Shaping futures

imageLong-range projections for business, like five year plans produced by central governments, are often more about justifying continued expenditures than they are about revealing the forward-looking thinking that is hidden in different parts of product teams. Rather than commitments to spend money forever, the best nuggets of good ideas from these teams should be periodically harvested and used to synthesize and weave together a long view, from which plans and estimates can then emerge and form into projects. This long-term perspective can help an organization recognize how and when they need to shift their investment priorities, organizational structures, and focus onto their next generation of new product for the future. The foundations required to produce such 'over the horizon' perspectives are well understood:

  • You must know where you are
  • You must know where you want to go
  • You need some idea of how you will get from where you are to where you want to go
  • You should understand how many resources you'll need, and what it will take to get them
  • You must believe that the journey is worthwhile
  • You must be able to measure your progress along the way, so mid-course corrections can be performed
  • You must make effective and timely decisions as the environment and your understanding of the landscape changes

As an example of the considerations required to form such a foundation, consider the product portfolio of a automotive manufacturer who must develop a long-term strategy for shifting their product lines so they primarily are made up of electric vehicles over a ten year period. Many companies are trying to successfully navigate through the decisions required to implement such a shift. Alan Mulally, Ford's CEO, described the factors that must be considered in making such transitions at the Tech Green conference in 2012. Alan indicated that there are three fundamental problems that the industry and communities around the world are all attempting to realize over the long term: economic development, energy independence, and environmental sustainability.

It may be impractical to simultaneously bet on multiple product configurations, each pursuing evolutionary or revolutionary steps in order to satisfy these needs; such a strategy would introduce uncompetitive compromises into each derivative product along the way, both due to the fragmentation of efforts, and the lack of focus on target benefits. For a while, Tesla's new product offerings appeared to approach perfection in consumer ratings, though in service, reliability has suffered, Evolution may offer appealing value propositions, while other new entrants are crashing and burning. Yet all have a cost that is out of the price range of most consumers, which diminishes the realistic number of units that development costs can be spread across. Despite extensive government subsidies, electric vehicle sales have been surprisingly weak most consumers are simply unwilling to pay a premium price for a commodity product, or said another way, customers don't change behaviors unless it is worth their time and effort.

Let's say the product development and startup costs for a new vehicle are about one billion dollars (close to Tesla Motors initial financing after government loans). Lets say you forecast a sales volume of 100,000 units for Tesla's product line over the first five years of production. That means the development costs which must be amortized over this sales volume amount to $10,000 per vehicle. This cost must be layered on top of the incremental manufacturing and component costs specific to this type of product (which already exceed other luxury vehicles, due to the high price of batteries). Roll up these costs together, layer on something for profit, and you've determined the price you must sell the product for in order to survive as a business. Yet no electric vehicle manufactured in the United States has yet been able to sell enough of these vehicles for it to be a good investment for their businesses; that's even with the government incentives (a form of subsidy) that are available to consumers.

When looking at revolutionary, rather than evolutionary products, our long-term view thus becomes increasingly foggy, as we try to push our crystal ball's reach farther and farther into the future. The basis of decisions which must inevitably be made along this journey will increasingly rely upon territory and environmental conditions which we can't control, history we can't alter, and often hidden, but corrosive technical debt which we've incurred in the past. The possible combinations necessary deal with these factors quickly explode, making outcomes difficult to realize, and costly to even evaluate. Chet Richards describes the implications this has for planning over long-term horizons:

A plan is an intention about how to get from where we are now to where we want to be in the future. It is an intention because although we may plan to accomplish certain things, whether we actually do, and whether they have the effects we want, depends on factors beyond our control: customers, competitors, governments, and acts of God, to name a few.

You will spend a lot of time trying to map out the future: “If this happens, I’ll do that. On the other hand, if that happens, I’ll do this. On the third hand . . .” You might want to add a fortuneteller or a prophet to your staff. At the same time you’ll discover that these models, which are a type of “decision tree,” quickly become complex. Instead of helping you deal with the confusion of the world, this approach can easily add to it.

Do you believe that you can create a plan for every eventuality? Each scenario will suggest ten more, and you will soon wrap yourself in a web of nodes and arrows from which you will never escape, and you will have to pay in money and time for the effort...

Problems can arise if you have a formal planning process with an attendant, enforcing bureaucracy. With all this investment, there is a tendency to take the output of the process (the plans) too seriously, locking you into the plan as the real world goes on its way.

