Strategy & competition
Three Horizons
A portfolio view of growth that asks a business to manage three horizons at once: defending and extending the core, building emerging businesses, and seeding options for the future, so that today's earnings do not quietly starve tomorrow's.
Also known as Three Horizons of Growth, McKinsey Three Horizons. First set out by Mehrdad Baghai, Stephen Coley and David White (McKinsey & Company) in 1999; the primary source is cited in full below.
- Format
- Structural model
- Level
- Corporate · Business unit
- Best for
- Allocate resources · Evaluate options · Plan execution
- Decision stage
- Explore options · Plan · Review
- Difficulty
- Introductory
- Time to apply
- A half-day workshop for a first portfolio view; the value compounds when it becomes a quarterly governance rhythm.
Plate · The model
The components
Horizon 1: defend and extend the core
The mature businesses that generate today's cash flow and profit. They are run for performance: defending competitive position, extending incrementally, and funding everything else.
Signals of strength
Delivers the bulk of current revenue and profit · Managed for margin, share and operational excellence · Innovation is incremental and close to existing customers
Horizon 2: build emerging businesses
Rising businesses with proven concepts, scaling toward material contribution. They consume investment, need entrepreneurial management, and are the bridge between the core and the options portfolio.
Signals of strength
Fast-growing revenue with still-immature economics · Needs protected funding and dedicated leadership · Measured on growth and milestones rather than profit
Horizon 3: create options for the future
The seeds of future businesses: research projects, pilots, alliances and minority stakes. These are options rather than commitments, held cheaply and in numbers, with success measured in learning and the right to play later.
Signals of strength
Small, cheap probes rather than committed businesses · A deliberate spread of bets across plausible futures · Judged on option value and learning, never on next quarter's margin
When it earns its keep
- Growth has stalled or is about to, and you suspect the pipeline of future businesses is empty behind a still-profitable core.
- Annual planning reliably funds the core and starves everything else, and you need a shared language to make that trade-off visible to the board.
- Leadership debates lurch between protecting the crown jewels and chasing shiny new bets, with no framework for holding both concerns at once.
- You are reviewing whether management styles, metrics and talent are matched to the maturity of each business rather than applied uniformly.
And when it doesn't
- You compete where disruption arrives at software speed. Blank's critique bites here: treating Horizon 3 as distant work you can schedule for later is how incumbents get blindsided by rivals who ship Horizon 3 ideas in months.
- You are a single-business startup. There is no portfolio to balance yet; discovery tools such as Lean Canvas and build-measure-learn fit the job better.
- You need to evaluate specific initiatives rather than balance a portfolio. The horizons classify businesses by maturity; they contain no test of whether any individual bet is good.
- The business is in a survival crisis. When cash decides everything, portfolio balance is a conversation for after the turnaround.
How to run it
Before starting, gather the inputs the analysis depends on:
- An honest classification of current businesses and revenue by horizon, including how much of today's profit depends on the core.
- A list of emerging businesses and initiatives with their stage, resourcing and trajectory.
- A view of the market forces likely to erode the core, and on what timescale.
- Clarity on how each horizon will be measured: profit and share for the core, revenue growth for emerging businesses, learning and option value for seeds.
- 1
Classify the portfolio
Map every business and initiative to a horizon. Horizon 1 businesses generate today's cash flow and profit; Horizon 2 businesses are scaling toward material contribution; Horizon 3 holds options: pilots, research, minority stakes. Classify by maturity of the business, and resist classifying by how futuristic the technology sounds.
- 2
Interrogate Horizon 1
Ask whether the core is genuinely defended and extended: competitive position, margin discipline, incremental innovation. A neglected core cannot fund the other horizons, and a coddled one absorbs every pound of investment that comes near it.
- 3
Build Horizon 2 deliberately
Emerging businesses fail in the middle: too big for skunkworks, too small to win budget fights against the core. Give them dedicated entrepreneurial leadership, protected funding and revenue-growth metrics rather than profit targets.
- 4
Seed Horizon 3 cheaply and widely
Options should be numerous, inexpensive and pointed at plausible futures: pilots, partnerships, small stakes. The aim is the right to play later, and success at this stage is measured in learning rather than earnings.
- 5
Check the balance and the staircase
Step back and look for the empty horizon. The book's research found stalls are usually visible years earlier as a thin Horizon 2. Check that options actually graduate: Horizon 3 seeds should feed Horizon 2, and Horizon 2 businesses should be on a path to becoming the next core.
- 6
Govern the transitions
Set a rhythm for promoting initiatives between horizons, killing options that have taught what they can, and protecting each horizon's budget from the others. Managing all three concurrently, with different metrics and managers, is the discipline the model exists to enforce.
Reading the result
A one-page portfolio view showing which businesses occupy each horizon, where the gaps and imbalances sit, and a governance agenda: what to defend, what to scale, what to seed, and how each will be measured.
