The Fatal Flaw of the Three Horizons Model

A version of this article first appeared in the Harvard Business Review

I’m a big fan of McKinsey’s Three Horizons Model of innovation. (if you’re not familiar with it there’s a brief description a few paragraphs down.) It’s one of the quickest ways to describe and prioritize innovation ideas in a large company or government agency.

However, in the 21stcentury the Three Horizons model has a fatal flaw that could put companies out of business and government agencies behind their adversaries. While traditional analysis suggests that Horizon 3 disruptive innovations take years to develop, in today’s world this is no longer the case. The three horizons are not bound by time. Horizon 3 ideas – disruption – can be delivered as fast as ideas for Horizon 1 – existing products.

In order to not be left behind, companies / government agencies need to focus on speed of delivery and deployment across all three horizons.

When first articulated by Baghai, Coley and White in the 20th century, the Three Horizons model was a simple way to explain to senior management the need for an ambidextrous organization – the idea that companies and government agencies need to execute existing business / mission models while simultaneously creating new capabilities.

The Three Horizons provided an incredibly useful taxonomy. The model described innovation occurring in three time horizons:

  • Horizon 1 ideas provide continuous innovation to a company’s existing business model and core capabilities.
  • Horizon 2 ideas extend a company’s existing business/model and core capabilities to new customers, markets or targets.
  • Horizon 3 is the creation of new capabilities to take advantage of or respond to disruptive opportunities or to counter disruption.

Each horizon requires different focus, different management, different tools and different goals. McKinsey suggested that to remain competitive in the long run a company allocate its research and development dollars and resources across all three horizons.

And here’s the big idea. In the past we assigned relative delivery time to each of the Horizons. For example, some organizations defined Horizon 1 as new features that could be delivered in 3-12 months; Horizon 2 as business/mission model extensions 24-36 months out; and Horizon 3 as creating new disruptive products/business/mission models 36-72 months out.  This time-based definition made sense in the 20th century when new disruptive ideas took years to research, engineer and deliver.

That’s no longer true in the 21st century.

Today, disruption Horizon 3 ideas – can be delivered as fast as Horizon 1 ideas.

For example, Uber took existing technology (smartphone app, drivers) but built a unique business model (gig economy disrupting taxis) and the Russians used existing social media tools to wage political warfare. Fast disruption happens by building on existing technologies uniquely configured, packaged and/or delivered, and combining them with a “speed of good-enough deployment as a force multiplier” mindset.

What’s an Example of Rapid Horizon 3 Implementation?
In the commercial space AirBnB, Uber, Craigslist, Tesla, and the explosion of machine learning solutions (built on hardware originally designed for computer graphics (Nvida)) are examples of radical disruption using existing technologies in extremely short periods of time.

In the government space, Russian interference with elections, and China building island bases in the South China Sea as well as repurposing ICBMs as conventional weapons to attack aircraft carriers, are examples of radical disruption using existing technologies deployed in extremely short periods of time.

What’s Different about Rapid Horizon 3 Disruption?
These rapid Horizon 3 deliverables emphasize disruption, asymmetry and most importantly speed, over any other characteristic. Serviceability, maintainability, completeness, scale, etc. are all secondary to speed and asymmetry.

To existing competitors or to existing requirements and acquisition systems they look like minimum viable products – barely finished, iterative and incremental prototypes. But the new products get out of the building, disrupt incumbents and once established, they then refactor and scale. Incumbents now face a new competitor/threat that obsoletes their existing product line/infrastructure/business/mission model.

Why Do the Challengers/new Entrants Have the Edge?
Ironically rapid Horizon 3 disruption is most often used not by the market leaders but by the challengers/new entrants (startups, ISIS, China, Russia, etc.). The new players have no legacy systems to maintain, no cumbersome requirements and acquisition processes, and are single-mindedly focused on disrupting the incumbents.

