In 2010, McKinsey conducted confidential interviews with over 2,000 executives, revealing that only 28 per cent believed that “the quality of strategic decisions in their companies was generally good, 60 per cent thought that bad decisions were about as frequent as good ones, and the remaining 12 per cent thought good decisions were altogether infrequent.” A Harvard Business Review Analytic Services Report in 2016 surveyed 646 managers and showed that many rely on gut feeling rather than data-based, systematic, well-organized judgments.

The challenge of making strategic decisions amid high uncertainty is one faced by entrepreneurs, innovators, and management teams. If you wait for certainty, you will never make a decision. So how can we make the best decisions under such circumstances?

There is more data generated and available to help with decision-making than ever before. As suggested by the Cynefin Framework, depending on the decision to be made, analysis of data may give you the answer you are seeking. Alternatively, it may give you more insight into the decision you should make or demonstrate the limits of what is plausible—or, quite possibly, with higher levels of uncertainty, it might only give you a false sense of security about the decision you are making. The answer is not simply to perform more analysis. One problem is that data is often historical or valid as of today. There is no hard, factual data available for the future.

I first wrote about the concept of strategic assumptions in an article published by Ivey Business Journal in 2008. I subsequently published the book Strategic Assumptions: The Hidden Beliefs That Control Every Decision You Make in 2014.

The model I put forward basically posits that there are nine categories of assumptions we must consider as we work through the process of creating a new three-to-five-year strategic plan—indeed, when we are making any strategic decision (defined as one where timelines are longer than usual and the consequences are high).

These nine categories include, firstly, assumptions around (1) the macro-environment (often described through the famous PEST or PESTEL model). For example, the assumption that companies today hold as to whether a certain ex-president will be re-elected in November 2024 in the United States would fall into this category, as does your estimate regarding the outcome of Putin’s invasion of Ukraine, whether the current war between the government of Benjamin Netanyahu and Hamas will end or expand, or the future price of oil.

Similarly, until recently there was a very broadly held macro-assumption that the projected growth of online shopping had been accelerated by five years due to the COVID-19 pandemic—with both online and brick-and-mortar retailers building their long-term planning with that in mind. This has proven to be invalid, with sales in most categories in early 2024 having returned to pre-pandemic levels. For many, it proved a costly assumption to hold.

The other categories of assumptions include (2) how we believe our markets and competition will evolve and behave in the future; (3) the future expected behaviour of our customers (e.g., what, how, and when they will purchase); (4) how stakeholder behaviour will change or not; (5) assumptions we have regarding the state of our physical and non-physical assets; (6) our financial assumptions; (7) how our current internal processes will serve us or need changing; and (8) assumptions regarding our employees (their projected availability, pay demands, turnover, etc.).

The ninth category is mysteriously called “background of shared obviousness” (9). It refers to the personal or organizational beliefs we have about ourselves and the world. These assumptions override all others. For example, whether you believe in a God or not changes your entire “operating system”—the way you see the world. This category also includes the power of a belief in MAGA (that perhaps blinds you to everything else) or whether you believe in humanity’s impact on climate change. Whether you do or do not believe in the latter will influence every strategic decision you make about your company’s production footprint, packaging, and processes.

Many of us have been taught that making assumptions is bad. Making assumptions about other people usually is. However, imagine that a company’s sales distribution, profitability, or optimal production footprint is highly dependent on the exchange rate between the Canadian and U.S. dollars. With today’s technology, I can know the exact exchange rate—to many decimal places—in real time. However, it is not the exchange rate today that is important to my decision. What I would really love to know is what the exact exchange rate will be in five years. If I knew that, it would make my decision easy. But, without a crystal ball, it is unknowable. I have no choice but to make an assumption, either consciously or unconsciously. (I recommend the former.)

Today, it is perhaps unthinkable to assume that the Canadian dollar could be worth more than the U.S. dollar. However, one of my clients believed it could in 2002 (see Figure 1). Challenging every assumption is an essential part of the strategic decision-making process.

