Asymmetric learning: what, why, when...
Better strategy early means better execution later...
I realized that I hadn’t written a simple guide to "asymmetric learning" as a strategy for achieving competitive advantage in pharmaceutical portfolios, so this is an attempt to do just that…
The central argument is that traditional pharmaceutical development often operates under a "prediction paradigm" that limits the potential for outperformance. We all know what that means: from the earliest stages of development, we’re trying to predict all kinds of things, sometimes 10 to 15 years out - probabilities of success, market size, competitor behaviour and more. Unfortunately, we depend on doing the same things as our competitors, and the same way.
Thanks for reading Asymmetric Learning - Pharmaceutical Innovation ! Subscribe for free to receive new posts and support my work.
Asymmetric learning proposes a shift towards a "learning by doing" approach, emphasizing flexibility, continuous information gathering, and stage-appropriate decision-making to maximise the potential of drug assets. Along the way, you’ll see that I advocate for strategies such as parallel development teams and focusing on target opportunity profiles rather than premature product specialisation.
Main Themes and Important Ideas
The Goal of Asymmetric Learning:
Learning as the Differentiating Factor:
Moving Beyond the Prediction Paradigm:
Embracing "Learning by Doing":
Stage-Appropriate Decision Making:
The Power of Parallel Development Teams:
Focusing on the Target Opportunity Profile (TOP):
Planning to Learn vs. Executing Against Prediction:
Conclusion:
The concept of asymmetric learning presents a compelling alternative to traditional pharmaceutical development strategies. By prioritising learning, flexibility, and stage-appropriate decision-making, companies can potentially unlock hidden opportunities and achieve a competitive edge. The proposed shifts in mindset and practices, such as employing parallel teams and focusing on target opportunity profiles, offer concrete ways to move away from a potentially limiting "prediction paradigm" and embrace a more adaptive and ultimately more successful approach to bringing innovative medicines to market.
Pharmaceutical companies can strategically adopt asymmetric learning to gain a distinct competitive advantage by fundamentally shifting their approach from a prediction-based model to one that prioritises learning and adaptability throughout the drug development lifecycle. This involves several key changes in mindset and practice:
? Embrace a Learning-Centric Approach Over Prediction: Instead of primarily focusing on predicting outcomes early in development, companies should prioritise the continuous acquisition of knowledge. The core idea of asymmetric learning is that the difference between a preclinical molecule and a launched product is the learning that has occurred. Therefore, improving the learning process itself becomes a source of competitive advantage.
? Adopt Stage-Appropriate Decision Making: Companies should avoid "premature specialisation" by refraining from making all critical decisions about a product's indication and market positioning too early. Asymmetric learning advocates for making decisions that are appropriate for each stage of development, allowing for flexibility and adaptation based on emerging data.
? Plan to Learn, Not Just to Prove: Early-phase planning should focus on identifying and addressing key uncertainties rather than solely aiming to prove a pre-determined hypothesis. This involves actively seeking information about the drug's potential, including where it works, its side effects, and the target population. Teams should spend more time listening and evaluating what is known and unknown about their path to market.
? Explore Multiple Potential Paths and Indications: Rather than locking into a single development path early on, companies should explore a range of potential "destinations" and be prepared to pivot based on what they learn. This allows them to uncover unexpected benefits or identify the most promising applications for their assets.
? Foster Internal Competition and Diverse Learning: A powerful strategy is to assign the same drug asset to multiple independent teams within the organisation. While there might be a time and energy cost, different teams will inevitably learn in different ways, and one is likely to achieve superior results. This allows the company to identify more effective development and execution strategies.
? Focus on Gathering Superior Information: The competitive advantage gained through asymmetric learning stems from the ability to collect and utilise better information than competitors. This superior information informs better decisions at each stage of development.
? Shift from a Product-First to a Target Opportunity-First Approach: Instead of starting with a molecule and trying to find its place in the market, companies should consider starting with an understanding of unmet needs and market opportunities (the "target opportunity profile") and then identifying or developing drugs that address those needs. This approach significantly reduces the risk associated with making early predictions about a molecule's market potential. The "three-legged stool" model – approvability, commercial attractiveness, and feasibility of development – should ideally guide the selection and progression of assets.
? Recognise and Leverage the Empirical Nature of Drug Development: Given the complexity of biological systems, a significant amount of drug discovery and development is serendipitous. Asymmetric learning embraces this by prioritising learning through experimentation and observation over purely rational, predictive approaches.
