline: you set the decision date now, while you're still clear-headed, so a fond attachment to the idea can't quietly turn the test into a permanent, unfunded science project.

You've spent ~7% of the full commitment to retire most of the uncertainty around the other 93%. If the test is a clear "no," that's not a failed project, it's the cheap failure doing exactly its job. This is also where the boundary with expected-value thinking matters: expected value tells you which bet is worth making on average; optionality tells you how to structure the bet so a bad draw costs you little.

Frequently asked questions

Isn't this just an excuse to procrastinate?

It can become one, which is the honest risk. The difference is a deadline and a rule. Real optionality means you've named the decision date and the criteria that will trigger "go" or "kill" before you start. Open-ended waiting with no exercise condition isn't an option, it's avoidance with better vocabulary.

How is this different from just running a pilot?

A pilot is one shape of a real option (the option to expand). The wider habit is to ask, for any uncertain commitment, "which future choice am I trying to keep open, what's it worth, and what will I pay to keep it?", and to apply that to leases, contracts, hires, and architecture decisions, not only product trials.

Do I need to run the financial maths?

Almost never. Copeland and Tufano (HBR, 2004) warn that the formal pricing models are easy to misapply to strategic bets. For nearly all leadership decisions, the reasoning, stage the spend, cap the downside, set kill criteria, captures the value. Treat any precise option valuation with suspicion.

When is preserving optionality the wrong call?

When delay buys you no new information, when the commitment is cheap to reverse anyway, or when holding the option costs more than the flexibility is worth. If waiting won't teach you anything, stop waiting and decide. Speed beats optionality when the cost of being wrong is low and recoverable.

How do I stop my team hoarding options?

Force exercise. Every open option gets an owner, a decision date, and a default of "kill unless we actively choose to continue." Reward clean, early kills as much as wins, that's the cultural signal that keeps optionality from curdling into permanent indecision.

Related in the Toolkit

Where to go next