FIGURE 1: History of Canadian Dollar–U.S. Dollar Exchange Rate

Source: Adapted from “Pacific Exchange Rate Service,” https://fx.sauder.ubc.ca.

Over the course of the past few years, I have worked with many management teams to uncover, identify, challenge, and validate the strategic assumptions that shape their decision-making.

As a result of my consulting work, I have a lot of anecdotal (though confidential) evidence and personal experience of how difficult it can be for members of any management team to come to an agreement on their final shared assumption about an important yet “unknowable” factor. However, I have never had the time or resources to undertake in-depth, long-term academic research to scientifically prove the value of identifying strategic assumptions during any strategic decision-making process.

Fortunately, in a recent book by Adam Grant called Think Again, the author referred to a scientific article published in Management Science that inadvertently researched deeply into strategic assumptions. It described the findings of a research project in Italy that examined decision-making in 116 entrepreneurial start-up companies in Italy over a one-year period. This first piece of research work was then followed by an even deeper, broader study that included three additional trials for a total of 754 start-ups, including small and medium-sized enterprises in Milan, Italy; Turin, Italy; and London, England.

Using a very thorough, scientific, randomized control trial (RCT) approach to their studies—which included half the companies acting as the treatment group and half as the control group—the management teams in both groups were trained and encouraged to use four widely known models and concepts, which were (1) the Business Model Canvas, (2) customer interviews, (3) the minimum viable product concept, and (4) the iterative prototyping of early product versions.

The only meaningful difference between the start-ups in the treatment group and the control group was the introduction of one more concept, a fifth one, to only the treatment group. With this latter group, the researchers observed how the rigorous identification, testing, and monitoring of “hypotheses” (a synonym for strategic assumptions) in their decision-making during the 12-month period affected the decision-making performance of the treatment group. The control group relied only on “entrepreneurial intuition” and heuristics in its decision-making.

The researchers mainly focused on two categories of hypotheses and assumptions: the market and customers the start-up entrepreneurs focused upon (the second and third categories of assumptions mentioned previously). For these two categories, the entrepreneurs were asked to frame the problem, build a theory, and formulate assumptions—and test these using rigorous experiments with valid and reliable metrics, while setting thresholds for these metrics to make decisions.

These thresholds were, in essence, a synonym of a set of criteria or a set of simple rules that told them when, how, and on what basis to make decisions. For example, one particular company chose to reject an assumption if less than 60 per cent of its customer interviews (sample size of 50) did not provide corroborating evidence to support that assumption.

At the time of the first study (2016–2017), one company, Inkdome, which today is a virtual tattoo studio that uses AI to connect tattoo lovers with the best ink artists, articulated four clear assumptions about its customers: (1) tattooed people do not always use the same tattooist, (2) they often choose new tattooists online, (3) choosing a new tattooist takes time and is difficult, and (4) tattooed people can find online all of the information they need to make their choice. Inkdome’s “simple rule” was that it would only pursue its original business model if all four of these hypotheses were corroborated—otherwise, it would abandon that idea and investigate alternative solutions, i.e., make a strategic pivot, explore a different strategic alternative, or exit. If an assumption was changed, it would then move forward while looking for ways to test and corroborate its revised set of strategic assumptions.

In the second research study, the researchers again cite several examples. One of them is Mimoto, a start-up that initially planned to offer an electric moped-sharing service in Milan. The founders originally assumed that such a service would be analogous to a car- or bike-sharing service and would likely serve a broad range of customers. They tested this assumption. They found that, instead, the service was more likely to be used by specific groups of people. Their assumption was that young people, with mobility needs and the ability to pay, would be the best clientele to target. This prompted them to focus on college students.

Upon testing this assumption, they found that with college students—and others—results were disappointing. They realized that their assumption of targeting “young people with mobility needs” was too vague. College students’ schedules are highly predictable due to them having fixed class schedules. A moto service is really useful when you unexpectedly have to grab a convenient means of transport. The reality was that the company needed to look for customers with unpredictable mobility needs.

