For a long time, the McKinsey Three Horizons model has been a systematic approach to look at the strategy of innovation in larger companies. It laid out three horizons, or time lines, that innovation comes in. Many enterprises have used the model to stay competitive in emerging markets. But we do so with a lower impact today than we used to. Even though incumbents in markets continue to grow quickly. Until we stop.
Defining The Three Horizons Model
The model came out in 2000, the same year the dot com bubble burst. Think about this: when Baghai, Coley, and White wrote The Alchemy of Growth, it was a disruptive model. And selling a good or service online was actually considered innovative. Like Pets.com.
The goal was to execute business models while researching and building new capabilities. And many of those took years to develop. Today, we can probably launch a new product in months if not weeks. That might be going up or down market with an existing product, or tinkering with the whole new disruptive approach for our existing models.
Before going into what’s so broken about the model these days, let’s look at what the horizons are:
- Horizon 1 includes ideas that provide a continuous stream of innovation to an organization’s business model. Anything continuous is usually building on core capabilities and usually includes investments that pay back quickly so provide an ROI in the short-term.
- Horizon 2 ideas extend business models into adjacent markets for core capabilities or extends selling targets. These are often referred to as adjacencies these days.
- Horizon 3 ideas involve bringing on new capabilities and entirely new streams of business in order to disrupt industries. This might mean using disruptive technology or models to protect market share as well.
Thee Horizons seems great on paper. Just define a percentage of the product budget for each and a time horizon unique to our industries and organizations. I mean, now we have a spreadsheet, so innovation just kinda’ solves itself, right? As organizations scale, they become less nimble. But now we have an innovation strategy. We have a presentation that identifies the amount of management overhead, the tools, and goals for each horizon.
Issues With Three Horizons
No offense but these days the competitors for organizations who have established themselves as incumbents in markets are already going up and down market. Competitors are also going vertical, building specialized solutions to reach beachheads, or micro markets within customer segments. They don’t need to think about compliance or politics like getting buy-in. They just do. Customers can buy from them without a lot of friction. Sales are often direct, so they don’t have to be concerned about confusing the channels when they release. In short, they’re faster, they’re smaller, and they’re more nimble. They’re probably also cheaper.
Over time, they may mature. But many won’t need to, or will mature in ways away from us. But many will see our core products as goldmines and likely be specifically looking to disrupt us. And each individually likely can’t do so quickly. But combined they can chip away at the customer base, fragment the market, and erode trust. They make it seem we aren’t innovating fast enough. But they aren’t constrained by the same limitations we are.
A Level Playing Field With Startups
The playing field has been leveled in many ways. Since the rise of public clouds like Amazon Web Services and the Google cloud platforms, competitors no longer need to go get a colocation facility and build clusters of servers as was necessary when the 3 Horizons model was developed. There are a lot more developers today, including some who work for some of our larger customers. And there are sites today that allow for social coding, so many who don’t know how to perform certain tasks can get there quickly. Even technology like machine learning can be taken out of a box and put into a product quickly so models can start maturing early.
Development efforts can also be outsourced in about as many ways as there are companies these days. Outsourcing can be done to small, targeted companies who are highly specialized with specific technology. Outsourcing can be to larger organizations who can cover more. And with websites that allow for specific outsourcing it can be as low as a half-hour task where just a tiny bit of help is needed. If concerned about intellectual property, don’t get confined with legacy tools but instead create a whole new toolchain.
Given a solid understanding of a given domain knowledge, a new product that provides a subset of the features of a large, incumbent, can be launched in days or weeks. Sure those types of products would target a subset of customers but with a highly targeted product, incumbents are left with prescribed workflows appearing as though they’re innovative approaches. These kinds of products can be launched cheaply and angel or venture capital is cheap and abundant, if it’s even needed.
The Mythical Man Month is as true today as it was in the 1970s. When IBM developed the groundbreaking S/360 mainframe at a cost that exceeded the cost of the Manhattan Project, they identified that throwing more humans at a project usually actually reduces the productivity. So a competitor with 4 people can be perceived externally as actually accomplishing more than an incumbent with 100 or 1,000 times that many employees.