Four Strategies to Deal With Disruption
For incumbents, there are four ways to counter rapid disruption:

  • Incentivize external resources to focus on your goal/mission. For example, NASA and Commercial Resupply Services with SpaceX and OrbitalATK, Apple and the App Store, DARPA Prize challenges. The large organizations used startups who could rapidly build and deliver products for them – by offering something the startups needed – contracts, a distribution platform, or prizes. This can be a contract with a single startup or a broader net to incentivize many.
  • Combine the existing strengths of a company/agency and its business/mission model by acquiring external innovators who can operate at the speed of the disruptors. For example, Google buying Android. The risk here is that the mismatch of culture, process and incentives may strangle the newly acquired innovation culture.
  • Rapidly copy the new disruptive innovators and use the incumbent’s business/mission model to dominate. For example, Microsoft copying Netscape’s web browser and using its dominance of operating system distribution to win, or Google copying Overture’s pay per click model and using its existing dominance in search to sell ads. The risk here is that copying innovation without understanding the customer problem/mission can result in solutions that miss the target.
  • Innovate better than the disrupters. (Extremely difficult for large companies/government agencies as it is as much a culture/process problem as a technology problem. Startups are born betting it all. Large organizations are executing and protecting the legacy.) Successful examples, Apple and the iPhone, Amazon and Amazon Web Services (AWS). Gov’t agency and armed drones.

Lessons Learned

  • The Three Horizons model is still very useful as a shorthand for prioritizing innovation initiatives.
  • Some Horizon 3 disruptions do take long periods of development
  • However, today many Horizon 3 disruptions can be rapidly implemented by repurposing existing Horizon 1 technologies into new business/mission models
  • Speed of deployment of a disruptive/asymmetric product is a force multiplier
  • The attackers have the advantage, as the incumbents are burdened with legacy
  • Four ways for the incumbents to counter rapid disruption:
    • Incentivize external resources
    • Acquire external innovators
    • Rapidly copy
    • Innovate better than the disrupters

17 Responses

  1. Dear Steve, Thank you for sharing this important insights with regards to the Three Horizon Model of Innovation. I agree that Horizon 3 (and Horizon 2) should involve much faster iteration than traditional waterfall innovation processes and even possibly find success more quickly. However, I suggest one needs to consider the fact that most disruptive ideas will require many iterations before they find a repeatable, scalable, and viable business model (if ever). So this could take many months, even years. Not many have the luxury of being able to parallelise this search (as early stage investors do). I see a lot of “let’s just do it” mentality after minimal real validation. Of course, this does not mean innovators aren’t “getting out of the building,” literally or through the Web, but it should also mean that they are searching not executing. These, as you have explained so clearly before, are very different activities. Cheers, Ashley.

  2. Happy New Year, Steve! What a great New Year gift your insightful post is. I have been struggling with the McKinsey Three Horizons Model for some time. It is still useful, but as you remark we need to apply it less rigidly.
    Disruptive models can apply to all three horizons, and lean startup iterative methods can at times be applied also to Horizon 1 & 2.
    Digitalization, Apps and platform solutions bring new opportunities for radical innovation. Frameworks like the Three Horizons need to be applied with flexibility.
    Thanks so much for sharing your insights. This will need some more mulling over from my side.
    All the best,

  3. Thanks Steve. As always, your insights are appreciated. In this case, I think you are being too kind. One could argue that the McKinsey model was never a good idea. Christensen published The Innovator’s Dilemma in 1997. McKinsey published their Three Horizons model in 2000. The evidence was readily available to indicate that low-end disruptive business model innovation was already happening when McKinsey published their model. Secondly, it gave many large companies further support for their risk aversion by leading them to believe that disruption was too far away to either (1) invest in or (2) be afraid of. In my opinion, the Three Horizons model was the result of a compromise between new thinking and old thinking that large companies are often susceptible to. So what;’s the point? Picking on McKinsey? No. But it should beg the question, “which frameworks are being adopted today due to the size of the author’s megaphone and reputation over better frameworks from new and original thinkers?” Am I being unreasonable?

  4. Hi Steve, I agree companies can get disruptive innovations to market faster than ever before, but it still takes time (perhaps less as well) to scale them and make them profitable. The first Uber pitch deck was 2008, and they’ve still never made $. However, they certainly became a viable service for consumers very quickly and disrupted the taxi industry. As you point out repurposing SW, cloud, mobile tech etc make is faster/easier to create new ventures. A related point is that this tech has also made it easier for new ventures reach their intended audiences.

  5. One of the ‘problems’ with the Horizon model is that it has too many different interpretations and usages, particularly in a corporate context. It has morphed and changed over time and usage which is normal.

    I have seen (and used) the model to describe ‘disruptiveness – how much it challenges the existing products, services, offers, capabilities, resources…’ and ‘radicalness – the degree to which it is outside of the core existing products and services’. Both these additional definitions don’t look at it through a time lens per se. You also have existing, adjacent, radical framings and how much resources should be allocated strategically.

