Three horizons framework

The three horizons framework is a portfolio lens that splits effort and spend across three buckets: the core business earning today, the emerging business scaling next, and the speculative bets that may never pay. It is an allocation view, not a plan.

It came out of The Alchemy of Growth, the 1999 book by Mehrdad Baghai, Stephen Coley and David White, and became a standard way of describing where a company’s money actually sits.

The value is diagnostic. Lay your projects on the three buckets and the imbalance usually shows up immediately — horizon one always wins the argument for resources.

That is the point of the lens. It makes the quiet starvation of horizon three visible on one page, before the board notices it as a growth gap three years later.

Key takeaways

• Horizon one is the core business that funds everything else and dominates reported revenue.
• Horizon two is the emerging business that already has customers but not yet scale.
• Horizon three is the speculative set of bets, most of which are expected to fail.
• The framework measures allocation, so each horizon needs its own success test.

How it works

The three horizons framework works by sorting every initiative into one of three buckets, then comparing the split against your stated growth ambition. Where the two disagree, the allocation is the problem, not the ambition.

Horizon one holds the products and contracts that pay the bills now. It is judged on margin, retention and efficiency, and it absorbs the bulk of management attention — its problems are the urgent ones.

Horizon two holds offers with real customers but no scale yet. It is judged on growth rate and unit economics, not on profit, and it fails when it is measured with horizon one’s yardstick.

Horizon three holds options. Judge it on learning speed and on how cheaply a bet can be killed — most of them should be.

The buckets are not equal in count, either. A healthy portfolio usually carries a handful of horizon one lines, several horizon two offers and many small horizon three bets, most of which will close.

Spending data shows the usual tilt. Of the $692 billion United States businesses spent on domestic research and development in 2022, 79% went to development rather than research, with only 6% on basic work.

Treat the lens as a view onto the innovation portfolio, not as a schedule. An innovation strategy decides what to back; the three horizons framework only shows what you have backed so far.

Examples

The lens travels well across sectors. It is used to read a software company’s product mix, a service provider’s delivery base, and an outsourcing buyer’s split between proven offshore markets and newer ones.

Horizon one in outsourcing is the mature contract. A finance and accounting account in Manila running at steady margin pays for everything else, and core outsourcing decisions sit squarely in this bucket.

Horizon two is the offer that has traction but no scale. Analytics-led service lines and managed AI operations have sat here for most providers since 2023, growing fast on a small base.

Horizon three is where emerging BPO destinations belong. A pilot delivery site in a market you have never staffed is an option with a real chance of closing, priced accordingly.

Transformational growth is the stated reason most companies keep a horizon three at all — and it is the first budget line cut when horizon one has a bad quarter.

Reading the split also tells you what to hand out. Work sitting in horizon one and outside your core competency framework is the obvious outsourcing candidate, while horizon three rarely survives a vendor hand-off.

Steve Blank’s 2019 critique is worth reading against all three. He sets out the classic windows of 3 to 12 months for horizon one and 36 to 72 for horizon three, then argues the horizons are no longer bound by time.

That matters for how you use the lens. Read the buckets as risk and maturity, not as a calendar, and the framework survives the criticism intact.

Related terms

Several adjacent terms describe parts of the same allocation question. The distinctions below matter because a portfolio lens and a growth plan are routinely confused, and the confusion is what turns the framework into a timeline.

• Innovation portfolio: the full set of initiatives the lens is applied to.
• Innovation strategy: the choices about where to play, which the lens then tests.
• Transformational growth: the step change horizon three is meant to produce.
• Core competency framework: the capability view that decides what belongs in horizon one.
• Growth loops: the compounding mechanics that move an offer from horizon two to one.

FAQ

Five questions recur: what each horizon contains, whether the horizons are time periods, how to split spend, what belongs in horizon three, and how outsourcing decisions map onto the lens.

Are the three horizons time periods?

No. They describe maturity and risk, not dates on a calendar. Treating them as timeframes is the most common misreading, and it pushes genuinely new work into a queue it never leaves.

How much should each horizon receive?

There is no universal split, and published rules of thumb are starting points rather than findings. Set the ratio against your own growth target, then check it annually. A split that never changes usually means nobody is reading it.

What belongs in horizon three?

Work with real uncertainty about whether it will ever earn: new markets, new delivery models, unproven technology. If you can forecast it confidently, it is horizon two.

Can a service business use the framework?

Yes. Delivery lines, service offers and new geographies sort into the three buckets as cleanly as products do, and the starvation pattern is identical.

Who owns the horizon split?

The executive team, because the split is a funding decision. Pushing it down to business-unit leaders returns a portfolio where every bucket is horizon one.

When mapping delivery options across all three horizons, start with the market view at Outsource Accelerator.