In an era defined by unprecedented volatility and rapid technological shifts, the imperative for organizations to anticipate and adapt to future landscapes has never been more acute. While scenario planning has long been championed as an indispensable tool for navigating radical uncertainty, its traditional methodologies, often protracted and resource-intensive, struggle to keep pace with today’s demand for agile insights. The conventional approach, which can span many months and consume significant managerial bandwidth and financial capital, is increasingly being challenged by a streamlined, user-centric paradigm augmented by the transformative capabilities of generative artificial intelligence. This evolution promises to deliver bespoke, impactful scenarios with unprecedented speed, enabling executives to proactively interrogate their strategic foundations and prepare for futures that defy linear extrapolation.
The global business environment, characterized by geopolitical instability, accelerating climate change impacts, and disruptive technological advancements, has amplified the need for robust strategic foresight. Organizations across sectors face complex challenges ranging from supply chain fragility and inflationary pressures to evolving consumer behaviors and talent shortages. In this context, a reliance on single-point forecasts or simple trend extrapolations is a perilous gamble. However, the deep dives required for comprehensive scenario planning – involving extensive research, workshops with diverse stakeholders, and meticulous narrative construction – historically presented a significant barrier. These processes often cost hundreds of thousands, if not millions, of dollars and could take six months or more to yield actionable intelligence, by which time the underlying assumptions might have already shifted. Furthermore, the trend towards delayering central management and strategy teams has left many managers stretched thin, making the allocation of substantial "person-days" to such endeavors increasingly difficult. The critical challenge, therefore, lies in forging a method that delivers timely, relevant insights without exhausting already scarce resources.
A new approach, pioneered by forward-thinking companies like Nordic fast-moving consumer goods giant Fazer and UK employee benefits provider Unum Ltd., addresses these bottlenecks head-on. Central to this methodology is framing scenarios around the specific individuals who will utilize them for strategic decision-making, ensuring immediate relevance. This user-centric lens focuses on challenging the "ghost scenario" – the often-unspoken, implicit assumptions that underpin an organization’s current investments and strategic direction. By explicitly identifying and interrogating these ingrained beliefs, the process grounds the scenarios in the immediate strategic concerns of the leadership team, making the insights directly applicable and impactful.
The "ghost scenario" is, in essence, the default future that senior leaders unconsciously anticipate and plan for. It comprises a constellation of assumptions about market growth, regulatory stability, technological adoption, competitive dynamics, and consumer preferences. For example, a company might implicitly assume sustained demand for a particular product category, stable raw material prices, or a gradual pace of digital transformation. When these assumptions remain unexamined, they can lead to strategic blind spots and an inability to recognize emerging threats or opportunities. The power of scenario planning, particularly when focused on the ghost scenario, is to bring these implicit assumptions to the surface, allowing them to be questioned, enriched, and ultimately reevaluated against a backdrop of plausible alternative futures. This ensures that the strategic conversation is anchored in what truly matters to the decision-makers here and now, preventing the common pitfall of developing abstract, irrelevant scenarios.
The process of surfacing these core assumptions typically involves in-depth interviews with senior management, reviews of existing strategic documents, board minutes, and internal communications. For Fazer, this entailed interviewing senior managers to compile a list of 15 widely held business assumptions, subsequently narrowed down to a critical few for focused exploration. Unum Ltd. similarly extracted assumptions from executive surveys and internal strategy documents. The deliberate selection of a small number of core assumptions for challenge is crucial; it avoids the "boiling the ocean" effect, where overly broad scope dilutes focus and delays actionable insights. By concentrating on specific, high-stakes assumptions, scenario planning teams can generate bespoke narratives that directly confront and enrich the existing mental models of key decision-makers, thereby enhancing strategic resilience.
The second transformative element in this accelerated approach is the judicious integration of generative AI (GenAI) tools, particularly large language models (LLMs). While traditional scenario planning relies heavily on human research and intuition, GenAI offers unparalleled capabilities for rapidly synthesizing vast quantities of information and generating creative outputs. LLMs can be deployed in two primary phases: initial research and scenario narrative generation. In the research phase, once core assumptions are identified, GenAI can swiftly analyze technological, economic, political, social, environmental, and legal (PESTEL) factors that could either reinforce or radically challenge these assumptions. This capability significantly reduces the manual effort and time typically required for comprehensive environmental scanning, enabling teams to explore a broader spectrum of developments that might spark imaginative thinking.
Following initial workshops where human experts outline potential future trajectories, GenAI truly shines in crafting compelling scenario narratives. These narratives, often told from the perspective of a future observer, describe how a particular future state might plausibly unfold. The key here is plausibility, not probability; scenarios are designed to stretch the imagination and consider what could happen, not to predict outcomes. By feeding workshop highlights and key building blocks into an LLM, a draft narrative can be generated in a fraction of the time it would take human writers. Iterative prompting allows for refinement, ensuring each scenario is distinct, vivid, and directly challenges the ghost scenario. Furthermore, these AI-generated narratives can be seamlessly integrated into video generation tools, creating engaging multimedia outputs that enhance accessibility and comprehension for senior management. Both Fazer and Unum reported efficiency gains of 40-50% in producing these outputs, alongside the added benefit of incorporating rich visual and auditory elements that would have been impractical without AI. This not only accelerates the process but also elevates the quality and impact of the final deliverables.