    Add to this that it is dependent upon the meaning to the specific company as to what is regarded as actually ‘a long time’, ‘radical’, ‘disruptive’ ‘a portfolio’ etc and you have a messy tool that can become locally ‘meaning subjective’ on top of all this.

    I find it is ultimately a useful tool to get a high level overview of innovation efforts and the literature should take stock of the different iterations of the meanings and usages of the tool in order to arrive at some clarity at the different lenses (time being one of them), that can be kept at the forefront to help manage innovation.

  6. Brilliant analysis. Well Presented. Great stuff! Jerome S. Engel U.C. Berkeley @jeromesengel

  7. I was thinking about the Horizon theory whilst reading Clayton Christensen’s book The Innovators Solution and I think both theories have a fundamental flaw. In horizon there is the point that you make around timeframes, which is great but horizon theory also didnt deal with the cultural issues of internal teams creating new growth businesses. (ie they cant and they fail often) This is dealt with more in the innovators solution but the solution proposed there is wrong. In the book Christensen’s solution is to build businesses that quickly become profitable to avoid the potential drastic cuts a board or leadership might make in the future to kill the project. Think about how that business launched internally would compete with a standalone startup or external company without these pressures. The external team will always out execute by focusing more on the customer and over delivering, on growth and market share. It points towards a model where corporates should co invest in spin offs or corporate venturing (problem here is not enough of the upside is captured by the corporate). What are your thoughts on this execution problem Steve?

  8. Steve,

    Thank you for posting your note on the Three Horizons model and for your kind words about it. I am delighted that you have found it helpful and that you are a fan of it. I am equally a fan of your work.

    I wanted to point out a few points about the framework and its use that might convince you to alter the title of your post though.

    First, we created the Three Horizons as a growth strategy framework, not an innovation framework. There are important differences between the two. I agree entirely with your points about the shifting nature of innovation but our model was not so much about that. It was created to address the pipeline of growth businesses in a portfolio and how they should be managed inside the company. To be honest, the definition of the Three Horizons in your post are not the same as our definitions from the book. I am not sure about the origin of your version of the definitions.

    Second, you are correct that the time to Horizon 3 has been speeding up. In the book, we pointed out that every company needs to decide what the right timing for each Horizon should be. For example, we believed some forest product companies we looking at Horizon 3 being 10-20 years away while semi-conductor companies looked at Horizon 3 being 18 months away. The way to address your observation is by adjusting the timing of the Horizons. They were never intended to be fixed as your post appears to suggest.

    Third, it is true that we published the Three Horizons model in various forms during the 1995-1997 period (not 2000 as one of the comments above suggests). While technically those years fall in the 20th century, it is a bit misleading to portray our thinking as being limited to or characteristic of the 20th century. In fact, the Three Horizons were quite forward thinking at the time and the framework really grew in popularity and usage in the first decade of the 21st century. It is probably more accurate to refer to it as an early 21st century framework.

    Like you, we have been advising our clients about another important change in the last two decades. Not only have the timelines accelerated but the intensity of innovation in the market ecosystem is now even greater than before. No company can out-innovate the entire market. Therefore, we strongly advise greater reliance on external identification of innovation to accelerate the timing of the company response. I completely endorse your recommendations.

    Where I beg to differ is that, if properly understood and applied, the Three Horizons model is perfectly compatible today’s competitive reality. There is no “fatal flaw” to “fix”. I would suggest modifying the title for your post to focus on the nature of accelerated context. What you are suggesting is perfectly compatible with our description of the model two decades ago. Sensationalizing the message only takes away from your extremely valid and very helpful prescriptions.

    With greatest respect and admiration for your body of work,
    Mehrdad Baghai
    (co-author of the Three Horizons model)

    • All relevant points and great to read the facts.
      Titled changed.

      Your model is not only a defacto standard to explain innovation to leadership, but over the years time have become associated with each of the horizons.
      I was trying to send a message that it was detrimental to do so

      • Thanks Steve. Makes total sense. And very much appreciate the changed wording. You are a true gentleman.

