The three horizons of innovation is probably the most widely used portfolio framework in corporate boardrooms. Originally published in The Alchemy of Growth by Baghai, Coley, and White in 1999 and popularized by McKinsey, the three horizons model offers a simple structure: Horizon 1 is your core business, Horizon 2 is your emerging opportunities, Horizon 3 is your future bets. Simple. Clean. Easy to put on a slide.
That simplicity is both its greatest strength and its biggest weakness. After using the three horizons framework at 50+ industrial companies over 25+ years, I have a nuanced view. It is genuinely useful for specific situations. It is actively misleading in others. And most companies use it in exactly the wrong context.
This article gives you my honest practitioner assessment: what the model gets right, where it breaks down, and when to reach for a different tool.
Request a Strategy Call about your innovation portfolio
In 30 minutes, I'll review how you use frameworks like Three Horizons and identify where your portfolio planning needs sharper tools. Or book a leadership workshop where your team builds a portfolio view that works for your industry.
What the three horizons model actually is
Before assessing the framework, let me make sure we are looking at the same thing.
The McKinsey three horizons model organizes your innovation portfolio into three categories based on time to revenue:
Horizon 1: the core. Your current business models generating revenue today. The products and services that pay everyone’s salary. Optimization, efficiency improvements, incremental extensions. Time horizon: now to 12 months.
Horizon 2: emerging growth. New opportunities that are gaining traction but are not yet at full scale. These might be new market segments, new product lines, or new business models that show early revenue. Time horizon: 2-3 years to significant revenue.
Horizon 3: future bets. Early-stage ideas, experiments, research projects. No revenue yet, possibly no product yet. High uncertainty, potentially high return. Time horizon: 5-10 years.
The visual is a set of overlapping S-curves. As H1 eventually matures and declines, H2 takes over as the new core, and H3 matures into H2. It suggests a continuous pipeline: you are always growing the next horizon while maintaining the current one.
On paper, this is elegant. In practice, it gets complicated.
What the three horizons framework gets right
I want to be fair to the framework before I criticize it. There are real situations where Three Horizons earns its place.
Board communication
Three Horizons speaks the language of boards. Directors think in planning cycles, budget allocations, and time to return. When a head of innovation walks into a board meeting and says “here is what generates revenue now, here is what we are building for the next three years, here is what we are betting on for the long term,” the board gets it immediately. No training required. No new vocabulary to learn.
At one chemical company, the innovation director had been struggling to get board support for long-term R&D projects. The board kept asking “when does this make money?” He mapped everything onto Three Horizons and suddenly the conversation shifted from “why are we spending money on this?” to “do we have enough in H3?” That reframing happened in a single meeting.
Making the implicit explicit
Most companies have an innovation portfolio. They just do not call it that. Projects sit in different departments, funded through different budgets, managed by different people. Nobody sees the full picture.
Three Horizons forces a conversation about what exists across all three time frames. When I ask a leadership team “show me your H3 projects,” and they realize they have none, that is a diagnostic moment. When they see that 95% of their budget sits in H1 and they have been talking about “being more innovative” for three years, the gap between rhetoric and reality becomes visible. The model is useful for exposing that gap.
Annual planning context
During annual strategy cycles, Three Horizons provides a structure for portfolio allocation conversations. How much do we invest in maintaining and optimizing the core? How much do we put toward scaling our emerging opportunities? How much do we allocate to early-stage exploration?
The framework does not answer those questions, but it creates the categories that make the questions possible. That is worth something.
What the three horizons framework gets wrong
Here is where my practitioner experience diverges from the textbook. These are not theoretical objections. They are patterns I see repeatedly in industrial and B2B companies.
The sequential progression myth
The three horizons model implies a neat progression: H3 ideas gradually mature into H2, and H2 matures into H1. Like a conveyor belt. One stage flows into the next.
Reality does not work that way. Some H3 ideas skip H2 entirely and disrupt H1 directly. Think about what digital photography did to film. It did not gradually mature through an “emerging” phase at Kodak. It replaced the core business faster than anyone expected.
In manufacturing, I see the opposite pattern too: H3 projects that stay in H3 for a decade because the technology works in the lab but cannot be manufactured at scale. There is no smooth S-curve when you need €2M in tooling to move from prototype to production.
The timing illusion
Three Horizons suggests that you can plan when innovations will mature. H2 in 2-3 years. H3 in 5-10 years. That implies a level of predictability that simply does not exist for anything beyond incremental innovation.
At one agricultural equipment manufacturer, the leadership team placed an autonomous harvesting project in H3 with a “5-7 year” timeline. Three years later, a competitor launched a working product. Their H3 was suddenly someone else’s H1. The framework had given them a false sense that they had time.
This is the danger of the three horizons of growth model: it can make leadership comfortable with a timeline that the market does not respect.
The missing evidence dimension
Three Horizons organizes projects by time. It tells you nothing about evidence quality. An H2 project could be well-validated with real customer data and a working prototype. Or it could be an executive’s pet project with nothing but a business case built on assumptions. Three Horizons treats both the same.
This is a serious problem for portfolio governance. When you sit in a portfolio review meeting and need to decide which projects get continued funding and which get killed, “it is an H2 project” is not enough information. You need to know what has been tested, what evidence exists, and what the remaining risks are.
