OECD member countries spent roughly $1.9 trillion on research and development in 2023, about 2.7% of their combined GDP, with United States business R&D alone near $729 billion (OECD, 2023). The returns on that spending are usually positive, but they vary widely from one company to the next, and a large part of that variation comes down to a management factor rather than a scientific one: how well the R&D portfolio is aligned with the corporate strategy it is meant to serve.
That alignment is harder to achieve than it sounds. The R&D portfolio is the CTO's responsibility, while corporate strategy belongs to the CEO acting with the Board, and the two are connected by a valuation process that few people outside the R&D function fully understand. Most portfolios are assembled through some combination of weighted scoring models and financial calculations, and when the CTO presents the result, the Board is asked to trust a method it cannot easily audit. The question we are asked to help answer is a practical one: how do you build an R&D portfolio that is defensible in the language the Board actually uses?
Financial valuation quietly biases the portfolio toward the near term
Two families of method dominate practice. The first relies on evaluative judgment through checklists, scorecards, stage-gate criteria, and weighted scoring. The second values projects in monetary terms through net present value, internal rate of return, payback, and real options. Both are useful, and most companies run some blend of the two.
The blend creates a recurring organizational tension. CFOs tend to prefer financial valuations, while CTOs maintain that a credible monetary figure often cannot be produced for early-stage work, where the eventual application, and therefore the revenue, is still years away. When the financial number wins by default, the portfolio drifts toward later-stage, easier-to-monetise projects. That drift may be the right strategic choice, but more often it is an accident of the valuation method rather than a decision anyone made on purpose.
Reframe the question from monetary value to business impact
Our approach reframes the conversation. Instead of asking what a project is worth in dollars before the science is settled, we ask what business outcome it is ultimately meant to move, and by when. We classify every project on three dimensions: the area of business impact it targets, the relative magnitude of that impact, and the approximate time horizon over which it is expected to land.
A Business Impact has to be more specific than "increase revenue" or "lower cost." It is the highest-level answer to how, precisely, R&D can contribute to a top-level corporate goal, and the cleanest source for it is the set of KPIs a given industry already lives by. A retailer may frame impact as sales per square foot rather than total revenue; an automotive OEM as warranty cost per vehicle; a miner as cost per tonne or worker exposure hours; an oil and gas producer as carbon intensity. We recommend no more than five impact areas, chosen so that revenue, cost, and a third category such as sustainability, safety, or regulatory standing are all represented.
"Insisting on a single financial number for every R&D project asserts a precision the analysis rarely has. The honest comparison is the business outcome each project is meant to move, and by when."
A common scale makes different-in-kind projects comparable
The obvious objection is that comparing a market-share project against a fuel-efficiency project is comparing apples to oranges. The answer is a single ordinal scale, with three or four intuitive points such as Low, Moderate, High, and Very High, applied across every impact area. The design work sits in defining what each point means for each area: a KPI that typically moves only in single-digit percentages might earn a "High" at a 5-to-10% gain, while one that is easier to inflect might need 20-to-30% for the same rating. Historical examples from the company or its competitors anchor those ranges and discourage inflation, and "Very High" is reserved for the exceptional result that may have been achieved only once in a decade, perhaps not even in-house.
Time horizon does similar work. The simplest framework is three horizons, where the short term runs roughly 12 to 18 months, the medium term out to about four years, and the long term beyond that, with the durations adjusted to the industry. The horizon also serves as a proxy for technical uncertainty, mapping cleanly onto technology-readiness levels: near-term projects should rest on mature technology, while long-horizon projects are where exploration and genuine uncertainty belong.
Decide the portfolio you want before you ask for proposals
The most consequential move comes before any proposal is written. The chosen impact areas and time horizons form a simple matrix, and the desired allocation across it, the Target Portfolio, should be agreed between the CTO and CEO, with reference to Board priorities, and then published alongside the call for proposals. This is a reversal of the usual order, in which proposals are solicited first and a portfolio is reverse-engineered from whatever arrives.
Setting the target first does three things. It cascades corporate strategy into high-level R&D goals before any effort is spent. It lifts the quality of submissions by telling proposers what the business actually wants, rather than inviting a scramble that then has to be rationalised. And it removes the late-stage manipulation that occurs when a portfolio has to be bent into shape after the fact. The method extends cleanly: a balanced scorecard can add technology-roadmap compliance, company readiness to adopt a given technology, and the breadth of a technology's application as further scoring dimensions, without losing the business-impact spine.
The portfolio becomes legible, and harder to cut by accident
Tying each project to a business impact and a horizon produces a portfolio that can be read at the Board level. It answers questions the Board actually asks, such as how much of a $50 million R&D budget supports greenhouse-gas reduction, by reading a subtotal off the matrix. When strategy shifts, when economic headwinds turn the focus from revenue growth to cost containment, the portfolio response is a deliberate reallocation between impact areas rather than a panicked round of cuts.
That last point matters most in exactly the conditions where R&D is most exposed. A portfolio expressed in the Board's own terms is far harder to cut capriciously, because every project is already attached to an outcome the Board has said it wants, and the trade-offs of cutting it are visible rather than buried in a scoring spreadsheet.
The lessons are clear: value R&D in the language of business outcomes rather than forcing a monetary figure the early science cannot support; agree the portfolio you want before you invite proposals, not after; and keep the whole thing legible enough that a Board can audit it. None of this removes judgment from R&D decisions, and it is not meant to. It gives that judgment a common frame of reference, so that the conversation between the R&D function, the C-suite, and the Board is about the same portfolio, valued the same way.