        • Steve and Mehrdad, I found this exchange to be confusing. if I understand Mehrdad’s post correctly, he is saying that the x-axis in the Three Horizons framework should, in fact, be viewed as representing Time but that each Horizon has been accelerated over the past few decades. Do I have that right? If so, put me on record as contesting the notion that Time is the best way to differentiate sustaining vs breakthrough/disruptive innovation. In fact, I think it can even be harmful. For example, if a Pharma company wants to consider both biosimilars and digital opportunities in its portfolio, then it might put biosimilars in Horizon 1 (sustaining) and digital opportunities (potentially disruptive) in Horizon 3. Yet, the biosimilar development could be 5-7 years in development and digital development could be 2-3 years. How should we accommodate such a distinction in the Three Horizons model with Time on the x axis? Thanks in advance for your insights.

          • Brian. Apologies for the confusion. One of the things I tried to point out is that the Three Horizons was designed as a growth strategy framework, not an innovation framework. For us, the primary definition of the Horizons has to do with the nature of the revenue and earnings over time. You can have disruptive innovation in Horizon 1 if it has real bottom-line impact in whatever your short-term Horizon is defined to be. Equally, you could have incremental innovation be in Horizon 3 if the timing of impact on value creation is more consistent your defined long-term Horizon. I would agree with your contested point about innovation but draw the distinction with how horizons are meant to be used as a pipeline of growth instead. Does that make sense?

          • Thanks Mehrdad. Sort of. But I had to go back to your original Alchemy of Growth book to help. Now I understand better the nuance of growth strategy vs innovation. But I do think its a subtle nuance and lost on most people who have applied the Three Horizons framework. I have seen the Three Horizons model applied at many of our clients and the equivalence of breakthrough/disruptive innovation to Horizon 3 is common, if not exclusive. And therefore the equivalence of breakthrough/disruptive innovation to long time horizons…which is my concern. Its not the first time that a published framework has gotten misused to fit another purpose. Clayton Christensen would say the same about his theory of disruptive innovation. Thanks for engaging.

          • Fair point Steve. Makes total sense.

  9. Thanks for this Steve. Longtime reader, first time commenter. I can always trust your blog to get a fresh take on business and innovation.

  10. I totally agree! fair point and utterly important point indeed in order to avoid taking the wrong strategic decisions and specially strategic decisions while designing business models.

    Like you say horizon 3 may indeed be much closer than many people may think
    the 3 horizon is great is somehow a limiting view I use it but just as a mental framework to keep the techs in mind for a industry
    but It ought to be used with caution regarding the times and not believe horizon 3 for instance is far away when taking strategic decision regarding the future of your firm and so for instance in most industries horizon 3 ought not to be seen as something to be considered lightly and for the future as horizon 3 tech may well right now be part of your competitor’s strategy and so probably been used and tested right now as we speak by competitors to launch disruptive innovation new business models, very soon as like you mention delivery time is not a bog constraint anymore and the time factor has indeed changed..

    By the way I loved your commencement speech at ESADE ‘s graduation..

    I had to tell you.. I was invited to see it. and personally love it!
    Warm Regards,

    • What I meant in my above comment is that the 3 horizons is a useful tool (I use it but very carefully) a tool to be used as a mere framework but when taking important strategic decisions, decisions a firm’s future depends on .. Then the CEO or CIO of a firm ought to be utterly careful and not think horizon 3 is not down the corner ..
      Otherwise it could be a mistake that could put the firm at risk …

      So I personally use it merely as basic framework to keep in mind all the different tech.. but not as a strategic decision making tool

      As I do not consider horizon 3 for instance as something far away but rather the opposite, as like Mr. Steve Blanks points out.. delivery TIMES constraints are NO LONGER an issue and DISRUPTIVE INNOVATION coming from your direct competitors, new comers in your sector, new start-ups or spin-offs… may indeed very well be arriving soon to the sector and industry market you may be leading right now and much sooner than many might think.. or even just down the corner..?

      and it is a very high price to pay for the firm and CEO of a firm if strategic decision are taken wrongly.. a CEO cannot afford to be surprised by what others are doing or business models they are bringing to the market (competitors) …

      you don´t want to be that CEO having to say thad we didn’t do anything wrong.. but somehow we lost..suddenly bursting in tears..

      do you?

      if you know what I mean..

      The only fixed horizon thing I know of is this one:

      Competition is tough and disruptive innovation is here to stay not in horizon 3 but rather in horizon 1 ..

      Like we say in Spanish (well-know Spanish say)

      El que avisa no es traidor…

Leave a Reply

%d bloggers like this: