Unsure how to build a business innovation strategy that goes beyond buzzwords? Split your ideas across three horizons, run each with its own rules, and test fast. That balance keeps today’s revenue safe while you search for what’s next.
This guide is part of the wider topic of business strategy, and it focuses on how to organise innovation inside an existing company.
| Item | Value |
|---|---|
| Horizon 1 delivery window in the older time-based model | 3-12 months |
| Horizon 2 delivery window in the same model | 24-36 months |
| Horizon 3 delivery window in the same model | 36-72 months |
| Size of a lean Horizon 3 team | Fewer than 5 people |
| Customer conversations such a team can hold | 100+ in 10 weeks |
| Speedup claimed for lean corporate innovation | Several times faster, as the cited source claims |
The time windows come from Steve Blank’s post on the Three Horizons model, and the team-size and customer-conversation figures come from his lean innovation management essay on VentureBeat. They are one author’s examples and claims, and figures like these can change, so check the sources before you rely on them.
What a business innovation strategy has to decide
A business innovation strategy answers two questions: where the company will look for new growth, and how it will protect the business that pays the bills today. Most of the difficulty sits in the second half of that sentence.
In his 2015 essay on lean innovation management, Steve Blank notes that companies pursuing innovation can buy, build, partner or use open innovation. Each route is a different way to get new capability:
- Buy: acquire a company or product that already exists.
- Build: develop the capability internally.
- Partner: share the work and the risk with another organisation.
- Open innovation: look outside the company for ideas and technology.
Choosing a route is only half the job. The same essay points to the ambidextrous organization, an idea from O’Reilly and Tushman, which says companies that want continuous innovation must execute their core business model while innovating in parallel.
The three horizons as a planning frame
The Three Horizons model was developed by Baghai, Coley and White, and Blank describes the idea in his 2015 essay as a way to spread innovation across three categories. His 2019 post gives these working definitions:
- Horizon 1 provides continuous innovation to a company’s existing business model and core capabilities.
- Horizon 2 extends the existing business model and core capabilities to new customers, markets or targets.
- Horizon 3 creates new capabilities to take advantage of, respond to, or counter disruptive opportunities.
The same post says each horizon needs a different focus, different management, different tools and different goals. It adds that McKinsey suggested a company allocate its research and development dollars and resources across all three horizons to stay competitive in the long run.
Treat this as a portfolio decision. Your plan should show what share of effort goes to improving the core, extending it, and searching for something new.

Why do time labels no longer define the horizons?
Older versions of the model tied each horizon to a delivery time. According to Blank’s 2019 post, some organizations used these windows:
- Horizon 1: new features delivered in 3-12 months.
- Horizon 2: business model extensions 24-36 months out.
- Horizon 3: new disruptive products and business models 36-72 months out.
Blank argues this time-based definition made sense when disruptive ideas took years to research, engineer and deliver. His claim is that disruption can now be delivered as fast as ideas for the existing product, so the horizons are not bound by time.
His example is Uber, which took existing technology (a smartphone app and drivers) and built a unique business model on top of it. He also lists Airbnb, Craigslist, Tesla and machine learning solutions as disruption built from existing technologies in short periods, in the same post.
The practical lesson for your plan is to stop sorting ideas by how long they will take. If you want the underlying logic of how newcomers unseat incumbents, read up on disruptive innovation theory.
Running each horizon with its own rules
Blank’s lean innovation essay replaces the time labels with a question about the business model. As that essay frames it, a horizon is defined by whether the business model is being executed or searched for:
- Horizon 1 supports existing business models, with repeatable and scalable processes, procedures, incentives and KPIs to execute and measure the model.
- Horizon 2 extends existing businesses with partially known business models, and works by pattern recognition and experimentation inside the current model.
- Horizon 3 is focused on unknown business models.
Horizon 3 groups, the essay says, need to be physically separate from operating divisions, for example in a corporate incubator or their own facility. They also need their own plans, procedures, policies, incentives and KPIs.
The tools differ too. Horizon 1 product management uses existing tools such as StageGate, while Horizons 2 and 3 use lean tools such as Lean LaunchPad or the NSF I-Corps, per the same article. Small Horizon 3 teams of fewer than 5 people can talk to 100+ customers in 10 weeks, so a company can afford to run many in parallel.
Making the strategy stick inside a large company
Blank warns that early attempts to overlay lean startup tools on large organisations produced innovation theater: great projects and wonderful press releases, but no real substantive change in product trajectory. A written strategy has to deal with the following points to avoid that outcome, according to his article.
- Support staff: people from legal, finance and procurement are assigned to work inside Horizon 3 teams, to help them navigate to a “yes”.
- Incentives: if supporting Horizon 2 and 3 is not part of Horizon 1 goals and incentives, there is no real commitment to corporate innovation.
- Cleanup: successful projects leave technical and organisational debt, so a dedicated corporate group should refactor it.
- Scaling: folding the founders back into a Horizon 1 organisation is typically a bad idea for all involved.
Blank also claims that lean innovation management can deliver 10x the number of initiatives in 1/5 the time, a 50x speedup, in the 2015 essay. Treat that as one practitioner’s claim, not a benchmark.
A practical sequence for drafting your plan
Use the model as a checklist rather than a template. A simple order of work looks like this:
- List every current initiative and tag it Horizon 1, 2 or 3 by whether its business model is known, partly known or unknown.
- Decide what share of resources each horizon should receive, as the model suggests.
- Give each horizon its own goals, metrics and tools instead of one scorecard for all.
- Assign support people and align Horizon 1 incentives with the new work.
- Plan in advance what happens when a project succeeds, including who owns the cleanup.
Turning this plan into repeatable routines is the subject of the innovation management process.
Frequently asked questions
What is the Three Horizons model?
Show answer
It is a way to spread innovation across existing business, extensions of it, and new disruptive capabilities. Blank credits Baghai, Coley and White for it in his 2015 essay.
Does a business innovation strategy need a separate team?
Show answer
For Horizon 3 work, Blank says groups need to be physically separate from operating divisions, with their own plans, policies, incentives and KPIs. They still receive help from legal, finance and procurement staff, as described on VentureBeat.
Are the horizons still tied to time?
Show answer
Blank argues not. In his January 2019 post he says Horizon 3 ideas can be delivered as fast as Horizon 1 ideas.
What is innovation theater?
Show answer
It is activity that looks innovative, such as projects and press releases, without changing the product trajectory. Blank describes it in his 2015 essay as the result of overlaying lean tools on a corporation without adapting the structure around them.