? Continuously Re-evaluate Assumptions and Decisions: Asymmetric learning encourages a dynamic approach where companies regularly revisit their understanding of an asset and the market. The key questions become: "where are we now, what do we know now that we didn't know before, and what decision would we take today if we were about to embark upon it?".
By strategically implementing these principles, pharmaceutical companies can move away from a "rules of average" mentality and build a competitive advantage based on superior learning, adaptability, and the ability to identify and pursue the most promising opportunities for their drug portfolios. This shift towards a growth mindset for individual assets and development teams is crucial for outperforming competitors.
How does asymmetry relate to desired market outperformance?
Asymmetry is fundamentally linked to desired market outperformance because it represents a deliberate departure from average or symmetrical performance, with the specific goal of achieving superior results. The asymmetry is your desire to outperform. In markets where performance tends to be symmetrical, meaning companies achieve similar outcomes in areas like share of voice or the relationship between investment and drug approvals, asymmetry is the key to breaking away from the pack.
Here's how asymmetry, particularly through the lens of asymmetric learning, relates to desired market outperformance:
? Moving Beyond Average Performance: Symmetrical performance indicates that a company is performing in line with industry averages. If a pharmaceutical company aims to achieve market outperformance, it needs to adopt strategies that create asymmetry, allowing it to exceed these averages. Asymmetric learning provides such a strategy by encouraging companies to think and act differently from their competitors.
? Focusing on Superior Learning: The core of asymmetric learning lies in the idea that the difference between a drug in preclinical development and a successful launched product is the knowledge gained throughout the development process. By adopting a better approach to learning – one that is more active, adaptive, and exploratory – a company can gain a significant advantage over competitors who rely on traditional, less effective learning methods. This superior learning leads to superior information, which in turn informs better decisions and ultimately, better outcomes in the market.
? Challenging the "Rules of Average": Many companies operate based on industry norms and past practices, essentially following the "rules of average". Asymmetric learning encourages companies to challenge these norms and develop unique approaches tailored to their specific assets and market opportunities. This departure from the average is essential for achieving asymmetrical – i.e., outperforming – results.
? The Pharmaceutical Innovation Index: The concept of asymmetry is central to measuring pharmaceutical innovation, as highlighted by the Pharmaceutical Innovation Index, which assesses companies that outperform their direct competitorsin drug development and execution. This underscores the direct link between asymmetric approaches and achieving superior competitive results.
? Growth Mindset and Competition: Fostering internal competition by assigning the same asset to multiple teams can drive a growth mindset and lead to superior learning within the organisation. This internal asymmetry in learning approaches can ultimately translate to better market performance for the company as a whole.
In essence, asymmetry, as embodied by the principles of asymmetric learning, is not just about being different for the sake of it; it's about intentionally adopting strategies that enable a pharmaceutical company to learn faster, more effectively, and in ways that uncover opportunities and mitigate risks better than their competitors, thereby leading to desired market outperformance.
Why is a symmetrical learning approach often insufficient for pharma?
A symmetrical learning approach is often insufficient for pharma because it tends to perpetuate average performance and prevents companies from achieving a distinct competitive advantage. The video source highlights that in many markets, including aspects of the pharmaceutical industry, performance can be symmetrical, meaning companies achieve similar results in areas like investment leading to approvals. However, if a company desires to outperform its competitors, this symmetrical or average approach to learning is inadequate.
Here are the key reasons why a symmetrical learning approach falls short in the pharmaceutical industry:
? It Leads to Average Outcomes: A symmetrical approach implies that a company is learning and operating in a way that is similar to its competitors. By following the "rules of average" and adopting standard practices, companies are likely to achieve average results, rather than the desired outperformance. Asymmetric learning, in contrast, is specifically about having a "desire to outperform".
? Reliance on Past Practices: Symmetrical learning often means that companies are "learning the way that it used to be done" and applying "the lessons of the past". This can hinder innovation and prevent the adoption of more effective learning strategies that could lead to a competitive edge.
? Following Industry Norms: Companies engaging in symmetrical learning may adhere to industry averages, such as conducting certain activities two years before launch because "on average" companies do so. This fails to recognise that to outperform, a company needs to deviate from these averages and adopt more strategic and potentially unconventional approaches.
