08 · Innovation strategy and portfolio

If the portfolio is the unit of management, its first designable property is its composition. The most widespread framework for reasoning about it is that of the three horizons (Baghai, Coley and White, The Alchemy of Growth, 1999): horizon 1 gathers the extension and defence of the current business (for an innovation portfolio: incremental projects on existing lines); horizon 2 the emerging opportunities under construction (projects that open new lines on adjacent capabilities); horizon 3 the options on the future (long-term explorations, immature technologies, markets that do not yet exist). The point of the framework is not the labels but the obligation it imposes: to make explicit the distribution of capital and attention among the three, because the implicit distribution, left to spontaneous dynamics, converges systematically on horizon 1 (which has the customers, the numbers and the urgency on its side) and starves the other two, reproducing at portfolio level the exploration-exploitation imbalance that chapter 16 will treat at organisational level. The right proportions are not universal: the often quoted 70-20-10 rule is a managerial anecdote, not an empirical result, and the correct split depends on the speed of the sector, the competitive position and the stage of the organisation. What is universal is the requirement of explicitness: the split must be a declared and revisited choice, not an accounting residue.

Alongside the horizons, classic balancing crosses risk and expected value: the resulting matrix distinguishes pearls (high value, high probability), oysters (high value, low probability: the bets that justify the portfolio), bread and butter (low value, high probability: useful, but a portfolio full of it is an annuity disguised as innovation) and white elephants (low value, low probability: candidates for stopping that survive thanks to sunk costs and attached sponsors). Here too the function of the matrix is diagnostic before it is prescriptive: it forces the shape of the portfolio to be seen, and the pathological shapes (all bread and butter, or a single large gamble that concentrates the risk) are visible only from above.

Entry and exit criteria are the immune system of the portfolio, and their practical asymmetry has to be corrected with the opposite asymmetry in the design: since opening is easy and closing is difficult, entry criteria must be demanding and exit criteria as automatic as possible.

Disciplined entry requires the candidate project to pass three examinations: absolute merit (the problem validated according to chapter 02, with a plausible route to a solution), relative merit (it yields more than the alternatives in the queue, including the alternative of reinforcing existing projects: every entry competes with everything), and compatibility (the necessary competencies exist or can be obtained; the balance of the portfolio allows it; the roadmap places it). The third examination is the one that organisations skip most often, and it is the cheapest to carry out.

Disciplined exit rests on the gates of chapter 07 (criteria declared beforehand, evidence examined by someone with no sunk costs) with the addition of the portfolio perspective: the question at the gate is not only "does this project deserve the next euro?" but "does this euro yield more here or in the best project in the queue?". It is the question that sunk costs and sponsors do not want to hear, and it is the reason why it has to be institutionalised: periodic portfolio reviews with the complete inventory in front of you, not project-by-project decisions in separate rooms.

The value of what is stopped. A project stopped with method does not return zero. It returns learning (the refuted hypothesis that will avoid the same error elsewhere), sometimes reusable technology, and above all a documented concept: the problem analysed, the mechanism verified as far as it was taken, the risks named, any intellectual property. Real options theory gives this residue its proper name: an option, the right without the obligation to reopen when circumstances change; and circumstances do change, because enabling technologies mature (the cost that made the concept uneconomic falls along its curve), markets move, regulations open spaces. The most cited historical case is Corning's Chemcor glass: developed in the 1960s, without a market at the time, archived with its documentation, and reopened more than forty years later when mobile telephony created demand for a thin, scratch-resistant glass; commercialised as Gorilla Glass, it has become one of the company's main products. The condition for the option to exist is archival discipline: the stopped concept must be documented as if it were to be reopened by others, the intellectual property maintained where the cost justifies it, and the inventory of concepts reviewed periodically in the light of what has matured in the meantime. A mature portfolio therefore has two inventories: the active projects and the waiting options, and the second is an asset, not a graveyard.

• R.G. Cooper, S.J. Edgett and E.J. Kleinschmidt, Portfolio Management for New Products (2nd edition, 2001): the empirical reference of the discipline, with the diagnosis of too many projects and the methods of balancing.
• M. Baghai, S. Coley and D. White, The Alchemy of Growth (1999): the original formulation of the three horizons.
• R. Phaal, C. Farrukh and D. Probert, "Technology Roadmapping: A Planning Framework for Evolution and Revolution" (Technological Forecasting and Social Change, 2004): the methodological frame of roadmapping in the Cambridge school.
• R.G. McGrath, "Falling Forward: Real Options Reasoning and Entrepreneurial Failure" (Academy of Management Review, 1999): real options reasoning applied to stopped projects, the theoretical foundation of the value of the stock of concepts.