What is the 3 Horizons Framework?

The 3 Horizons Framework is a growth and portfolio management model that sorts initiatives into three time-based buckets: Horizon 1 defends today's core business, Horizon 2 builds emerging revenue engines, and Horizon 3 seeds disruptive options for the long term. McKinsey popularized it through the 1999 book The Alchemy of Growth.

TL;DR

Three time horizons, one portfolio: H1 protects current profit, H2 incubates next-cycle revenue, H3 explores disruptive bets that may pay off in five to ten years.
The 70/20/10 default: A common starting allocation puts 70% of investment in H1, 20% in H2, and 10% in H3, then adjusts for industry maturity and disruption risk.
It is a planning lens, not a roadmap: Use it to balance a portfolio of bets; pair it with a delivery framework like OKRs to actually move H2 and H3 work forward.
The classic failure mode: H1 cash cows starve H2 and H3 of attention and capital, leaving the company without a successor business when the core declines.

Definition: The 3 Horizons Framework is a strategic planning tool that helps organizations envision and manage growth, by categorizing initiatives and innovations into three distinct time frames: the present, near future, and distant future.

Where the framework comes from

McKinsey consultants Mehrdad Baghai, Stephen Coley, and David White introduced the model in The Alchemy of Growth (1999) after a three-year research project at McKinsey's growth practice. The thesis is summarized in one sentence from the book: "The secret is to manage business opportunities across three time horizons at once: extending and defending core businesses, building new businesses, and seeding options for the future."

The secret is to manage business opportunities across three time horizons at once: extending and defending core businesses, building new businesses, and seeding options for the future.

― Mehrdad Baghai, Stephen Coley, David White, The Alchemy of Growth (1999)

McKinsey reissued the framework in its 2009 Enduring Ideas series, where it remains one of the firm's most-cited strategy artifacts.

The three horizons explained

Horizon 1: Defend and extend the core

Horizon 1 covers the businesses that generate most of today's profit and cash flow. Work in this horizon includes optimizing existing products and services, improving operational efficiency, defending market share, and squeezing remaining value from mature lines.

H1 funds the rest of the portfolio.

Horizon 2: Build emerging businesses

Horizon 2 covers rising business lines that already have customers or pilot revenue and are expected to become tomorrow's core. These initiatives typically include new products, new markets, or adjacent business models.

They require significant investment but show clear paths to scale. Amazon Web Services spent years as an H2 bet inside Amazon before it became H1.

Horizon 3: Create viable options

Horizon 3 covers nascent ideas, research projects, and venture-style bets that may or may not become businesses. These initiatives are evaluated as a portfolio of options, not as forecasted P&Ls.

Most will fail. The point of H3 is to ensure that something is ready when the current H1 declines.

Horizon 1 vs Horizon 2 vs Horizon 3 at a glance

Attribute	Horizon 1	Horizon 2	Horizon 3
Time to material impact	0 to 12 months	1 to 3 years	3 to 6+ years
Typical share of investment	~70%	~20%	~10%
Revenue today	Most of it	Small but growing	Negligible or zero
Primary metric	Profit, market share	Growth rate, pipeline	Learning, optionality
Management style	Operational excellence	Venture-style with stage gates	Research and experimentation
Risk of failure	Low	Medium	High
Example	Existing product line	New market entry, adjacent SKU	Moonshot R&D, new business model

The 70/20/10 split is a starting heuristic, not a rule. Mature, cash-rich incumbents in stable categories may justify a heavier H3 tilt; capital-constrained early-stage companies often run closer to 90/10/0 until the core is stable.

How to apply the framework

Inventory current initiatives. List every active project, product line, and bet. Assign each to H1, H2, or H3 based on time-to-impact and revenue maturity.
Check the balance. Compare your current spend split against the 70/20/10 default. Large gaps (for example, 95% in H1) signal a portfolio that will run out of road.
Define horizon-specific metrics. Hold H1 to profit and efficiency. Hold H2 to growth rate, pipeline conversion, and unit economics. Hold H3 to learning velocity and option value, not revenue.
Use OKRs to operationalize H2 and H3. Without explicit goals and owners, H2 and H3 work loses to whichever H1 fire is loudest. Linking each horizon to OKRs gives non-core work a defendable slice of attention.
Review quarterly. Initiatives migrate. An H2 bet that hits scale moves to H1. An H3 experiment that proves a market becomes H2. The portfolio is not static.