? Missed Opportunities for Superior Information: A symmetrical learning approach often doesn't prioritise the collection of "superior information". Asymmetric learning emphasises that a key differentiator between teams working on the same asset is their ability to learn better and find out more relevant information that informs their decisions.
? Reinforces a Prediction Paradigm: Traditional, symmetrical learning tends to be tied to a "prediction paradigm" where the emphasis is on being right from the outset. This can lead to "premature specialisation" where decisions about a drug's use are made too early, limiting the learning about other potential applications and hindering the ability to pivot based on new information. Asymmetric learning, conversely, prioritises "planning to learn, not just to prove".
? Limited Adaptability: Symmetrical learning may not adequately foster the adaptability required in the complex and uncertain environment of drug development. The regulatory landscape, commercial environment, and even the understanding of the drug's mechanism of action can evolve. A rigid, symmetrical approach may not be prepared to adjust to new information and changing circumstances, whereas asymmetric learning explicitly aims to cultivate this flexibility.
? Failure to Leverage Internal Diversity: The video source suggests that even within the same organisation, different teams working on the same asset will learn differently, with one likely to be more successful. A symmetrical approach might not exploit this potential for diverse learning and internal competition, which asymmetric learning actively encourages.
In essence, a symmetrical learning approach in pharma keeps companies on a level playing field, achieving similar outcomes to their competitors. To gain a distinct competitive advantage and outperform the market, pharmaceutical companies need to embrace asymmetric learning, which focuses on actively seeking superior knowledge, adapting to new information, exploring diverse possibilities, and challenging the status quo.
How does asymmetric learning aim to benefit pharma portfolios?
Asymmetric learning aims to benefit pharma portfolios by designing in competitive advantage across the entire collection of drug assets. Instead of treating each drug with a standard, "average" approach to development and learning, asymmetric learning encourages a more nuanced and adaptive strategy that ultimately seeks to outperform competitors at a portfolio level.
Here are some specific ways asymmetric learning benefits pharma portfolios:
? Enhanced Identification of Promising Assets: By encouraging the exploration of multiple potential paths and indications for each drug early in development, asymmetric learning helps identify the most promising assets within the portfolio and their optimal applications. This avoids "premature specialisation" and the risk of focusing too narrowly on a potentially less successful path.
? More Confident Late-Stage Decisions: The emphasis on "planning to learn" in the early phases means that by the time assets reach phase 3, more information has been gathered about their potential, leading to more confident decisions regarding endpoints, value propositions, and target populations across the portfolio. This reduces the risk of costly late-stage failures.
? Optimised Resource Allocation: By fostering a deeper understanding of each asset's potential and the associated uncertainties, asymmetric learning can inform better decisions about resource allocation within the portfolio. Resources can be directed towards the assets with the highest potential for success, based on superior information gathered through the learning process.
? Increased Overall Success Rate: By improving the learning process at the individual asset level, asymmetric learning aims to increase the likelihood of successful development and launch for a greater proportion of the portfolio. This is achieved by being more adaptable, pivoting when necessary, and making stage-appropriate decisions based on accumulated knowledge.
? Competitive Differentiation: When a company adopts asymmetric learning across its portfolio, it moves away from the "rules of average" that its competitors might be following. This creates a distinct competitive advantage at the portfolio level, as the company becomes better at identifying opportunities, mitigating risks, and ultimately bringing more successful products to market.
? Fostering a Learning Culture: The principles of asymmetric learning encourage a growth mindset and a culture of continuous learning within the organisation. This benefits the entire portfolio as teams become more adept at gathering and utilising information, leading to more informed decision-making across all development programs.
In essence, asymmetric learning aims to transform a pharma portfolio from a collection of independently developed assets to a strategically managed group where each asset benefits from a superior learning process, ultimately leading to a higher likelihood of overall portfolio success and a distinct competitive advantage in the market.
Why might multiple teams on one asset benefit learning?
Having multiple teams working on the same pharmaceutical asset can significantly benefit learning because different teams will inevitably learn differently, leading to a broader and deeper understanding of the asset's potential and limitations. This aligns with the principles of asymmetric learning, which emphasises the value of diverse approaches to gain superior information and outperform competitors.
