AI, automation, and the quiet risk of ‘vanishing strategy’ – How restructurings are eroding institutional logic and what organisations must do to protect it (interview with Janice MacLennan, founder of Nmblr)

  1. Thank you for speaking with me Janice. From your vantage point as a senior strategy consultant working closely with large pharmaceutical organisations, what do you mean by vanishing strategy, and why is this risk often invisible to leadership until it is too late?

“Most large pharmaceutical organisations are well supplied with decks, roadmaps, and governance forums. What vanishes is something more fundamental: the institutional logic that connects those artefacts into a coherent whole. It is the shared understanding of why the organisation is pursuing certain choices, which trade-offs it is consciously making, and what it has learned from prior cycles of success and failure. When that logic erodes, decisions may still be made and initiatives may still move forward, but they are no longer anchored in a common strategic memory.

“The risk is often invisible to leadership precisely because the organisation continues to produce outputs. Programs advance, milestones are met, and dashboards remain reassuringly green. In stable conditions, this creates the illusion of strategic health. The problem only becomes apparent when a shock occurs – be it a pipeline failure, a regulatory shift, or a competitive disruption. At that point, the organisation struggles to explain its own priorities, align quickly across functions, or adapt its course with confidence. By the time this fragility is exposed, the underlying strategy has not just weakened; it has effectively disappeared.”

  1. Over the past decade, restructurings, delayering, and cost-optimization programs have become almost continuous in big pharma. How do repeated organisational resets alter institutional logic and strategic memory over time, even when individual restructuring decisions appear rational in isolation?

“Over time, repeated restructurings fundamentally disrupt the continuity of institutional logic, even when each individual intervention appears rational on its own terms. Delayering, cost optimisation, and operating model resets tend to fragment the chains of accountability and narrative that explain why certain strategic choices were made in the first place. As people move on, roles are recombined, and mandates are redefined, the organisation steadily loses its ability to remember what was tried, what failed, and what unintended second-order effects emerged as a result.

“The cumulative effect is that strategy shifts from being something the organisation carries forward and refines to something it must repeatedly reconstruct from scratch. Each new leadership team or restructuring cycle feels compelled to ‘re-decide’ fundamental questions, not because the answers are wrong, but because the reasoning behind them has been lost. Over time, this erodes strategic coherence and increases the risk of oscillation – where organisations cycle through familiar moves without learning, mistaking motion for progress while institutional memory quietly disappears.”

  1. Much of strategic capability lives in tacit forms—judgment, historical context, pattern recognition. Which types of institutional knowledge are most vulnerable to loss during automation and workforce reduction, and why are these the hardest to replace once they disappear?

“The institutional knowledge most vulnerable to loss is not technical expertise or formally documented process. It is the tacit knowledge that sits between functions and across time. In large pharmaceutical organisations, this often includes the unwritten rationale behind portfolio and brand choices, the pattern recognition that links experiences across therapeutic areas and markets, and a practical understanding of the organisation’s true risk appetite – particularly the conditions under which that appetite expands or contracts. It also encompasses the ability to interpret weak signals, distinguishing early indicators of material change from background noise, as well as the informal influence map that determines who truly shapes decisions and where resistance is likely to emerge.

“These forms of knowledge are hardest to replace precisely because they are carried by roles that are frequently targeted during automation and workforce reduction. Efforts to remove cost and complexity tend to focus on routinised and mid-complexity work, yet these positions often act as connective tissue within the organisation. They translate strategy across functions, retain historical context, and accumulate judgment through repeated exposure to edge cases rather than through formal authority. Once this layer is removed, the knowledge does not migrate cleanly into systems or successors. It simply disappears, leaving organisations with data and processes, but without the interpretive capacity needed to use them well.”

  1. AI is often presented as a solution to complexity, speed, and efficiency challenges. In your experience, does AI primarily compensate for strategic capability gaps created by restructuring—or does it risk accelerating the disappearance of human strategic reasoning itself?

“In practice, AI does both – it compensates for certain strategic capability gaps while simultaneously creating a new class of risk if it is misapplied. On the positive side, AI is genuinely powerful at scanning volumes of information far beyond human capacity, surfacing patterns and anomalies, accelerating first drafts, and reducing friction in knowledge retrieval. In organisations thinned by restructuring, these capabilities can meaningfully offset lost capacity and help teams move faster with fewer resources.

“The deeper concern, however, is not AI itself but how its outputs are used. AI produces answers that are often highly plausible and delivered with speed, which makes them easy to accept – especially in environments with fewer people, less time, and diminished institutional context to interrogate them. When output becomes a substitute for reasoning, the risk is that teams stop practising strategic thinking altogether. Over time, that leads to atrophy of strategic muscle: fewer challenges to assumptions, weaker sense-making, and declining ability to frame the right questions. In that scenario, AI does not merely compensate for gaps created by restructuring; it accelerates the disappearance of human strategic reasoning by masking its absence with convincing outputs.”

  1. Digital transformation tends to privilege measurable outputs. How does the increasing reliance on dashboards, KPIs, and algorithmic recommendations reshape decision-making cultures—and where does this shift begin to crowd out deeper strategic thinking?

“Digital transformation inevitably reshapes decision-making culture by privileging what can be measured, compared, and optimised. As dashboards, KPIs, and algorithmic recommendations become more central, attention shifts toward variables that are easily quantified and updated frequently. This subtly pulls organisations away from weak signals, outliers, and long-cycle strategic bets – exactly the areas where early insight and judgment matter most, and where clean metrics are least available.