Fazer, a venerable Finnish consumer goods company with a diverse portfolio spanning confectionery, bakery, and plant-based products, exemplifies the benefits of this accelerated approach. Operating across multiple markets and industry categories, Fazer grappled with myriad unpredictable developments and sought to invigorate its annual strategy discussions with fresh perspectives. Initial attempts at scenario planning with self-focused "what-if" questions, such as "double-digit volume drop," proved insufficient for truly stretching strategic thinking. The adoption of the Oxford approach, centered on challenging the ghost scenario, refocused their efforts on broader contextual shifts. By interviewing senior managers, Fazer identified 15 core assumptions, ultimately narrowing them down to focus on critical supply-side dependencies, particularly related to raw material availability and climate change impacts.
A two-day workshop involving 20 senior representatives from across Fazer’s value chain – including supply chain, procurement, R&D, marketing, finance, and communications – further refined these assumptions. The insights from this diverse group, combined with rapid exploration facilitated by LLM tools, allowed Fazer to develop a half-dozen plausible futures within two months. A particularly impactful scenario explored a future where climate change led to significant negative impacts on staple crops, challenging the company’s implicit assumption that emissions reduction alone would mitigate all climate-related risks. This exercise forced Fazer to confront the "inconvenient truth" of drastic, direct impacts on its suppliers and operations, prompting a reevaluation of fundamental business model elements. The scenarios, integrated into Fazer’s strategic assessment processes, generated compelling insights that enabled executives to question long-held beliefs and develop strategies that differed significantly from current plans.
Unum Ltd., the UK subsidiary of a major US employee benefits provider, faced its own set of mounting uncertainties as it approached its next strategic planning cycle in early 2025. Following significant growth post-COVID-19, a sluggish UK economy, slowing employment, and pressures on the public healthcare system created a complex operating environment, further complicated by the unknown implications of AI on the labor market. With a tight deadline to deliver scenarios and a preliminary 2030 strategy draft to the board by May, Unum’s strategy team embraced the accelerated, AI-enabled approach. Their initial step was to identify the ghost scenario, revealing an executive team largely extrapolating current trends with only moderate disruption from AI. This led to 15 focal topics, from UK healthcare provision to the future of employment in an AI-driven corporate sector.
Unum leveraged LLMs to quickly identify credible data sources for each topic, streamlining the research process ahead of a full-day scenario development workshop. This workshop convened internal subject matter experts alongside external consultants, including a public affairs specialist and a former competitor executive, ensuring diverse perspectives. Participants engaged in an exercise using cards representing opposing hypotheses (e.g., "AI drives creation of high-skilled jobs" vs. "AI contributes to widespread job displacement") to map plausible sequences of events leading to future states by 2040. This process converged on two critical uncertainties for Unum: the extent of AI’s influence on human capital demand and the level of government intervention in the labor market, both profoundly impacting an employee benefits provider. Post-workshop, GenAI crafted vivid narratives for the ghost scenario and three alternatives, describing life in the UK in 2040. Crucially, three-minute AI-generated videos illustrating these narratives proved to be a powerful engagement tool for the executive team, fostering meaningful discussion and a shared understanding of complex future possibilities far more effectively than text alone. Unum’s experience underscored AI’s potential to accelerate complex strategic processes, freeing up valuable human capital for higher-value strategic thinking.
The successful application of GenAI in scenario planning hinges on a crucial philosophical shift: viewing LLMs not as objective sources of factual information to automate research, but as sophisticated collaborators in imaginative foresight. This "AI-in-the-loop" perspective places human reasoning at the core, with AI serving as an ideation partner whose resources extend beyond organizational boundaries and mitigate groupthink. Effective prompting, such as using personas (e.g., "You are an expert scenario planner looking to challenge…") and iterative refinement, becomes a continuous dialogue rather than a one-off query.
However, the outputs of LLM queries demand rigorous evaluation for plausibility, relevance, and usefulness. While scenarios are not predictions, they must resonate as believable futures. Human biases or knowledge gaps can still affect this assessment. Both Fazer and Unum addressed this by establishing external expert review processes or meticulously validating AI-cited sources. Fazer engaged external experts to review scenario descriptions, while Unum treated LLM outputs as "raw" material, scrutinizing cited documents for credibility. This underscores the importance of a critical, human-led oversight mechanism, ensuring that AI-generated insights are robust and genuinely inform strategic decision-making.
In conclusion, the escalating complexity and uncertainty of the global business landscape necessitate a paradigm shift in strategic foresight. The traditional, arduous nature of scenario planning is no longer fit for purpose in an environment demanding rapid adaptation and agile decision-making. By integrating a laser focus on user-centric problem framing, specifically challenging the implicit "ghost scenario" that guides current strategy, and leveraging generative AI to accelerate research, narrative creation, and multimedia production, organizations can unlock faster, more relevant, and deeply engaging strategic insights. This innovative approach not only significantly reduces the time and resource burden but also empowers leaders to proactively explore alternative futures, question fundamental assumptions, and cultivate a heightened state of strategic readiness, thereby enhancing their capacity to thrive amidst radical uncertainty.