Beyond Development
Customer acquisition is also easier today than ever before. Humans are just more interconnected in general. Online communities can create network effects quickly. Payment processing and tax lookup are available as a service, making it simple to transact when going to market. Successful startups have trained sellers to hop between products making it easier than ever to actually reach customers. Startup exists have given us thought leaders, podcasters, authors, and speakers who lay our sales, product, marketing, and development roadmaps bare for anyone to pick up and compete with us.
Once a startup has acquired customers, it’s easier than ever to communicate with them. Support tools are available as a service and internal chat can be deployed in minutes in apps. FAQs, support desks, support sites, call systems are all available in simple, monthly SaaS models. In short, it’s simple to launch a fledgling support organization. And it’s not uncommon to talk to a founder when we initiate a support request early-on in a product release – sometimes finding a product gap addressed within days if not hours.
And the result has been that the asymmetrical competition created means that Horizon 2 and 3 have become more acquisition-oriented for many organizations. Many of these startups get acquired by incumbents or larger organizations looking to move into adjacencies. This then drives the private equity industry to invest more, infusing startups and actually acts against incumbents. It turns out that once a disruptor has been acquired they end up being asked to operate at the speed of the acquirer and thus become less innovative.
Today, the more mature an organization if not an industry, the more vulnerable we are for disruption. We don’t want to accept this reality but it is a reality.
What’s Next
Ultimately, Three Horizons needs an update, but it’s not dead. Truly disruptive ideas where we’re creating categories or disrupting industries can still be considered Horizon 3 and new features of existing products or services can be considered Horizon 1. Or we can turn the model on its head for the right ideas.
But it’s clear that anyone still leveraging Horizon 3 should be thinking about how we should, dare we say, innovate that model. So what are some factors we can consider in iterations we can make to the model?
- Horizon 0. One of the best ways to block disrupters is to stay modern with technology. This might be considered defensive. There’s rarely anything externally facing so these are usually technical roadmap types of initiatives. They make us nimble, improve the time to market and allow our technical teams to take ownership of their aspect of innovation in an organization as well.
- Involve User Research. Product teams are more and more laser focused on defining what goes into products. UX Research teams have stepped up to help more clearly define what goes into products and how new features manifest themselves in products. Those User Research teams can also help us map out what a minimum valuable product looks like. Going a step further organizations can also extend the findings of user research to everyone, democratizing findings and bringing others into the process.
- Accept minimum valuable products as iterative in nature. Releasing a minimum viable product is, well, minimally viable to everyone. But understanding the customer problem still allows a disruptive product that can be iteratively finished as we get feedback from customers. Early adopters can then become our biggest advocates in the market, becoming personas, amplifying messages on social networks, providing case studies, and helping shape the product.
- Use Design Sprints for rapid prototyping. There are a few books on design sprints at this point, but rapid prototyping tools, sketching tools, and rapid application development tools allow organizations to ship a fast and easy app. Leaders might have to hold the antenna with one hand and point their other hand in the sky while standing on one foot in order for the prototype to work – but they’re useful and help us identify what we’re looking for in an investment and maybe if we’re lucky, what to expect
- Take prototypes to market quickly. Google launched Google Apps as a beta product and it sat there for a long, long time. And in that time they picked up market share. Branding the tool as beta meant that there weren’t as high an expectation of completeness and scale as there might otherwise be. It was a huge challenge to get support at first, but we all knew what we were getting into. Consider getting prototypes in front of customers quickly and soliciting feedback.
- Give ownership to the innovators. Run each initiative as a separate cost center. Each initiative doesn’t have to be run as an independent company, but each leader should understand what it means to run a business. Give them access to finance and Human Resources and make a horizon to profitability explicit, the same way we might do if we were investing in a company as a private equity firm or as venture capitalists. The innovators believe in the ideas and whether in a separate organization or at our organizations, can be inspired to create great products.
- Don’t forget that these are early-stage startups. Many of us have forgotten what it’s like to work in a startup. As our organizations have grown and become compliant, we naturally stop identifying with those early-stage problems, like building a marketing discipline. What works for mature products often causes new initiatives to fail faster than they might if given a little room to breathe. Don’t forget that these products don’t have to be included in every aspect of compliance, that there will be aspects like reporting that are necessary for larger products but can crush leaders that need to focus more time on growing a business than informing the rest of the organization about the business. We want these early innovators to be given the room to grow into a generation of leaders, not to make them gun-shy about taking new ideas to market.