Here's a breakdown of why this approach can be beneficial, drawing from the sources:
? Diverse Learning Approaches: As mentioned above, if the same asset is given to two different teams within an organisation, it's highly likely that one team will perform better than the other not necessarily because they are smarter or more talented, but because they learned better. This highlights that different teams will naturally adopt different strategies for investigating the asset, asking different questions, and interpreting data in unique ways. This diversity in learning approaches can uncover a wider range of information and insights than a single team might achieve.
? Superior Information Gathering: The more successful team will likely be better at "finding out about opportunity, about planning, about the collection of superior information". By having multiple teams explore the asset simultaneously, the organisation increases the chances that at least one team will excel at gathering crucial information that might be missed by another. This collection of superior informationis a key tenet of asymmetric learning.
? Identification of Unforeseen Opportunities and Risks: Different teams, with their varied perspectives and learning methodologies, may identify different potential applications, patient populations, or even unexpected side effects that a single team might overlook due to their specific focus or biases. This broader exploration can reveal hidden opportunities or potential pitfalls associated with the asset.
? Internal Competition and Growth Mindset: Source points out that the element of competition between the teams can be beneficial. This internal competition puts the teams into a growth mindset, encouraging them to think differently about their interactions and strategies. The desire to outperform the other team can drive more creative and rigorous learning processes.
? Validation and Cross-Verification of Findings: When multiple teams are investigating the same asset, their findings can be compared and cross-verified. If different teams independently arrive at similar conclusions, this strengthens the confidence in those findings. Conversely, discrepancies in their results can highlight areas that require further investigation and a deeper understanding.
? Faster Learning Overall: While there might be a perceived cost in terms of time and energy by having multiple teams, the accelerated and more comprehensive learning that can occur may ultimately lead to faster and more informed decision-making in the long run. The combined knowledge generated by multiple teams could surpass what a single team could achieve in the same timeframe.
In the context of asymmetric learning, having multiple teams working on the same asset can be seen as an active strategy to introduce asymmetry within the learning process itself. By deliberately fostering diverse approaches and internal competition, a pharmaceutical company can move away from a more uniform, and potentially less effective, learning model, increasing its chances of gaining a competitive advantage through superior knowledge acquisition.
What is the key difference between pre-clinical and launched assets?
The key difference between a pre-clinical asset and a launched asset in the pharmaceutical industry is the knowledge acquired about the molecule during its journey through the development pipeline.
As highlighted in the thought experiment above, the molecule itself remains the same from the pre-clinical stage through to launch. The transformation from a pre-clinical entity to a marketable product is fundamentally a learning process.
This learning encompasses various critical aspects, including:
? Where it works: Identifying the therapeutic indications and patient populations for which the asset is effective.
? Side effects: Understanding the safety profile and potential adverse reactions associated with the asset.
? Effect size: Determining the magnitude of the therapeutic benefit the asset provides.
? Target population: Defining the specific patient subgroups that are most likely to benefit from the treatment.
? Pricing: Establishing an appropriate market value for the asset.
? Mechanism of action (to varying degrees): While the exact mechanism isn't always fully understood even post-launch, significant learning occurs regarding how the drug exerts its effects.
In essence, a pre-clinical asset is a molecule with potential, but with significant unknowns. A launched asset, on the other hand, is a molecule for which a substantial amount of information has been gathered through rigorous research and clinical trials, allowing for its safe and effective use in a defined patient population. This accumulation of knowledge is the defining factor that distinguishes the two.
Why might firewalled teams studying one asset perform differently?
Firewalled teams studying the same pharmaceutical asset are likely to perform differently primarily because of variations in their learning processes and approaches. While they are working on the same molecule, the way each team investigates, interprets information, and makes decisions will inevitably diverge, leading to different levels of success.
Here are some key reasons why firewalled teams might exhibit different performance:
? Distinct Learning Methodologies: As highlighted in source and our previous discussion, different teams will naturally employ different strategies for learning about the asset. One team might prioritise certain experiments or data points over others, leading to a different understanding of the asset's potential and limitations. As stated in source, the difference in performance is likely "not because one is smarter or more talented or they know more about something than the other, they're going to be better because they learned better".
? Variations in Information Gathering and Interpretation: A more successful team will be better at "finding out about opportunity, about planning, about the collection of superior information". Firewalled teams, operating independently, will collect and interpret information in their own way. One team might be more effective at identifying crucial data, asking the right questions, or drawing insightful conclusions from the available information, while another might miss key details or make different assumptions. The ability to gather and utilise "superior information" is a crucial differentiator.
