The three horizons model was first published in The Alchemy of Growth by Merhdad Baghai, Stephen Coley, and David White in 1999. The fundamental idea behind the model is that we need to be thinking about innovation across three time frames. Sheldon Laube recently wrote a good post on this model as well, which included a nice visualisation of it, which apparently comes from Innovation Tournaments by Christian Terwiesch and Karl Ulrichwhich. I have adapted it slightly here:

When you innovate using the three horizons framework, the first horizon involves implementing innovations that improve your current operations, horizon two innovations are those that extend your current competencies into new, related markets, and horizon three innovations are the ones that will change the nature of your industry. In general, H1 innovations tend to be incremental, while H3 are more often radical innovations. There are several key ideas that arise when using the three horizons model.

• Innovation must be linked to strategy: tools like the 3 Horizons framework are good ways to connect innovation with strategy. As you develop strategy, it will provide some guidance about where you want to end up in the future – you can then orient your H2 & H3 efforts in alignment with these future goals.
• Think of innovation efforts as a portfolio: we need to have innovation taking place across all three time horizons. Thinking about this directly helps us ensure that we have a balanced portfolio. We need to be innovating enough in H1 to make sure that we stay competitive. H2 is the area where we use existing competencies to drive growth. To protect ourselves against disruptive innovations, we also need to be placing some small bets (or taking real options) on where our industry might go in the future – this is the H3 domain.
• A balanced portfolio does not mean 1/3 or our resources in each area: Google uses a 70/20/10 split – most of their innovation resources are put into improving what they are currently doing – H1 initiatives. They have relatively less invested in the longer-term ideas. In most cases, these seem like reasonable investment proportions. If your environment is particularly turbulent, you can increase the investments into H2 and H3 innovation – it all depends on the situation you face.

The tricky part in all of this is management. It takes different skills to innovate in H1, where most of the ideas are incremental, than it does to innovate in H3, where many of the ideas are more radical. Nevertheless, organisations that are successful over time figure out how to do this well.

Thinking of innovation as a portfolio is one good way to manage the process more effectively.