Real-world portfolios

Amazon. Retail and AWS now sit firmly in H1. H2 bets include advertising, healthcare (One Medical), and grocery. H3 includes Project Kuiper satellites and longer-horizon robotics work.

Alphabet. Search and YouTube are H1. Google Cloud, Waymo Commercial, and Pixel are H2. Verily (life sciences), X moonshots, and quantum computing are H3.

These examples illustrate the discipline, not the precise math. Both companies regularly shut down H3 bets when option value disappears.

Where 3 Horizons rollouts typically break

The framework's reputation has taken legitimate criticism. Steve Blank's 2019 critique argued that the model assumes H3 is far away, when in fast-moving industries H3 disruption can arrive on H2 timelines.

Three failure patterns recur:

The H1 gravity well. Quarterly earnings pressure pulls leadership attention and capital back to the core, starving H2 and H3 for years.
Confusing H2 with H1 extensions. Adjacent SKUs and feature releases get labeled "Horizon 2" but are really line extensions of H1, leaving the portfolio with no real new-business bets.
Treating H3 like an H1 P&L. Asking H3 experiments to project revenue and ROI on H1 timelines kills the option before it can be tested.

The fix is structural: separate teams, separate budgets, separate success metrics, and a strategic planning process that explicitly reviews each horizon on its own terms.

Using the 3 Horizons Framework with OKRs

The 3 Horizons Framework decides what to fund. OKRs decide who owns what and how progress is measured. The two pair cleanly:

H1 OKRs are short-cycle, mostly committed OKRs tied to efficiency and revenue retention.
H2 OKRs are mixed: some committed (pipeline, sign-ups) and some aspirational (market expansion).
H3 OKRs are almost entirely aspirational, with key results that measure learning, prototypes shipped, or hypotheses validated rather than revenue.

Who invented the 3 Horizons Framework?

McKinsey consultants Mehrdad Baghai, Stephen Coley, and David White introduced the framework in The Alchemy of Growth (1999). McKinsey re-popularized it in its 2009 Enduring Ideas series, and it remains one of the firm's most-cited strategy frameworks.

What is the 70/20/10 rule in the 3 Horizons Framework?

70/20/10 is a common default for splitting investment across the horizons: roughly 70% to Horizon 1 (core business), 20% to Horizon 2 (emerging businesses), and 10% to Horizon 3 (long-term options). It is a starting heuristic, not a rule; the right split depends on industry maturity, disruption risk, and balance-sheet capacity.

How does the 3 Horizons Framework differ from a typical roadmap?

A roadmap sequences specific deliverables on a timeline. The 3 Horizons Framework is a portfolio lens that classifies bets by time-to-impact and maturity, without prescribing dates. Most organizations use both: 3 Horizons to decide the mix of bets, and roadmaps or OKRs to execute within each horizon.

What are the main criticisms of the 3 Horizons Framework?

The most cited critique, from Steve Blank, is that the model implicitly assumes Horizon 3 disruption is years away, when fast-moving industries can produce H3 disruption on H2 timelines. Other critiques note that "horizon" is an imprecise metaphor and that the framework is a preliminary lens rather than a detailed planning tool.

Can small companies and startups use the 3 Horizons Framework?

Yes, but the allocation usually skews heavily toward H1 until the core is stable. Early-stage startups often run close to 90/10/0 or 80/15/5 rather than 70/20/10, because survival depends on the core business reaching escape velocity before H3 bets are affordable.

When should you not use the 3 Horizons Framework?

Skip it when your business is small enough to be effectively a single horizon, when the strategic question is operational rather than portfolio-level, or when you need a detailed delivery plan rather than a high-level lens. Pair it with a strategic planning process and execution framework rather than using it standalone.