Three Horizons Model

Origin

The model was developed by Mehrdad Baghai, Stephen Coley and David White (The Alchemy of Growth, Perseus, 1999), based on McKinsey research into companies with sustained growth. The finding: companies that grow over decades defend their core business while simultaneously building new businesses and future options — they treat the horizons as a continuously filled pipeline, not as phases to be entered in sequence.

Typical use

In portfolio reviews, strategy retreats and growth planning, the model provides a shared language for a structurally uncomfortable fact: the core business finances the present but does not earn a future. It helps make resource conflicts between today's results and tomorrow's growth explicit, and it supplies the frame for giving each horizon its own goals, metrics and owners — instead of measuring everything by the yardstick of the current financial year.

Procedure

1. Map the portfolio

   Assign every business, project and initiative to a horizon — by maturity and earnings logic, not by revenue size. Honesty matters: a cost-cutting programme in the core is horizon 1, even if it trades under the label “transformation”.

2. Check the balance

   The map exposes imbalances: typically horizon 2 is thinly populated and horizon 3 is empty or filled with fig-leaf projects. Assess whether the pipeline is sufficient to replace the foreseeable erosion of the core business.

3. Define horizon-appropriate governance

   Each horizon gets its own yardsticks: horizon 1 is measured by margin and market share, horizon 2 by revenue growth and evidence of scaling, horizon 3 by validated learning and option value. Leadership styles and talent profiles differ by horizon as well.

4. Organise migration

   Regular reviews decide on promotion and termination: when does a horizon-3 option become a horizon-2 business, when does horizon 2 become a new core — and which initiatives are ended decisively so the pipeline does not clog?

Limits and typical mistakes

The best-known misapplication is treating horizon 3 as a residual: it receives whatever is left once horizon 1 has been served — and in a downturn it is cut first, which defeats the model's purpose. Equally widespread is the sequential misreading that the horizons are phases a company passes through one after another; in fact the authors demand simultaneous management. Further mistakes: assigning by revenue size instead of maturity, applying uniform metrics across all horizons — which makes horizon-3 initiatives fail against horizon-1 return requirements — and the notion that fixed budget quotas per horizon can replace strategic judgement. Finally, the time spans are industry-dependent; in fast-moving industries the horizons converge so closely that the classification loses its discriminating power.

Relation to the Innovator's Dilemma