In addition, through testing/piloting its business model, the company realized that some people, particularly women, were having a hard time manoeuvring heavy motos. The assumption that motos would be the appropriate mode of transport was wrong. The company switched from motos to scooters, which are lighter in weight.

Both Inkdome and Mimoto are great examples of entrepreneurs consciously identifying their assumptions, testing them, making the appropriate changes, and then testing them again.

The two studies cited above go into great mathematical detail to demonstrate the value of entrepreneurs taking a more scientific approach to decision-making. More specifically, however, what are the major insights, learnings, and conclusions about the use of strategic assumptions in decision-making that can be taken from these empirical studies?

1. The studies found that the revenues of the treatment group (those adopting the use of strategic assumptions) grew much faster than those of the control group in the first year of operation. Over the course of the studies, firms in the treatment group earned around 70 per cent more revenue than those in the control group.

2. The studies found that the treatment group also exited more often and much earlier than the control group, i.e., by identifying and testing their assumptions, members of that group could recognize earlier that a business model was unviable and could cancel or exit the project (see Figure 2).The authors concluded that a scientific approach reduces the odds of pursuing projects with false negative returns (that is, “bad” business ideas). Identifying their assumptions helped firms find, weed out, or reduce errors in their collective thinking and business models.

FIGURE 2: Frequency of Termination

Source: Adapted from Arnaldo Camuffo, Alfonso Gambardella, Danilo Messinese, Elena Novelli, Emilio Paolucci, and Chiara Spina, “A Scientific Approach to Innovation Management: Theory and Evidence from Four Field Experiments,” CEPR Discussion Paper, March 27, 2022 (revision).

3. The treatment group was also less likely to pivot more than twice and was more likely to focus its efforts in a more limited business space—perhaps enabling members to better identify viable opportunities (see Figure 3). It was also less likely to be continually experimenting, jumping from one idea to another. The studies also suggest that once a choice was made, those in the treatment group committed to it.

 FIGURE 3: Frequency of Pivot

Source: Adapted from Arnaldo Camuffo, Alfonso Gambardella, Danilo Messinese, Elena Novelli, Emilio Paolucci, and Chiara Spina, “A Scientific Approach to Innovation Management: Theory and Evidence from Four Field Experiments,” CEPR Discussion Paper, March 27, 2022 (revision).

4. Both studies demonstrate the value of seeking to test or corroborate, as much as possible, all the strategic assumptions on which a business model is based.The use of simple rules in determining the level of corroboration required for a strategic assumption to be fully grounded is a much-needed addition to the strategic assumption process. Setting ourselves the “stretch” challenge of requiring that “all hypotheses have to be corroborated before being adopted” may be impossible in practice, but setting ourselves the challenge to do so would mean it would be almost impossible to move forward on a business idea based upon the unfounded personal beliefs/convictions of any given individual or team.

5. Using a minimum viable product and the iterative testing of prototypes represents an inexpensive way to receive rapid feedback as to whether our strategic assumptions about our markets and potential clients, their buying habits, our whole product offer, etc. are correct. Other categories of strategic assumptions may be more difficult to corroborate, but this example shows the value of attempting to monitor, test, and ground identified strategic assumptions on an ongoing basis.

6. If or when existing assumptions prove incorrect or highly risky, leadership teams should revise them and evaluate other strategic alternatives that are aligned with the new set of assumptions. Imagine, for example, that our initial business model is built on the strategic assumption that we will be able to attract and retain the skilled workforce needed to operate a local production facility. Once we begin to research and test this assumption, we may realize that the salaries and packages required to recruit and retain employees in our region will potentially threaten the financial viability of the project or are simply impossible given the limited local talent pool. Our operating assumption may then become: “We cannot hire the staff to do this locally in-house.” This would lead to the emergence of different strategic alternatives, e.g., we might examine the idea of partnering with or outsourcing our production to another company in the area or we may begin to explore overseas manufacturing alternatives. After consideration, we may choose to kill the first idea, pivot, and seek ways to test and corroborate the new business model and strategic assumptions with new research, tests, and prototypes.

