Every R&D leader eventually faces the same uncomfortable arithmetic. The business needs better products next quarter, credible new ventures in three years, and some bet on a future that may not arrive for a decade. All three compete for the same budget, the same talent, and the same executive attention.

Left unmanaged, this competition has a predictable winner. The near term almost always prevails. Its returns are measurable, its customers are already paying, and its advocates sit in the rooms where decisions are made. Long-horizon work is rarely cancelled outright. It gets starved, one deferred milestone and one reassigned engineer at a time.

Organizations that sustain breakthrough development do not solve this by caring more about the future. They treat innovation as a portfolio spanning distinct time horizons, each with its own logic, metrics, and management system. Understanding how those horizons differ, how to fund them, and how to structure around them is what separates firms that renew themselves from those that simply optimize their way into decline.

The three-horizon framework, popularized by McKinsey consultants Mehrdad Baghai, Stephen Coley, and David White in The Alchemy of Growth, divides innovation activity by time to impact. Horizon 1 extends and defends the core business. Horizon 2 builds emerging businesses that are already gaining traction. Horizon 3 seeds options on the future: research programs, experiments, and early ventures that may or may not become businesses at all.

The framework's real value lies less in the categories themselves than in the recognition that each horizon runs on different logic. Horizon 1 work is about execution and efficiency. Uncertainty is low, so the right tools are business cases, stage-gates, and return-on-investment thresholds. Horizon 3 work is about learning. Uncertainty is so high that a business case is fiction, and the meaningful question shifts from 'what will this earn?' to 'what will this teach us, and how cheaply?'

Horizon 2 occupies the most dangerous position. These initiatives have moved beyond exploration and show early market evidence, but they lack the scale to compete with the core on its own terms. Many technology transfer efforts die here, after the science has been validated but before the business model has been proven. They are too mature to be protected as experiments and too small to justify themselves as businesses, so they fall into the gap between research budgets and operating budgets.

This is why horizons are best understood as stages of maturity and uncertainty rather than strictly as calendar time. A Horizon 3 project in a fast-moving field may reach market in eighteen months. A Horizon 1 improvement in pharmaceuticals may take five years. What defines the horizon is how much remains unknown, and therefore which management tools are appropriate for reducing it. Applying the wrong tools to the wrong horizon is the most common portfolio failure.

Takeaway

Horizons describe levels of uncertainty, not dates on a calendar. The central management discipline is matching each initiative to the tools that fit how much remains unknown.

Once horizons are defined, the obvious question is how much each should receive. The most cited benchmark comes from research by Bansi Nagji and Geoff Tuff, published in Harvard Business Review. They found that outperforming firms tended to allocate roughly 70 percent of innovation resources to core initiatives, 20 percent to adjacent ones, and 10 percent to transformational bets. Strikingly, returns ran in roughly the inverse proportion, with transformational work generating a disproportionate share of long-term value.

The ratio is a starting point, not a prescription. A consumer goods company in a stable market might reasonably sit closer to 80-15-5. A technology firm facing platform disruption may need to push well beyond 10 percent in its exploratory horizon. The right mix depends on how quickly the core is eroding, how much competitive advantage depends on new technology, and how much risk the organization can absorb.

More important than the exact numbers is the mechanism that protects them. When all horizons draw from a single pool, near-term pressures reliably raid long-term budgets during every difficult quarter. Effective organizations ring-fence exploratory funding, committing it at the portfolio level so individual projects do not have to justify themselves against core business returns. They also release Horizon 3 money in tranches tied to learning milestones, much as venture capitalists fund startups in successive rounds.

Allocation is not only about money. The scarcest resource is often talent, particularly experienced people who understand both the technology and the customer. Horizon 1 managers naturally guard their best engineers. Without deliberate mechanisms for rotating capable people into emerging ventures, long-horizon projects end up with generous budgets and no one able to spend them well. Executive attention deserves the same scrutiny, since what leaders review regularly is what the organization learns to value.

Takeaway

A budget that is not protected is merely a suggestion. The ratio across horizons matters less than the mechanism that prevents near-term pressure from consuming it.

Even well-funded portfolios fail when every horizon runs through the same organizational machinery. The processes that make a core business excellent, including tight budgeting, standardized metrics, and risk-averse approval chains, are precisely the ones that suffocate early-stage ventures. Uncertain work needs different rhythms: faster experimentation, tolerance for failure, and leaders rewarded for what they learn rather than what they deliver this year. Asking a Horizon 3 team to report through Horizon 1 governance is like evaluating a seedling by its fruit yield.

Charles O'Reilly and Michael Tushman described the most effective response as the ambidextrous organization. Exploratory units are structurally separated, with their own processes, incentives, and cultures, yet remain tightly integrated at the senior leadership level. Separation gives new ventures room to operate by different rules. Integration ensures they can still draw on the parent company's distribution channels, brands, manufacturing capacity, and customer relationships, which are often their greatest advantage over independent startups.

IBM's Emerging Business Opportunities program, launched around 2000, illustrated this logic in practice. New ventures received dedicated leadership, often seasoned executives, along with protected funding and reviews focused on market learning rather than short-term profit. The program helped build businesses in areas such as life sciences and Linux that would have struggled inside divisions measuring success by quarterly revenue. Equally important, ventures that proved themselves eventually moved into the operating divisions to scale.

Structure must also account for the boundaries of the firm. Following the logic of open innovation, some Horizon 3 opportunities are better pursued outside entirely, through university partnerships, corporate venture investments, licensing, or spin-offs. The design question is not simply where a team sits on the org chart, but where each initiative has the best chance of maturing, and how its value eventually flows back to the organization.

Takeaway

Different horizons need different homes. Separate where the operating logic must differ, integrate where the parent's assets create advantage, and plan the handoffs between them.

Managing multiple horizons is less about predicting the future than about building an organization that can survive several possible ones. The three-horizon framework provides the vocabulary, but the discipline lies in execution.

A systematic approach starts with classifying every initiative by its level of uncertainty and assigning matching metrics. It continues with explicit, protected allocation ratios, reviewed annually against the health of the core business. It concludes with structures that separate exploratory work from operational routines while connecting it to the assets that give the parent organization its advantage.

Most innovation portfolios are not designed. They simply accumulate, and the core wins by default. Making the balance deliberate is a leadership choice, and few choices shape an organization's long-term trajectory more.