? Impact of Internal Competition and Mindset: The element of competition between firewalled teams can be a significant driver of performance. This internal competition puts teams into a growth mindset, encouraging them to think more creatively and diligently about their approach. The desire to outperform the other team can lead to more rigorous analysis, innovative problem-solving, and a greater focus on identifying and addressing knowledge gaps. Teams in a growth mindset are also more likely to focus on "learning by doing" and collecting "superior information".
? Differences in Planning to Learn: Source and emphasise the importance of "planning to learn" rather than just executing against initial predictions. One firewalled team might be more proactive in identifying key unknowns and designing experiments or analyses specifically to address those uncertainties. They might spend more time "listening" and "evaluating what's known and unknown", leading to a more focused and effective learning journey compared to a team that takes a more rigid or less inquisitive approach.
In essence, the act of firewalled teams working independently on the same asset introduces asymmetry into the learning process. Each team's unique approach, influenced by their methodologies, information processing, competitive spirit, and focus on learning, will lead to different rates and depths of understanding about the asset, ultimately resulting in variations in their perceived performance and potential outcomes.
Explain the significance of "planning to learn" in early drug phases.
The significance of "planning to learn" in early drug development phases is paramount because these stages are characterised by significant uncertainty and assumption risk. Rather than simply executing a pre-determined plan based on initial predictions, "planning to learn" emphasises a proactive and adaptive approach focused on gathering crucial information to inform future decisions.
Here's why this is so important:
? Addressing High Assumption Risk: In the early phases (pre-clinical and phase one), there are numerous assumptions about the science, clinical potential, regulatory path, and commercial landscape of a drug. Trying to predict everything with certainty at this stage is unreliable. "Planning to learn" acknowledges this uncertainty and prioritises the collection of data to validate or challenge these initial assumptions.
? Avoiding Premature Specialisation: A key danger in early development is "premature specialisation" – deciding too early on a single indication or target for the drug. This limits the potential to discover other beneficial uses or patient populations. "Planning to learn" encourages the exploration of a "range of potential destinations and a range of paths". By actively seeking information about different possibilities, companies can avoid prematurely narrowing their focus and potentially missing out on more successful applications.
? Enabling Pivoting and Adaptation: The information gathered through a deliberate learning process can reveal unexpected findings or challenges. "Planning to learn" prepares companies to be flexible and "ready to pivot". If early data suggests the initial hypothesis is incorrect or that another avenue is more promising, a learning-oriented approach allows for adjustments in the development strategy. This contrasts with a prediction-based model, which can be resistant to change even in the face of contradictory evidence.
? Leading to More Confident Later-Stage Decisions: When early phases are focused on learning, the decisions made at later stages, such as the design of phase three trials, are much more informed and therefore more confident. By exploring endpoints and building a value proposition based on accumulated knowledge, the risk of late-stage failure is reduced [4, our conversation history]. As stated above, if you avoid premature specialisation, "your phase three is a lot more confidently assigned".
? Facilitating the Collection of Superior Information: "Planning to learn" involves a mindset of actively seeking and evaluating information about what is known and unknown. This contrasts with simply running experiments to prove a pre-conceived notion. Teams that are "planning to learn" spend more time "listening" and "really evaluating what's known and unknown" about their path to market. This focus on collecting "superior information" is central to achieving asymmetric learning and outperforming competitors.
? Fostering a Growth Mindset: "Planning to learn" aligns with a growth mindset, where the focus is on acquiring knowledge and adapting strategies based on new findings. This is more beneficial than a mindset focused solely on being "right" from the outset, which can lead to overlooking important information or being resistant to changing course.
In essence, "planning to learn" in early drug development shifts the focus from simply executing a plan to actively seeking knowledge and understanding. This iterative process of learning, adapting, and making stage-appropriate decisions based on evidence is crucial for navigating the inherent uncertainties of drug development and ultimately enhancing the chances of portfolio success through asymmetric learning [4, 5, our conversation history].
How does focusing on opportunity profiles inform product selection?
Focusing on opportunity profiles significantly informs product selection by shifting the initial focus from a specific molecule or asset to the desired outcome in the market. This approach allows for a more strategic and less risky selection of products to develop.
Here's how this works, drawing on the sources:
? Starts with the Desired Outcome: Instead of beginning with a molecule and trying to find a market for it, focusing on opportunity profiles starts by defining the unmet need or the desired commercial opportunity. This involves understanding what kind of product would be approvable, commercially attractive, and achievable, as represented by the "three-legged stool" of target opportunity.