Since one size generally will not fit all needs without totalitarian power, the alternative futures we must plan for must anticipate mixes of many different solutions, each with uncertainty varying from wild to mild; some options would require pivots to be performed over time in order to be successful, so the organization's agility in accomplishing these pivots must be taken into account.

In our automobile industry example, the vehicles the market wants are different within urban, suburban, and cross-country usage scenarios, and each of these call for different design solutions for this usage to be optimum. In this environment, businesses must chose to either craft a mix of solutions (one or more lines of products) that will appeal to a target customers (or with greater difficulty, attempt to convince current customers to replace what are using today). Assumptions must be made about important things that aren't yet knowable. Hedging thus emerges as a frequently employed strategy. For example, in his talk at the conference, Alan offered "None of us know which one is going to end up being the solution depending on the technology development. So, in the meantime we're going to pursue all of them and make that power of choice available to the consumer".

Alan's strategy is visible in Ford's continued investments in hybrids, while concurrently exploring various alternatives for electric vehicles. He clearly is letting the technologies mature which are critical to market requirements in the long run, without tying his future to any one particular direction for future products. This allows Ford to develop many of the necessary underlying technologies for energy management, while delaying a broader commitment to all-electric vehicles until the market is willing to pay for them. As of 2013, this strategy appears to be succeeding, at least so far. Of course, if Tesla learns to make their cars for half the price at scale (which we should all hope for, but not be surprised if they are unable to), Ford's strategy will be in trouble. Clean sheet designs such as Tesla's nearly always win out over incremental improvements since they can optimize their design more completely, and they have less encumbrances to circumvent that are grounded in prior decisions; but such new designs must also survive long enough, and build enough momentum in the meantime, for investors to keep investing - otherwise, they'll just be like many other advances that never captured enough market to survive.

It can be helpful to explore several alternative futures by adopting visualizations to depict the possible timing and interdependencies of activities and decisions as the future unfolds, so these alternatives can be considered from different frames of reference, and under different planning scenarios. The intent of such depictions - usually called by the generic term of roadmaps - is to help decision-makers better anticipate and position their organizations for emerging technology disruptions and evolving environments. Businesses and industries use these depictions to portray how changes will be introduced and features will be evolved over time. These depictions are described in more detail in the below table:

Types of roadmaps

Sponsors of Roadmaps

Scope and reach Know why

Leverage Know what

Areas of application Know how

Objective Will Do

Technology roadmaps

Science and Technology Providers

  • Cooperative investigations
  • Target applications
  • Enabling technologies
  • Trends & discontinuities
  • Technology elements and applications
  • Competitive applications and costs
  • Balance investment portfolios
  • Communicate technology maturation strategy

Strategy roadmaps

Industry and government regulators

  • Industry structure and positioning
  • Market drivers
  • Procurement standards
  • Discoveries
  • Inventions
  • Breakthroughs needed
  • Architectural patterns
  • Technology alternatives
  • Growth profiles
  • Involvement in standards development

Product roadmaps

Product and platform teams

  • Market structure and size
  • Feature differentiation
  • Competitive strategies
  • Product placement in market
  • Targeted evolution of capabilities over time
  • Integration points
  • Competitive positioning
  • Target costing
  • Market segmentation
  • Allocate channel capacity
  • Develop strategy to leverage intellectual property
  • Develop strategy to mitigate risks

Many examples of these roadmaps are publicly available, and can provide ideas about formats and structure. A representative sample includes depictions on Homeland Security, Wind Energy Management, Photonics, Mobile telephony, Enterprise architectureQuantum computing, Cloud ComputingUbiquitous computing, the Railway Technology Advisory Group, the British Health Care' Strategic Plan, the Automobile Australia's 2020 vision, or Microsoft's computational science forecast. Yet each can only anticipate the future based upon the rose-colored glasses of the present, and the increments which can be anticipated from that perspective.

The main risk of using such roadmaps is that they may isolate decision-makers from the actual conditions that are occurring on the ground. William Manchester described how this isolation can play out, in his compelling account of Winston Churchill during World War 1:

He described red-tabbed officers in warm, safe châteaux confidently moving pins on maps, forgetting that each pin represented a multitude of human beings whose outlook was very different from their own. “The hopes of decisive victory” grew “with every step away from the front line,” reaching “absolute conviction in the Intelligence Department.” The result—doomed offensives—troubled him more than any other aspect of the government’s war policy.

