Three Horizons Model

Originated by Mehrdad Baghai, Stephen Coley, and David White (McKinsey) in 1999

A framework for managing innovation across three simultaneous time horizons (protecting the core, building the next, and seeding the future) without letting any one horizon cannibalize the others.

Here’s a pattern that shows up repeatedly in corporate decline: a company spends years executing brilliantly on its core business, harvesting healthy margins, while gradually defunding anything that doesn’t show a near-term return. Then a disruptor enters, or the category shifts, and the company discovers it has no second horizon to fall back on and nothing seeded in the third. By the time leadership recognizes the problem, it’s too late to build the next thing from scratch.

The Three Horizons Model was designed to prevent exactly that. It gives companies a language and a framework for managing growth across multiple time frames simultaneously: not by choosing between the present and the future, but by funding both at once.

What the Framework Actually Does

Developed by Mehrdad Baghai, Stephen Coley, and David White at McKinsey and published in their 1999 book The Alchemy of Growth, the Three Horizons Model maps a company’s activities across three overlapping time horizons based on two variables: when they will generate meaningful revenue, and what level of certainty is attached to that outcome.

Horizon 1 is the core: current products and markets generating most of today’s revenue and profit. H1 deserves strong management attention, continuous improvement, and protection. But it’s also the most vulnerable to assuming it will remain healthy indefinitely.

Horizon 2 is the bridge: emerging businesses that are beginning to demonstrate commercial viability but haven’t yet reached scale. These might be new product lines, adjacent markets entered in the last few years, or acquisition integrations. They need investment and patience. They’re real enough to measure but fragile enough to kill with quarterly pressure.

Horizon 3 is the seed bed: early-stage experiments, R&D projects, technology bets, or new business model tests that have no meaningful revenue today but represent options on the future. They’re cheap (relative to H1 and H2) and mostly uncertain, but some of them will eventually become tomorrow’s H2 initiatives.

The model’s central argument is that all three horizons must be actively managed simultaneously. Companies that focus only on H1 will be disrupted. Companies that over-rotate to H3 will run out of money before the future arrives. H2 is where many promising initiatives die from underinvestment or organizational impatience.

The Origin

McKinsey published the framework at the height of the late-1990s technology boom, a moment when many established companies were discovering that their core businesses were more vulnerable to digital disruption than they had assumed. The authors had studied how long-term, sustainably growing companies managed innovation across time and found that the most resilient ones were consistently running initiatives across all three horizons, not just the current quarter.

The model drew on portfolio theory, innovation management research, and the observation that Horizon 3 ideas typically take seven to twelve years to become Horizon 1 businesses. That long gestation time meant companies needed to plant seeds years before they could know which ones would grow. This implied a tolerance for uncertainty and a willingness to protect investment in distant futures, which organizations under quarterly earnings pressure systematically resist.

How to Apply It

The first step is an honest audit. Map everything your organization is currently working on across all three horizons. Most companies discover they are massively overweighted in H1 and have very little in H3. This concentration isn’t inherently wrong (H1 is where the money comes from), but it does reveal a vulnerability to category-level disruption.

For Horizon 1 management, the key question is: what are we doing to keep this business efficient, competitive, and customer-relevant in the near term? This includes pricing optimization, customer retention programs, operational improvement, and incremental product updates. Spotify Wrapped is an H1 initiative: it serves the existing product, retains existing subscribers, and deepens engagement without requiring a new business model. It’s low-risk, well-understood, and highly efficient per marketing dollar spent.

For Horizon 2 management, the question is: which emerging bets have shown enough commercial signal to deserve scaling investment? These initiatives often fail because organizations treat them like H1 businesses (demanding immediate profitability) or like H3 experiments (giving them too little structure and resourcing). H2 needs a different management approach: clear milestones, protected budgets, and leadership sponsorship that shields them from quarterly pressure.

Atlassian’s Playbook initiative started as an internal H2 bet on community-led product adoption before becoming a significant part of how Atlassian drives enterprise growth. It needed sustained investment and organizational commitment before it demonstrated the value that made it central to the H1 business.

For Horizon 3 management, the goal is optionality. Seed many bets cheaply. Maintain a portfolio of small investments in technologies, business models, or market experiments that might become relevant in five to ten years. Fail fast on the ones that don’t develop. Double down on the ones that start showing unexpected commercial signal.

The hardest organizational work is protecting H3 from H1’s budget pressure. In a difficult quarter, H3 experiments are the easiest line item to cut. Building explicit budget structures and organizational ownership for H3, separate from H1 P&L accountability, is the most practical way to defend long-term optionality.

A Real Example

Apple’s return under Steve Jobs in 1997 is a compressed version of three-horizon thinking under existential pressure. Apple at the time was an H1 crisis: the core Mac business was bleeding market share and the company was months from insolvency. Jobs’ first move was radical simplification of the H1 portfolio (collapsing dozens of Mac models into four) to stabilize and protect the core business. That’s classic H1 discipline.

The “Think Different” campaign and the iMac were H1 and early H2 simultaneously: stabilize the existing customer base while beginning to build toward a broader cultural relevance that would attract new buyers. The H3 bets, including the work that would eventually become iPod and then iPhone, were already being explored in parallel, even before the core was fully stable.

What made Apple’s recovery remarkable wasn’t any single product decision. It was running all three horizons at once under severe resource constraints, without letting H3 thinking crowd out the H1 discipline needed to survive the near term, and without letting H1 pressure eliminate the H3 bets that created the next decade of growth.

When the Framework Falls Short

The original model assumed each horizon took roughly a fixed period to mature: H1 was the current year, H2 was two to five years, H3 was five to ten years. In software and digital businesses, those timescales have compressed significantly. A startup can move from H3 concept to H1 scale in eighteen months with sufficient capital. This changes the math on how long incumbents have to respond to emerging threats.

The framework also doesn’t resolve the organizational tension it identifies. It describes the need to manage all three horizons simultaneously and notes that different horizons require different incentive structures, management approaches, and risk tolerances. But translating that into organizational design (who owns H3, how H2 teams are structured, how you prevent H1 incentives from colonizing everything) requires implementation work that the model doesn’t provide.

Defining which horizon a given initiative belongs to is more contested in practice than the framework suggests. Leadership teams regularly disagree about whether something is an H2 investment being scaled or an H3 experiment that hasn’t proven itself yet. The disagreement often reflects a genuine uncertainty about the initiative’s commercial maturity, and the framework doesn’t give you a reliable way to resolve it.

When to Use It (and When to Reach for Something Else)

Three Horizons is most useful when you need to have an explicit conversation about innovation investment allocation over time. It’s particularly valuable for organizations that are doing well in the present but have noticed that their pipeline of future options is thin or absent.

It’s less useful if you need to prioritize within a single horizon. For that, BCG Matrix or GE-McKinsey are better tools. If you’re thinking about which growth directions to pursue (rather than which horizons to invest in), Ansoff Matrix frames the question more practically. If you’re looking to escape a competitive market rather than evolve within one, Blue Ocean Strategy is the right conversation.

The Three Horizons Model works best as a recurring conversation, not a one-time planning exercise. Run it annually. Map your portfolio honestly. Ask where H2 is dying from underinvestment and where H3 is being stripped to fund short-term H1 performance. The answers are almost always uncomfortable, which means they’re almost always worth hearing.

See these frameworks in action: Marketing Case Studies