A scientific and sustainable approach to business decision making

Following Camuffo et al (2024), this replication project focuses on teaching entrepreneurs, in this case in Uruguay, how to use a structured, scientific approach to make business decisions. Instead of relying on intuition, the program designs an intervention to train entrepreneurs to analyze their ideas systematically. Participants are taught to break down their business models into key components, identify the assumptions behind their strategies, derive hypotheses, design experiments to test these hypotheses, and make decisions based on the results of their testing procedures. This method, modeled after the scientific method, helps entrepreneurs gather real-world customer feedback, test minimum viable products, and make data-driven decisions about whether to proceed, pivot, or abandon their business ideas. Entrepreneurs are divided into two groups: one receives traditional training, and the other learns to apply scientific techniques. The project measures and compares the outcomes of these two groups, particularly in terms of business performance, strategy changes (business model pivots), and whether participants decide to continue or end their ventures. By applying this structured approach, the intervention aims to improve entrepreneurs’ ability to adapt to uncertainty, focus on viable opportunities, and make empirical-based decisions, ultimately enhancing the success of their businesses in a challenging economic environment.

Camuffo, A., Gambardella, A., Messinese, D., Novelli, E., Paolucci, E., & Spina, C. (2024). A scientific approach to entrepreneurial decision-making: Large-scale replication and extension. Strategic Management Journal, 2024(1), 1–29

The intervention in this project is a structured training program designed to teach entrepreneurs how to apply a scientific approach to their business decision-making. The program follows a rigorous methodology inspired by the scientific method, helping entrepreneurs move beyond intuition and adopt evidence-based practices. The training is delivered through sessions that combine group workshops and one-on-one mentorship over four months, alternating weekly between the two formats.

The program begins with an introduction to key tools, such as the Business Model Canvas, to help entrepreneurs outline the components of their business idea, such as customer segments, value propositions, and revenue streams. In the treatment group, participants are explicitly taught to think critically about these components by formulating testable hypotheses about their business models. For example, they might hypothesize that a specific customer segment will find value in their product and be willing to pay a certain price. Entrepreneurs are then guided to design experiments—such as customer interviews, A/B tests, or minimum viable product (MVP) launches—to validate these hypotheses.

A critical part of the training is teaching participants how to analyze and interpret the results of their experiments. For instance, they learn how to collect data objectively, avoid biases, and use reliable metrics to assess whether their hypotheses hold true. Based on this evidence, they are encouraged to make strategic decisions, such as pivoting to a new idea, refining their model, or, if necessary, terminating the project to avoid pursuing non-viable opportunities.

In contrast, the control group receives traditional entrepreneurial training, focusing on frameworks and techniques without the emphasis on scientific rigor. They will follow their intuition. Both groups receive the same number of training hours and access to resources, but the treatment group is taught to approach decision-making with a hypothesis-driven mindset focused on the scientific method. The goal of this intervention is to enhance the precision and quality of decisions, ensuring that entrepreneurs focus their efforts on ideas with the highest potential for success.

The last is a replication experiment that adds another case the work of Camuffo et al (2024). In a following second experiment of this project, the intervention focuses on sustainable business models, introducing additional layers of complexity to the scientific approach. Entrepreneurs are trained not only to test traditional business hypotheses but also to address social and environmental sustainability. To achieve this, the standard Business Model Canvas is adapted into a new model that accounts for the sustainability of the business model, which incorporates environmental and social considerations alongside economic aspects. Participants are guided to identify and test hypotheses related to the sustainability of their ventures, such as the environmental impact of their operations or the social value they create for their communities.

The training emphasizes the importance of validating these sustainability hypotheses with diverse stakeholders, such as community members, regulators, and environmental experts, to ensure the business aligns with broader societal goals. Entrepreneurs learn to design experiments tailored to sustainable business contexts, such as measuring the willingness of customers to pay for environmentally friendly products or assessing the social impact of their offerings. They are also encouraged to integrate sustainability into their value propositions and evaluate whether these strategies enhance the overall viability of their business models. By doing so, the training helps entrepreneurs tackle the additional uncertainties associated with sustainability-focused ventures, such as balancing ecological and societal goals with financial feasibility.

This second experiment is designed to evaluate whether the scientific approach is equally effective in driving success for sustainable business models as it is for traditional ones, based on the idea that sustainable business models operate in settings with higher levels of risk. The program measures outcomes like the ability to make meaningful pivots, the integration of sustainability metrics into decision-making, and the overall performance of the business. By expanding the scope of the scientific method to include environmental and social dimensions, the intervention aims to empower entrepreneurs to create ventures that are not only financially viable but also contribute positively to society and the planet.