Given these opportunities for misplaced perceptions of control which these depictions can introduce, it often makes sense to preserve as many options as possible, so a particular organization is positioned to fit and adapt within a changing landscape; unfortunately, this often means products are portrayed as periodic 'releases', without providing much detail about what those releases would contain.

The challenge remains, as Mulally outlined above: to strategically position your business across several different planning horizons and across multiple dimensions, in a way that produces a set of meaningful strategies that inform those who must implement those strategies, while enabling you to survive long enough for the finances to work out along the way. The real threats in the long term arise from competitors that entrenched industries don't see coming, or would cannibalize their existing solutions. Success is thus more likely to be a result of increasing yields from existing investments, and increasing the pace of learning to learn, rather than swinging for the fences (no matter how appealing grand slams are). It is less painful to embrace a future that pretends that scarcity, competition, or ongoing obligations are not present, but it is not usually realistic to do so. Only when existing paths all lead to ruin, and significant change is recognized as essential to survival, should we pick up the torch and march for revolution.

Direction is essential to provide the scope and elaborate the foundation for these strategy discussions. A shared vision steered through a clear, actionable strategy provides a context for shaping and evaluating alternatives for these strategies, and provide a basis for enabling dialog and understanding of the commitments that will be required along the way. Yet sponsors, in search of predictability in an increasingly unstable world, may prefer to delegate such decisions, though this increases the likelihood of fragmentation, sub-optimization, and surprises later on. They may also withhold commitments for resources until they are convinced that their objectives are both realistic and worthwhile. With these limitations in mind, here is a roadmap for building useful product roadmaps, while managing the cognitive dissonance involved in creating them:

  1. Identify the stakeholders involved in producing and approving such roadmaps
  2. Use thought leaders representing these stakeholders to perform a CATWOE and SWOT exercise, with brutal honesty. To the extent you can, make the challenges concrete, with facts and data, so everyone recognizes what needs to change and by when
  3. Set the strategic and technological landscape for your product roadmaps, ideally by referencing a common source for this material, and agreeing on when significant changes in this landscape should be expected to occur
  4. Adopt a standard architecture for your roadmaps, so they will have a similar look and feel.
  5. Establish a level-setting set of assumptions for the roadmaps - for example, what should their scope and planning horizons be?
  6. Survey the decision-makers for the roadmapping effort, to establish the differences of the environment expected once roadmaps are in place. This can reveal how they are intended to be used.
  7. Explore the evolutionary changes roadmaps are expected to portray, such as providing new features, incorporating architectural shifts, or addressing emerging business situation
  8. Determine how decisions and alternatives will be depicted within the decision network, from both a technical and financial perspective
  9. Evaluate how to portray key attributes of elements on the roadmap; these include aspects such as levels of confidence, buy-in, business plan alignment, etc
  10. Generate a candidate set of potential depictions for the range of roadmaps that must be produced; address both the easy and hard end of the scale
  11. Walk through how these candidate depictions could be used with decision-makers, to think through how evolution would work, and validate the utility of the depiction
  12. Select the depictions with the highest stakeholder buy-in
  13. Determine how to scale, synchronize, and stabilize the selected example more broadly
  14. Establish a way to manage, authorize, and communicate changes to the collection of roadmaps as they will be evolved over time

One final caution: even with such a disciplined approach for roadmaps, growth is an emergent property of product interactions with environments and ecosystem. ​Charlie Anderson, the Architect of Borland Quattro Pro for Windows (the once-dominant, but now extinct company Borland), describes how software-intensive systems typically grow: "Software is like a plant that grows: you can’t predict its exact shape or how big it will grow; you can control its growth only to a limited degree. There are no rules for this kind of thing. It’s never been done before." Darryl Eaton, product manager at Yahoo, reinforces this perspective: "It's hard enough to get the next year's plan roughly right. Unconstrained, features grow like weeds. Over time, it will be necessary to thin out the weeds, so they don’t kill the roses." When gardening is a hobby, the effort to perform this thinning can still be satisfying. If you instead rely upon the food you are growing to get you through the next cold winter, it is probably better to be more selective up front. That's why this feature creep should be minimized up front by requiring linkage of roadmap elements to the business objectives which they are intended to support.