Manufacturing reality breaks the S-curves
The neat S-curve visual assumes that projects can start small, grow gradually, and scale smoothly. That works for software. It does not work for industries where:
- Development cycles run 3-5 years before you have a product to sell
- Tooling investments of €500K-€2M are required before you can test at production scale
- Regulatory approvals add 6-12 months after development is complete
- Customer qualification cycles require another 6-12 months after regulatory approval
In manufacturing companies, there is no smooth curve. There are large capital commitments followed by long periods of no visible progress, followed by a step-change when the product finally reaches market. The three horizons model does not account for this reality, and it can create false expectations about how innovation timelines actually work in capital-intensive industries.
No mechanism for kill decisions
Three Horizons tells you what time horizon a project lives in. It does not tell you when to stop investing. There are no built-in criteria for when an H3 project should be killed, when an H2 project is not gaining traction fast enough, or when an H1 optimization has diminishing returns.
This gap matters. One of the most common portfolio mistakes I see is keeping mediocre projects alive because they are “our H2 pipeline.” The framework gives them a category to hide in. Without evidence-based kill criteria, Three Horizons can actually protect projects that should have been terminated.
When to use the three horizons framework
Not every tool needs to do everything. Three Horizons has clear sweet spots:
| Use Three Horizons for | Because |
|---|---|
| Board presentations | Boards understand time-based categories |
| Annual strategic planning | Creates natural buckets for allocation discussions |
| Executive education | Simple enough to explain in 10 minutes |
| First portfolio conversation | Makes innovation spending visible for the first time |
| Communicating to non-innovation stakeholders | No jargon, no training required |
The pattern: Three Horizons works best as a communication tool, not a governance tool. Use it when you need to tell a story about your innovation pipeline. Do not use it when you need to make investment decisions.
When to use something else
| For this question | Use this instead |
|---|---|
| Which projects should we continue or kill? | Business Portfolio Map (evidence + return axes) |
| How much should we allocate to core vs. transformational? | Innovation Ambition Matrix |
| Is our portfolio balanced between explore and exploit? | Business Portfolio Map |
| How do we govern our innovation portfolio? | Evidence-based review process with kill criteria |
| Which business units deserve more investment? | BCG Matrix or GE-McKinsey Matrix |
For a detailed comparison of how these frameworks work together, see the full framework comparison.
Combining Three Horizons with the Business Portfolio Map
Here is what I actually do in practice. I do not throw Three Horizons away. I combine it with the Business Portfolio Map from The Invincible Company to cover both dimensions that matter: time and evidence.
Three Horizons for the boardroom. Once or twice a year, during annual planning or board presentations, I use Three Horizons to tell the pipeline story. “Here is our core business, here is what we are growing, here is what we are exploring.” Boards get it. They can ask the right questions. The conversation stays productive.
Business Portfolio Map for the review room. Monthly or quarterly, when the innovation team and leadership sit down to review the portfolio and make actual decisions, I use the Business Portfolio Map. It shows which projects have real evidence behind them and which are still running on assumptions. It makes the explore vs. exploit balance visible. It gives you something concrete to base continue-or-kill decisions on.
For a deeper look at how the Business Portfolio Map works as a standalone tool, see the practitioner’s guide to the Business Portfolio Map.
The combination works because each framework compensates for what the other lacks:
| Dimension | Three Horizons | Business Portfolio Map |
|---|---|---|
| Time perspective | Strong (H1/H2/H3 timeline) | Weak (no time axis) |
| Evidence quality | Absent | Strong (evidence axis) |
| Kill decisions | No mechanism | Supports evidence-based decisions |
| Board communication | Excellent | Requires explanation |
| Portfolio balance | By time only | By explore/exploit and evidence |
A practical example
At an industrial components manufacturer, I facilitated a portfolio review where we used both tools. On the Three Horizons view, the board saw a healthy pipeline: 8 H1 projects, 4 H2 projects, 3 H3 projects. It looked balanced.
Then we mapped the same 15 projects on the Business Portfolio Map. Two of the four H2 projects had been in “emerging” status for over two years with no customer validation whatsoever. They had a time label but no evidence. On the Portfolio Map, they sat in the high-return, low-evidence quadrant, which meant high risk with nothing to show for it.
That second view triggered a conversation the Three Horizons view never would have. The team decided to set 90-day validation milestones for both projects. If they could not produce customer evidence within that window, the projects would be stopped. One survived. One did not.
That is the difference between a communication framework and a governance framework. You need both.
For guidance on making kill decisions without damaging your innovation culture, see how to kill innovation projects the right way.
Starting with your portfolio
If you are currently using Three Horizons as your primary portfolio tool, here is how to assess whether that is enough:
-
Can you make kill decisions with it? If your portfolio reviews end with “let us keep going on everything,” Three Horizons is not giving you the governance you need.
-
Do you know the evidence status of each project? If you cannot distinguish between an H2 project with strong customer validation and an H2 project based on assumptions, you have a visibility gap.
-
Does your timeline match reality? If you placed a project in H3 two years ago and it is still in H3 today with the same assumptions, the time-based label is not helping you.
-
Are your manufacturing development cycles reflected? If your product takes 3 years to develop and you are using standard H1/H2/H3 time windows, the framework is misleading your board about when returns will appear.
If you answer “no” to two or more of these questions, it is time to add a governance layer on top of Three Horizons. The model itself is fine for what it was designed for. The problem is using it for things it was never designed to do.
Understanding your innovation readiness helps determine which portfolio tools fit your organization’s maturity level. And whichever portfolio framework you choose, the individual initiatives still need a solid business model and a systematic approach to testing assumptions before they earn continued investment.