Camuffo, A., Gambardella, A., Messinese, D., Novelli, E., Paolucci, E., & Spina, C. (2024). A scientific approach to entrepreneurial decision-making: Large-scale replication and extension. Strategic Management Journal, 2024(1), 1–29

2025-09-01

2025-11-17

Total Revenue Accumulated
Time to Revenue
-Time to First Client
-Termination
-Time to Termination
-Total Pivots

1) The first set of outcomes aims to measure business performance:

-Total Revenue Accumulated: The total revenue generated by a participant’s venture during the observation period, measured in Uruguayan pesos.
-Time to Revenue: The time (in months) it takes for a participant’s venture to begin generating revenue.
-Time to First Client: The time (in months) it takes for a participant’s venture to acquire its first client.

2) The second set of outcomes aims to measure strategic entrepreneurial decisions

-Termination: Whether a participant decides to terminate their venture project (dichotomous variable: 1 = terminated, 0 = not terminated).
-Time to Termination: The time (in months) it takes for a participant to decide to terminate their venture project.
-Total Pivots: The total number of strategic pivots (e.g., changes in business model components) made by a participant’s venture during the observation period.

The experimental design of this project involves two randomized controlled trials (RCTs), each tailored to address distinct entrepreneurial contexts: one focusing on traditional business ventures and the other on sustainable business models. Both experiments aim to evaluate the impact of teaching entrepreneurs a scientific approach to decision-making, emphasizing evidence-based methods for testing business hypotheses and refining strategies. The experiments are designed to ensure robust comparisons between a treatment group, which receives the scientific training, and a control group, which receives conventional entrepreneurial training.
In both experiments, participants are recruited through a comprehensive outreach campaign targeting aspiring entrepreneurs. The selection process involves screening applications to ensure participants have early-stage ventures suitable for the intervention. Once selected, participants are randomly assigned to treatment and control groups to minimize bias and ensure balance across key variables, such as industry, development stage, and geographic location. This randomization allows for a causal analysis of the impact of the intervention on various outcome variables, such as revenue generation, project termination, and strategic pivots.
The training intervention for the treatment group in the first experiment emphasizes using the scientific method to enhance decision-making in traditional entrepreneurial ventures. Participants are taught to identify assumptions in their business models, formulate testable hypotheses, and design experiments to validate these hypotheses. The training includes techniques like customer interviews, A/B testing, and minimum viable product (MVP) development. Control group participants receive conventional training covering similar topics but without the structured scientific framework for decision-making. Outcomes such as revenue growth, time to first client, and the frequency of strategic pivots are monitored over time.
The second experiment, focusing on sustainable business models, incorporates additional elements to address social and environmental dimensions. Participants in the treatment group use an adaptation of the traditional canvas aimed to integrate environmental and social factors into their business strategies. The training emphasizes engaging diverse stakeholders, such as regulators, community members, and environmental experts, to validate sustainability-related hypotheses. Control group participants in this experiment receive training that includes sustainability topics but lacks the hypothesis-driven scientific approach. Outcome variables for this experiment are the same as the first, including traditional metrics like revenue, client acquisition, project termination, and strategic pivots. By combining these designs, the project aims to generate actionable insights into the effectiveness of structured decision-making approaches across diverse entrepreneurial contexts.

