Home·Insights·Strategic alternatives: building a real choice

Strategic alternatives: building a real choice

Most companies «choose» between one real option and two strawmen built to lose. The quality of the decision was decided before anyone voted.

Growth Strategy2026-10-026 min readAtlas Strategy Group

A strategic decision presented as a choice is usually a ceremony. The outcome was known before the meeting; the alternatives existed to decorate it — a timid variation, an absurd extreme, and the option the sponsor wanted, sitting between them looking reasonable. The ceremony wastes more than the meeting: a company that never built a real alternative set has never tested what it believes, and its strategies inherit the unexamined assumptions of whichever option happened to be drafted first.

What makes a set of alternatives real

Three properties separate a genuine choice from a comparison. Each alternative must be someone's best case. Not a caricature, not a version of another option with the risk numbers inflated — the strongest honest version a competent advocate would defend, because alternatives built to lose prove nothing except that their author wanted them to lose. They must differ in the bet, not the decoration. Two plans with different slide decks but the same assumptions about the market, the customer and the company's advantage are one plan wearing two costumes; real alternatives disagree about what is true — where demand sits, what the company is good at, which risk dominates. At least one must be genuinely capable of winning. A set in which the sponsor's option is the only survivor of plausibility is a set that was fixed, and everyone in the room will behave accordingly — checking their calendars instead of the logic.

The standard failure: alternatives as theater

The theater version follows a known script. The decision is already made — by momentum, by a founder's conviction, by last year's plan. Someone is tasked with «showing alternatives». The task is performed honestly at first and then practically: the real options are hard to model, the timeline is short, and the path of least resistance is to construct comparisons the sponsor's option can beat. The presentation happens; the vote is unanimous; the minutes record a decision «after considering alternatives». The company has paid for a test of its strategy and received a test of its politics. The pattern compounds: teams that watch one such meeting stop bringing real alternatives to the next one, and the organization quietly loses the ability to generate them at all.

Building the real thing

The working discipline has three parts. Generate before evaluating: the options are built separately from the judgment, ideally by different people, so that drafting a case does not yet mean defending it. Write the strongest version of each: each alternative gets its advocate — a person who argues for it sincerely in the room, because an alternative no one argues for is not an alternative, it is set dressing. Pre-register the differences: before choosing, the team writes down what the alternatives actually disagree about — the assumptions, the risks, the evidence that would confirm one and kill another. This last step converts the choice into a test the company can learn from regardless of outcome: deciding under uncertainty works precisely because the decision's structure was made explicit while it was still cheap to change. The same discipline feeds scenario work: alternatives built on different beliefs about the world, checked against scenarios instead of a single forecast.

A decision made without real alternatives is not a decision. It is a confirmation of a draft, performed in front of witnesses.

Building genuine alternative sets — generation before evaluation, an advocate per option, pre-registered differences — is standing work in the growth strategy practice, and it is the single cheapest structural protection a company can buy against the most expensive failure in strategy: confidently choosing the only option it ever seriously considered.

All insights →