- Read the horizons as concurrent, never sequential. The book's central instruction is to manage all three at once; a plan that treats them as phases has deferred the future indefinitely.
- Look for the empty horizon. A dominant Horizon 1 with a thin Horizon 2 is a stall in the making, visible years before the revenue line shows it.
- Check that metrics and managers differ by horizon. A Horizon 3 seed judged on this year's margin is being quietly executed, whatever the strategy document says.
A worked example
A regional commercial law firm balances its growth portfolio
A twelve-partner commercial law firm in the north of England earns well from commercial property, corporate and employment work, but fixed-fee competitors and legal-technology products are eroding pricing. At the annual strategy day the partners run a Three Horizons review of everything the firm does and is trying to do.
- Horizon 1: defend and extend the core
- The three established practice areas produce over ninety percent of fee income, with strong client relationships but falling realisation rates as clients push for fixed fees. Defence means pricing discipline, matter-management efficiency and deliberate deepening of the top forty client relationships, none of which is currently anyone's explicit job.
- Horizon 2: build emerging businesses
- One genuine candidate exists: a productised employment-law subscription for SME clients, launched eighteen months ago. It is growing at forty percent a year from a small base but contributes under five percent of income and is run part-time by one partner around her billable work. It has proof of concept and no machinery to scale.
- Horizon 3: create options for the future
- Three options are live in name only: an AI-assisted contract review pilot with a legal-technology vendor, a data-protection advisory offer, and an exploratory referral alliance with a corporate finance boutique. All are funded from partner goodwill and evenings, with no budget, no owner and no kill criteria.
- Check the balance and the staircase
- The portfolio is bottom-heavy in the classic pattern the book warns about: a strong core, one undernourished Horizon 2 business, and hobby-funded options. The binding constraint is partner time, which the billable-hour targets allocate entirely to Horizon 1 by default.
The read. The honest reading is that the firm is one pricing shock away from a stall. The decisions: ring-fence the subscription business with a dedicated partner, a growth target and relief from billable-hour metrics; fund the AI pilot and the data-protection offer properly and kill the alliance, which has taught what it can; and review the portfolio quarterly rather than annually, because legal-technology disruption is unlikely to respect the leisurely timescale the horizon habit assumes.
Pitfalls
- Reading the horizons as a time sequence. They are concurrent portfolios to manage simultaneously; treating them as phases is the misreading that defers the future forever.
- Bolting on a fixed 70/20/10 allocation as if it were part of the model. The book prescribes no ratio; the right balance depends on the industry's clockspeed and the state of the core.
- Judging every horizon with Horizon 1 metrics. Emerging businesses measured on this year's profit die of respectability.
- Treating Horizon 3 as synonymous with distant time. The horizons describe the maturity of a business, and in fast markets an option can become a core-threatening rival within quarters.
- Letting the horizons harden into silos, with an innovation team owning Horizon 3 and the core owning Horizon 1, so that nothing ever transitions between them.
What the critics say
Blank argues the model's implicit time dimension has become a fatal flaw: disruptive Horizon 3 business models can now be assembled from existing technologies and shipped at Horizon 1 speed, as Uber, Airbnb and Netflix did, so incumbents that file disruption under 'later' are structurally late. McKinsey-schooled portfolios, he notes, were defined by an assumption of years that no longer holds.
Blank, S. (2019) 'McKinsey's Three Horizons Model Defined Innovation for Years. Here's Why It No Longer Applies', Harvard Business Review, February 2019.
Futures-studies practitioners found the original framing too firm-centric and too tied to managed growth, and rebuilt it: Curry and Hodgson, then Sharpe, recast the horizons as three ever-present patterns contending in the present (a declining first horizon, an entrepreneurial second, an emergent third). The adaptation is a deliberate departure, and its existence is itself a critique of the original's limits.
Curry, A. and Hodgson, A. (2008) 'Seeing in Multiple Horizons: Connecting Futures to Strategy', Journal of Futures Studies, 13(1), pp. 1–20; Sharpe, B. (2013) Three Horizons: The Patterning of Hope. Axminster: Triarchy Press.
The framework classifies but does not evaluate. It offers no test of whether any given Horizon 2 or 3 bet is sound, so a portfolio can look balanced on the chart while consisting of poor bets in every band.
Sources and further reading
- Baghai, M., Coley, S. and White, D. (1999) The Alchemy of Growth: Practical Insights for Building the Enduring Enterprise. Reading, MA: Perseus Books.
- McKinsey & Company (2009) 'Enduring Ideas: The three horizons of growth', McKinsey Quarterly, December 2009. ↗
- Blank, S. (2019) 'McKinsey's Three Horizons Model Defined Innovation for Years. Here's Why It No Longer Applies', Harvard Business Review, February 2019. ↗
- Curry, A. and Hodgson, A. (2008) 'Seeing in Multiple Horizons: Connecting Futures to Strategy', Journal of Futures Studies, 13(1), pp. 1–20. ↗