7. The authors observed that the effects of teaching entrepreneurs to use a more scientific decision-making approach had a statistically significant effect for younger entrepreneurs with lower education levels and less experience. This suggests that even entrepreneurs with little formal business education can greatly benefit from learning and implementing the concept of strategic assumptions in their decision-making.

I have always believed that the optimal strategic decision-making process includes data collection and analysis, but that the notions of intuition, gut feelings, and instinct—along with the co-creation of a collective, emergent conversational environment—are equally important, particularly in times of high uncertainty. A “background of shared obviousness” assumption that I personally hold is that today, every industry is characterized by high levels of uncertainty.

Strategic assumptions represent an easily adopted, transparent generative mechanism through which decision makers can gain the maximum from fusing data analysis and intuition together in a generative conversational process.

Having witnessed the value of using strategic assumptions many times over the past years, it is now gratifying to have found large, long-term RCT studies that provide real-world empirical evidence to support the argument that identifying, corroborating, and finally validating a set of strategic assumptions can improve everyone’s decision-making capability and performance.

Why does this approach help us all? Many of us, even if we work in large organizations, now operate in more “entrepreneurial” environments and many of the decisions we make are strategic (i.e., the consequences are high), if not for the organization then certainly for our careers.

We can each begin to identify the strategic assumptions we are making in the nine different categories and seek corroboration when we have difficult decisions to make. The theory and anecdotal and scientific results now show that the “quality” of our decision-making will improve by doing so. This approach will help us eliminate biases and decide which strategic alternatives are viable, when to pivot, when to exit, and when to drive forward and commit our full resources to the implementation rollout.

The strategic assumption that I now hold and that I hope will prove true is that “You will do so.”

I would like to express my deep appreciation of and gratitude towards the authors of all the research I refer to in this article (in particular, Arnaldo Camuffo, Alfonso Gambardella, Danilo Messinese, Elena Novelli, Emilio Paolucci, and Chiara Spina) and apologize if I have misinterpreted, poorly summarized, or inadvertently misrepresented any of their findings. My effort was solely to apply their findings to the area of strategic assumptions and decision-making and to help expose their work to new audiences.

REFERENCES

• Dan Lovallo and Olivier Sibony, “The Case for Behavioral Strategy,” McKinsey Quarterly (2010).
• Mark Hollingworth, “Strategic Assumptions: The Essential (and Missing) Element of Your Strategic Plan,” Ivey Business Journal (November/December 2008), https://iveybusinessjournal.com/publication/strategic-assumptions-the-essential-and-missing-element-of-your-strategic-plan.
• Mark Hollingworth, Strategic Assumptions: The Hidden, Yet Powerful Beliefs That Control Every Decision You Make (5i Strategic Affairs, 2014).
• Matthew Townsend, “The Great Post-Covid Online Shopping Bet Was a Costly Delusion,” Bloomberg Businessweek, October 11, 2022.
• Adam Grant, Think Again: The Power of Knowing What You Don’t Know (New York: Viking, 2021).
• Arnaldo Camuffo, Alessandro Cordova, Alfonso Gambardella, and Chiara Spina, “A Scientific Approach to Entrepreneurial Decision Making: Evidence from a Randomised Control Trial,” Management Science 66, no. 2 (2020): 564–586.
• Kathleen M. Eisenhardt and Donald N. Sull, “Strategy as Simple Rules,” Harvard Business Review 79 (2001): 107–116.
• INKbusiness (website), accessed March 19, 2024, https://business.inksearch.co.
• Arnaldo Camuffo, Alfonso Gambardella, Danilo Messinese, Elena Novelli, Emilio Paolucci, and Chiara Spina, “A Scientific Approach to Innovation Management: Theory and Evidence from Four Field Experiments,” CEPR Discussion Paper, March 27, 2022 (revision).
• Mimoto (website), accessed March 19, 2024, https://helbiz.com/enCA/mimoto.