? Reduces Prediction Risk: The traditional approach of starting with a molecule and then trying to predict its market segment, pricing, and other commercial factors involves significant "risk". By contrast, beginning with a well-defined opportunity profile allows companies to ask more pertinent questions about their existing molecules and assets.
? Enables Better Questioning of Existing Assets: When you have a clear understanding of the desired opportunity profile, you can then assess which of your existing molecules or assets has the potential to meet those criteria. This allows for a more informed selection process, rather than simply pushing forward with whatever molecules happen to be in the pipeline.
? Avoids Premature Specialisation at the Molecular Level: Starting with a molecule can lead to "premature specialization" where the drug's potential is narrowed down too early based on initial assumptions. By focusing on the opportunity profile first, you maintain flexibility and can consider a broader range of potential applications for different molecules.
? Aligns Development with Market Needs: This approach inherently increases the likelihood of developing a product that the market will actually want. By understanding the target opportunity profile – including factors like commercial attractiveness – you are more likely to select a product that addresses a real market need and has a viable path to success.
? Facilitates "Planning to Learn": Focusing on opportunity profiles aligns with the principle of "planning to learn". Instead of trying to force a molecule into a predefined market, you are using the understanding of the desired opportunity to guide your learning process as you evaluate different product options [5, our conversation history]. This involves asking questions about which assets have the potential to fit the profile and then designing learning activities to gather the necessary information to make informed selection decisions.
In essence, focusing on opportunity profiles provides a strategic framework for product selection. It reverses the traditional approach of molecule-first thinking and instead prioritises understanding the desired market outcome. This enables companies to make more informed decisions about which assets to invest in, reducing risk and increasing the likelihood of developing commercially successful products.
What is the significance of stage-appropriate decision-making?
The significance of stage-appropriate decision-making is central to the concept of asymmetric learning and offers a more effective approach to pharmaceutical development compared to traditional methods.
Here's a breakdown of its importance:
? Reduces Risk in Early Phases: Making all critical decisions at the beginning of Phase 1, before sufficient learning has occurred, exposes the entire development process to significant "risk". If the initial assumptions about the drug's profile (such as indication or target patient population) are incorrect, the subsequent development is based on a flawed foundation. Stage-appropriate decision-making mitigates this by delaying key decisions until more data and insights are available.
? Avoids Premature Specialisation: By not locking into a specific product profile or market too early ("avoid premature specialization"), stage-appropriate decision-making allows for the exploration of a wider range of potential applications and target populations as development progresses [3, 4, our conversation history]. This flexibility increases the chances of identifying the most viable and commercially successful path for the asset.
? Enables Learning to Inform Decisions: The core idea is that decisions should be informed by the learning that takes place at each stage. As the molecule progresses through preclinical studies, Phase 1, and Phase 2 trials, valuable information is gathered about its efficacy, safety, pharmacokinetics, and potential patient populations. Stage-appropriate decision-making ensures that this evolving knowledge directly influences the subsequent choices, rather than adhering rigidly to initial predictions.
? Increases Confidence in Later Phases: When decisions about Phase 3 endpoints and the value proposition are made later in the development process, they are based on a much more robust understanding of the drug's potential. This leads to "a phase three is a lot more confidently assigned" because more information has been gathered and alternative paths have been considered.
? Enhances Understanding of the Market: Stage-appropriate decision-making extends to understanding the market as well. By delaying definitive decisions about pricing, dosage, and target segments, companies can gather more information about the competitive landscape and market needs as their asset progresses. This ensures that the final product is more likely to be well-received by the market.
? Shifts Focus from Prediction to Learning: It moves away from the "prediction paradigm" where companies try to forecast everything at the outset. Instead, the emphasis is on "planning to learn" throughout the development process, with decisions acting as milestones based on the accumulated knowledge [5, our conversation history].
In essence, stage-appropriate decision-making acknowledges the inherent uncertainties in early-stage drug development and advocates for a more flexible and adaptive approach. By making key decisions based on the evidence gathered at each stage, companies can reduce risk, maximise the potential of their assets, and ultimately increase their chances of bringing successful products to market. This contrasts sharply with the traditional approach of making early predictions and sticking rigidly to them, even as new information emerges.