“Over time, teams adapt their behaviour accordingly. They learn to manage to the dashboard, optimising performance against visible indicators rather than exploring uncertain or ambiguous terrain. In this environment, optimisation begins to crowd out exploration, and leaders may conflate the speed and precision of reporting with the depth and robustness of underlying thinking. The result is a culture that appears analytically rigorous, yet is increasingly brittle – highly responsive to what is measurable, but less capable of sensing, interpreting, and acting on what truly shapes long-term strategic outcomes.”

  1. Many executives now inherit organisations already stripped of layers and memory. What responsibilities do today’s leaders have to actively protect institutional logic, rather than assuming strategy will “emerge” from tools, processes, or AI systems?

“In organisations already stripped of layers and institutional memory, leaders cannot assume that strategy will simply emerge from tools, processes, or AI systems. Institutional logic does not self-organise; it must be actively stewarded. Protecting it is a core leadership responsibility, not a by-product of efficient operations or sophisticated analytics.

“At a practical level, this stewardship has three dimensions. First, coherence: leaders must ensure that decisions made across functions and time horizons fit together into a comprehensible whole, rather than optimising locally at the expense of the system. Second, memory: they must deliberately preserve the rationale behind past choices – what was tried, what was learned, and why certain paths were taken or abandoned. Third, interpretation: leaders need to legitimise time spent making sense of signals, debating implications, and exercising judgment, rather than treating reporting speed as a proxy for strategic quality.

“Ultimately, this means recognising that leadership is not just about running the operating system efficiently, but about stewarding the organisation’s thinking system so that strategy remains cumulative, intelligible, and resilient.”

  1. Based on your consulting work and your perspective as founder and CEO of Nmblr, what concrete organisational practices—structures, roles, or cultural interventions—can realistically prevent strategic erosion while still embracing AI and automation?

“I would focus on a small set of interventions that are realistic, durable, and compatible with AI and automation – rather than idealised transformations that organisations struggle to sustain.

“First, organisations need to make decisions portable. Every material decision should leave behind a lightweight but explicit record: what was decided, the underlying rationale and assumptions, the confidence level at the time, what evidence would change the decision, and who owns it with a defined review date. This creates continuity of logic even as people, structures, and tools change.

“Second, leaders must create explicit space for strategic thinking. Performance reporting and operational reviews are necessary, but they are not substitutes for interpretation. Strategic sense-making requires protected time and forums of its own. Without that separation, the urgent will always crowd out the important, and strategy will be reduced to commentary on last month’s numbers.

“Third, institutional memory needs formal ownership and incentives. Someone – or a small group – must be accountable for maintaining decision logic over time: connecting past choices to current debates and ensuring that learning accumulates rather than resets with each cycle. When no one is responsible for memory, it predictably disappears.

“Finally, AI should be deployed with governance that protects human judgment rather than bypasses it. For consequential decisions, teams should be required to surface assumptions, limitations, and counterarguments, with clear human sign-off on the final call. From my perspective at Nmblr, this is where structured shared workspaces add real value – not as ‘another tool’, but as a place where strategic logic lives in one view. Assumptions, evidence, trade-offs, and narrative remain visible, revisit-able, and shared, helping teams to keep moving as one, rather than just producing outputs in parallel.”

  1. Finally, looking 5–10 years ahead, what distinguishes organisations that will retain strategic depth in an AI-rich future from those that may become operationally efficient but strategically hollow—and what early warning signs should boards and regulators watch for?

“Looking ahead five to ten years, the organisations that retain genuine strategic depth in an AI-rich environment will be distinguished less by the sophistication of their tools than by the strength of their underlying thinking systems. Strategically deep organisations maintain a living map of their assumptions, risks, and narrative choices. They can explain not only what they decided, but why they decided it, what trade-offs were accepted, and under what conditions those decisions would be revisited or reversed. Strategy remains cumulative, interpretable, and adaptive rather than episodic.

“By contrast, organisations that become strategically hollow will push automation and efficiency to their logical extreme, without making a corresponding investment in human judgment. What remains is often a thin layer of executives sitting atop powerful analytical and AI systems, but with weakened connective tissue across functions and time. Decisions may be fast and well-instrumented, yet poorly integrated, fragile under stress, and difficult to defend when conditions change.

“As for the early warning signs that boards and regulators should take seriously, these include the same strategic questions resurfacing repeatedly without resolution; frequent decision reversals with little or no learning captured; persistent inconsistency in strategic narratives across functions; and a growing governance load that adds process without increasing clarity. Perhaps most concerning is an increasing dependence on a narrow set of metrics, models, or external vendors to justify major strategic decisions, accompanied by limited documented debate or exploration of alternative scenarios. Taken together, these signals point to organisations that are becoming operationally efficient, but progressively less capable of thinking strategically over time.”

Interviewer: Prof. Atanas G. Atanasov

Janice MacLennan is CEO of Nmblr, a collaborative strategy platform for Biopharma and MedTech. With 30+ years’ experience guiding companies through complex commercial challenges, she helps teams align, adapt, and deliver innovations effectively. Passionate about humanising strategy, through Nmblr, Janice empowers organisations to make better decisions and bring breakthrough therapies to market.