Foresight in Action: Corporate Foresight Case Studies
What companies with real foresight functions actually do, what the evidence says it is worth, and the four ways of building a foresight system.
The most famous story in corporate foresight is that Shell's scenario planners saw the 1973 oil shock coming and the company profited handsomely from being ready. It is told in almost every introduction to the field.
The people who ran the team say it is not what happened. Angela Wilkinson and Roland Kupers, both former members of Shell's scenario team, wrote in Harvard Business Review in 2013: "We have no solid examples of Shell's having anticipated future developments better than other companies, the mythology around anticipation of the 1970s oil crises notwithstanding." Their claim is narrower and more useful. Shell was perhaps faster than others at catching on once change began. One team member's description of the job: "We're not forecasters. We're your personal trainers."
The historian Jenny Andersson goes further in the Business History Review, arguing that Pierre Wack's account of the 1972 scenarios was mythologised after the fact, and that the scenarios worked as instruments for framing arguments about OPEC and resource limits rather than as prediction.
This matters because the case for foresight is usually made with that story, and the story does not support the claim. The evidence that does exist is more modest, more recent, and worth stating accurately.
What the evidence shows
The best quantitative study remains Rohrbeck and Kum's longitudinal analysis in Technological Forecasting and Social Change in 2018. They measured foresight maturity across a set of firms in 2008 and looked at performance through to 2015.
Firms they classed as vigilant, meaning their foresight capability matched the turbulence of their environment, averaged 16% profitability against an industry average of 12%. That is the source of the widely-quoted "33% more profitable" figure, and it is a relative premium rather than an absolute one. Over the same period vigilant firms grew market capitalisation by 75%, and the authors describe this as 200% additional growth over the comparison group. Firms two or more maturity levels below what their environment demanded ran 44% lower profitability.
Two caveats belong with those numbers. The profitability analysis covers 70 firms and the market capitalisation analysis only 42, in a self-selected sample of mostly European multinationals. And no study since has replicated the link at comparable scale. This is the strongest evidence the field has, and it is thin. Anyone quoting it without the sample size is overselling.
Rohrbeck's earlier maturity model, built from 19 case studies and 107 interviews, is the more durable contribution. It assesses five dimensions: information usage, method sophistication, people and networks, organisation, and culture. The point of the model is that foresight capability is not a single thing you have or lack. A firm can have excellent scanning and no route from scanning to a decision, which is the most common failure we see.
How many companies actually do this
Jan Oliver Schwarz and Katharina Kleine surveyed 400 senior executives at Forbes Global 2000 companies for The State of Corporate Foresight, published in 2025. Their findings are more sobering than the case-study literature suggests.
42% have a dedicated foresight unit with its own budget and staff. Another 36% embed foresight in existing functions. Only 4% have no formal structure at all. So the practice is widespread among very large firms, and it has grown: a comparable 2020 survey found around 34% with a dedicated unit.
The interesting numbers are about horizon and use. 86% plan on a three to ten year horizon, and only around 1% look beyond ten years. Foresight is being used to extend the planning cycle rather than to think about structurally different futures. And while 81% use it for goal-setting, only 43% use it for long-term strategic planning and 40% for exploratory work.
That gap between having a foresight function and using it for anything exploratory is the real story in corporate foresight, and it is where most of the value is left on the table.
The cases that hold up
Shell is still the deepest practice, and it is still running. The scenarios team published the 2026 Energy Security Scenarios in January 2026, offering three futures named Archipelagos, Surge and Horizon, following the 2025 edition and the first Energy Security Scenarios in 2023. Sixty years of continuous scenario practice, beginning with Ted Newland's Long-Term Studies group in 1965 and reshaped by Pierre Wack from the early 1970s, is a longer institutional memory than any other company has. The value is in the continuity and in what it does to how executives think, and not in a forecasting record.
AXA runs the best-evidenced corporate foresight function outside the energy sector. AXA Group Foresight is a named unit with a named head, Olivier Desbiey, and it has published an annual foresight report for eight consecutive years, most recently The Atlas of New Futures in 2026. Separately, the AXA Future Risks Report reached its twelfth edition in October 2025, built on 3,600 experts across 57 countries and 23,000 members of the public in 18 countries, run with Ipsos. Climate change has ranked first for five consecutive years. Published method, repeatable instrument, and a time series long enough to show movement. That combination is rarer than it should be.
The Foresight Academy is the most structurally interesting case, and the least discussed. Initiated by Audi and built around Rupert Hofmann's Trend Receiver method, it is a shared foresight platform with partners including Porsche, Allianz, Swiss Re, Siemens, SAP, E.ON, Beiersdorf, Ferrero, BSH, Lidl, Kaufland, Hornbach and Ströer. Competitors and companies from unrelated sectors pooling foresight infrastructure is a different model from building your own unit, and for organisations too small to staff a function it is the more realistic one.
PepsiCo is worth including with its date attached. From 2008 the company ran a programme applying foresight methods to its research portfolio, creating two new departments aimed at breakthrough innovation and using seven methods including implications wheels, inductive scenarios, weak signal scanning and participatory futures. It is documented properly, by Farrington, Henson and Crews in Research-Technology Management in 2012. It is also a practitioner self-report about a programme that began eighteen years ago, and we found no evidence the function still exists in that form.