The experimental design of this project employs two randomized controlled trials (RCTs), carefully tailored to explore the efficacy of a scientific approach to decision-making across two entrepreneurial contexts: traditional business ventures and sustainable business models. Each RCT is designed to compare the outcomes of a treatment group, which receives training in structured, evidence-based decision-making, with a control group, which receives conventional entrepreneurial training. This setup allows for a rigorous examination of the causal impact of the intervention on key outcomes, such as revenue generation, client acquisition, project termination, and the frequency and nature of strategic pivots.
The first experiment replicates the original study by Camuffo et al. (2020), which was later expanded in Camuffo et al. (2024). For the first time, this experiment will be conducted in a Latin American country—Uruguay—introducing a new context to the research and advancing the evidence on the scientific method in entrepreneurship incrementally, or “in stages.” This approach follows a methodological recommendation to adapt the original design and test it in different settings—in this case, a developing economy—to understand how these contextual changes influence the findings of the original study.
The second RCT also follows this incremental approach by exploring uncharted territory: it will generate new evidence on the impact of the scientific method on nascent ventures aiming to create sustainable businesses. This extension allows for a deeper understanding of how structured decision-making affects entrepreneurial outcomes in ventures that integrate environmental and social considerations alongside financial goals. By adding these dimensions, the study expands the scope of the scientific method’s applicability in diverse entrepreneurial ecosystems.
Participants for both experiments are recruited through a broad outreach campaign using social media, entrepreneurial networks, and targeted advertising. Applicants must demonstrate they are working on early-stage ventures, as these are the most suitable for testing the intervention's effects. After screening applications, participants are randomly assigned to either the treatment or control group. Randomization ensures balance across key variables, such as industry, venture development stage, and geographic location, thereby minimizing selection bias and ensuring comparability between groups.
In the first experiment, the intervention for the treatment group focuses on applying the scientific method to traditional entrepreneurial ventures. Participants are trained to identify and articulate the assumptions underlying their business models using tools like the Business Model Canvas. They then learn to translate these assumptions into testable hypotheses and design experiments to validate or refute them. Techniques such as customer interviews, A/B testing, and the development of minimum viable products (MVPs) are core components of this training. Participants are guided to interpret their experimental results objectively and to adjust their strategies accordingly, whether by pivoting, refining their model, or terminating their venture project.
The control group in the first experiment receives a traditional entrepreneurship training program that covers similar topics, such as market research and product development but lacks the structured hypothesis-driven approach of the treatment group. This group is encouraged to rely on intuition and conventional business heuristics, providing a clear contrast for measuring the impact of the scientific method. Key outcomes tracked in this experiment include revenue growth (measured in Uruguayan pesos), time to first revenue, time to first client, total pivots, and project termination rates. These metrics are analyzed to assess the effectiveness of the scientific training relative to conventional methods.
The second experiment expands the scientific approach to sustainable business models, integrating additional elements that address environmental and social dimensions. The treatment group uses a business model canvas that incorporates economic, social, and environmental factors, enabling participants to consider sustainability alongside financial viability. Training sessions focus on identifying hypotheses related to sustainability, such as environmental impact reduction or social value creation, and testing these hypotheses through stakeholder engagement, including consultations with regulators, community leaders, and environmental experts.
The control group in the second experiment receives traditional entrepreneurship training with a sustainability component but without a rigorous scientific framework. For example, they are exposed to topics like sustainable supply chains and market trends in green products but are not guided to test hypotheses systematically. Outcome variables for this experiment mirror those in the first, including revenue, time to the first client, project termination rates, and strategic pivots.
Statistical analysis is a cornerstone of this project, ensuring robust and reliable conclusions. The primary analysis involves DiD estimates to measure the overall effect of assigning participants to the treatment group or control group. A secondary analysis uses instrumental variable regression to estimate the effect of actual compliance with the intervention, leveraging random assignment as the instrument. This approach helps isolate the causal impact of adopting the scientific approach, accounting for potential noncompliance or varying levels of engagement in applying this method by the nascent entrepreneurs.
The outcomes are modeled using linear regressions for continuous variables (e.g., revenue, time to first client) and logistic regressions for binary outcomes (e.g., project termination). Survival analysis techniques, such as Cox proportional hazards models, are used to examine time-to-event outcomes like time to revenue, first client, or project termination. To ensure robustness, the analyses control for participant characteristics (e.g., industry type, prior experience) and include fixed effects for session instructors and regional factors. The data collection process spans up to 12 months post-training in each experiment, with periodic surveys and structured interviews capturing quantitative and qualitative data. This longitudinal approach allows for the assessment of both short- and long-term impacts of the intervention.

Participants will be randomly assigned to balanced groups based on the same variables used in the original experiments. These variables will be collected through an application form that aspiring participants must complete during the selection process. Additionally, applicants will be asked whether they know other entrepreneurs applying to the training course. After acceptance, they will be asked again if they know any other accepted participants. To address potential biases, if any participants know one another, they will be assigned to the same stratum to ensure that such cases are consistently placed in either the treatment or control group to avoid contamination between groups.
The randomization process will utilize a computer-generated algorithm in R, a widely used and reliable method for RCTs. This approach ensures assignments are random and unbiased, giving each participant an equal chance of being placed in either the treatment or control group, except for cases involving prior knowledge among them. To maintain balance across key variables, such as industry, development stage, and geographic location, statistical tests will be conducted post-randomization to confirm the balance between groups. These tests will align with the protocols followed in the original studies conducted in Italy and the UK.
The randomization process will be thoroughly documented to ensure transparency and adherence to protocol. This includes maintaining and making the codes of the randomization output traceable and recording all decisions made during the process.

Nascent entrepreneurs; except those that know each other. If we have participants who knew each other previously, they will be assigned to strata to ensure they belong to either treatment or control groups. However, it is worth mentioning that the described situation has not been relevant in previous versions of this experiment.

No

116 nascent entrepreneurs

If no entrepreneur abandons their project during the observation period, the experiment will yield a total of 1,392 observations. However, this scenario is highly unlikely. For participants who declare project abandonment, their data will be recorded up to the point of their declaration. Consequently, while the maximum number of observations per experiment is 1,392, the actual total may vary depending on the number of project abandonments.

Each experiment requires a minimum of 116 nascent entrepreneurs, with 58 participants allocated to the treatment group and 58 to the control group.