- Let acquisitions run semi-independently. We acquire companies for a reason. But we often forget about that reason and expect them to act just like the rest of the organization. Do we really need every product to be compliant with all the same compliance or governance initiatives in a portfolio? Can each run using a separate framework? Leaders at the top of the organization need to see performance and especially over time. But we can establish an office to help take reporting and other burdens off younger products and work with leaders to integrate systems slowly. In the beginning we should focus on ramping up customers with initiatives like product integrations and plan to reduce costs by integrating backend systems later. Allowing for a flexible operating paradigm means meeting with each existing leader and the new incoming leadership for a product to review the strategy and tactics for integration into the organization and lay out a time frame to expect these. And importantly, to allow for flexibility.
- Find ways to compensate sellers for promoting the cause. Most successful organizations have a large, mature sales and marketing organization. Sellers by nature follow their incentive plans. New products can disrupt our existing models and to keep sellers from acting competitively against new initiatives provide compensation for doing so early. Don’t forget that a good sales organization is a great channel and expanding portfolios to allow for innovative concepts to flourish both make sellers feel better about the future of the organization and when done properly allow new products to expand quicker. Just be deliberate about these allocations and make sure the new product can expand at the pace and that sellers are well informed about the cohort of customers we might be looking to direct to each product in the portfolio. After all, customers usually have a finite budget in the first year and can’t buy everything we make up front.
- Use a Marketplace to get prototype features or integrations shipped quickly. The buzzword of becoming a “platform” has been on the tip of everyone’s lips since the early days of the Internet when Netscape managed to command an over 80% market share in the browser market. Many an organization built a marketplace to become a platform and may even be taking a cut of profits from other organizations in those marketplaces. But those Marketplaces can be leveraged to release prototypes as “beta” or “lab” projects quickly and with limited liability or support.
- Use the gig economy the same way competitors are. This doesn’t mean getting a corporate Uber account. It means using targeted consulting to get various tasks done quickly. Sites like Upwork give access to an infinite number of software developers. Sure, they’re not approved by IT and they’ll be working on their own systems, but carving up projects smartly so a couple different developers are building back-end micro services likely can’t connect the dots for who they’re working for and even if they posted their work product to GitHub the exposure would be minimal (and might just benefit an organization anyway).
- Use customer advocacy programs to incentivize feedback and get our most ardent fans involved early in product cycles. That might mean using an existing advocacy portal to get customers to try out new products or write reviews of products. Advocacy programs can also get much more private feedback at more stealthy stages and maybe even bring in new employees who believe in the mission. After all, innovators should be missionaries, not mercenaries.
- Establish incentive programs to include the community and foster external innovation. Think of NASA giving a prize to SpaceX, DARPA prizes, Netflix algorithm challenges, Apple bug bounties, and many others. This model can be outsourced but we should start out doing it ourselves in order to best understand what outsourcing attributes to look for. This might be an entirely new product or looking for ways to integrate existing products. And there’s a possibility that we end up hiring or outsourcing to those involved early.
- Copy new competitors back. Many incumbents created categories. New competitors copy the incumbent and target market segments. Rather than discount what these new competitors are doing, embrace the ideas and take the narrative back. Heck, even give them credit publicly. Quickly understand the problems, validate the model from market entrants, do better and do it faster to disrupt the disruptors.
Ultimately, incumbents have more capital and can innovate faster than our competition. But will we? It turns out that innovation isn’t as simple as three bullet points. We just don’t have years for Horizon 3 ideas to emerge and mature as maybe we once did. Or maybe it’s that the timeline is now the same for Horizon 2 and Horizon 3 ideas.
Don’t fall in the trap of a traditional Three Horizons model. Instead, use it as a general outline. Consider the value of building products, getting to market quickly, and comparably the true cost of acquisitions. We see competitors emerge quickly and innovate quickly. We then try to adopt their strategies but expect those same strategies to work in a larger organization when many just can’t. Taking the model and applying some hard learned lessons over the past few years prepares us to get out there and build a better world.