Four ways to build a foresight system
The question we are asked most often is not whether foresight works but where to put it. Pihlajamaa and colleagues studied eleven large companies through 43 interviews for a 2025 paper in Technological Forecasting and Social Change, and identified four ways of organising a corporate foresight system. Each has a characteristic strength and a characteristic failure.
Function-driven. Foresight lives inside an existing function, usually strategy, innovation or R&D. Cheapest to start, and the work inherits that function's credibility and its blind spots. The failure mode is that the foresight only ever addresses questions that function already cares about.
An independent foresight unit. A dedicated team with its own budget and mandate, which is what 42% of Global 2000 firms now have. It buys independence and depth of method. Its failure mode is distance: a unit that produces excellent work nobody in the business asked for, and which is the first thing cut in a bad year.
Executive team-led. Foresight owned directly by the leadership group, often with light external support. It solves the distance problem completely, since the people who commission it are the people who decide. It is fragile in a different way, because it depends on individuals and rarely survives a change of chief executive.
Platform-based. Shared infrastructure across business units, or as the Foresight Academy shows, across companies. It spreads cost and widens the range of signals. It needs governance, agreed processes and someone accountable for interpretation, or it becomes a shared library nobody uses.
None of these is correct in the abstract. The right one depends on where decisions actually get made in your organisation, which is a question worth answering before choosing a structure.
Five things that make organisations futures ready
Every organisation's route differs, but the same five capabilities recur in the ones that make foresight stick.
A futures culture. Permission to discuss futures that are inconvenient to the current strategy. Without it, the scanning finds only what the strategy already assumes.
Foresight capability. Enough method in enough hands that futures work does not depend on one enthusiast. When that person changes role, an organisation without distributed capability loses the practice entirely.
Scenario-based strategy. Not a scenario set delivered as a report, but strategy that has been tested against several futures and knows which of its commitments only work in one of them.
Continuous futures intelligence. Scanning on a rhythm, feeding a named decision, with indicators worth checking annually. A scan with no owner and no decision behind it is a hobby, and it gets cut.
Anticipatory leadership. Leaders who can hold several futures in mind while still deciding. This is the scarcest of the five and the one that determines whether the other four produce anything.
A note on AI
The World Economic Forum and the OECD surveyed 167 foresight practitioners across 55 countries for AI in Strategic Foresight, published in November 2025. Practitioners value AI mainly for the labour-intensive parts: scanning, sorting, summarising. Their reservations are about quality, hallucination, and limited inductive reasoning.
The structural point in that report is the one worth keeping. AI draws on existing knowledge and struggles with forward-looking perspectives. It is constitutionally backward-looking, which is precisely the wrong bias for this work. It is very good at the half of foresight that is retrieval, and it does not do the half that is judgement.
Where this comes from
The framing of ambidexterity used throughout the corporate foresight literature is older than most citations suggest. Robert Duncan first used "ambidextrous" of organisations in 1976. James March supplied the underlying distinction between exploration and exploitation in 1991, arguing that the basic problem for any organisation is to exploit enough to stay viable while exploring enough to have a future. Tushman and O'Reilly formalised organisational ambidexterity in 1996. The Invincible Company by Alexander Osterwalder, Yves Pigneur, Alan Smith and Frederic Etiemble, published in 2020, is where most practitioners meet the idea: constant reinvention, competing on business models over products, and running exploration alongside exploitation.
We work as futurists-in-residence, providing sparring, facilitation, research and insight for organisations building a futures culture. If you are deciding where foresight should sit in your organisation, or you have a foresight function that produces good work nobody acts on, that is the problem we spend most of our time on.
Sources
Wilkinson, A. and Kupers, R., "Living in the Futures", Harvard Business Review (May 2013)
Andersson, J., "Ghost in a Shell: The Scenario Tool and the World Making of Royal Dutch Shell", Business History Review 94(4) (2020)
Shell, The 2026 Energy Security Scenarios (January 2026)
Rohrbeck, R. and Kum, M. E., "Corporate foresight and its impact on firm performance: A longitudinal analysis", Technological Forecasting and Social Change 129 (2018)
Rohrbeck, R., Corporate Foresight: Towards a Maturity Model for the Future Orientation of a Firm, Physica-Verlag (2011)
Schwarz, J. O. and Kleine, K., The State of Corporate Foresight: Global Study on the Application of Corporate Foresight, Bavarian Foresight-Institute and Nuremberg Institute for Market Decisions (2025)
Pihlajamaa, M., Komonen, P., Huuhanmäki, J. and Kurki, S., "Four archetypes of organizing corporate foresight at the intermediate maturity stage", Technological Forecasting and Social Change 215 (2025)
Farrington, T., Henson, K. and Crews, C., "Research Foresights: The Use of Strategic Foresight Methods for Ideation and Portfolio Management", Research-Technology Management 55(2) (2012)
AXA, Future Risks Report, twelfth edition (October 2025), and AXA Group Foresight annual reports
World Economic Forum with the OECD, AI in Strategic Foresight: Reshaping Anticipatory Governance (November 2025)
Backler, Iny and Turner, "What Companies that Excel at Strategic Foresight Do Differently", Harvard Business Review (January 2026)
Osterwalder, A., Pigneur, Y., Smith, A. and Etiemble, F., The Invincible Company, Wiley (2020); March, J., "Exploration and Exploitation in Organizational Learning" (1991); O'Reilly, C. and Tushman, M., "Organizational Ambidexterity" (2013)
