1.1 Effectuation: The Original Study of Expert Entrepreneurs

The theory, or more accurately the logic of effectuation, originated in a study of entrepreneurial expertise (Reference SarasvathySarasvathy, 2009). In conversation with Herbert Simon, advisor to my doctoral dissertation, I designed a think-aloud protocol study of expert entrepreneurs. Drawing upon studies of expertise in over 200 domains, I defined an expert entrepreneur as someone with ten or more years of full-time immersive experience starting and running multiple ventures, including successes and failures, and at least one public company. The objective for the sampling frame was to ensure breadth and depth of experience, as well as proven verifiable performance. A total of 245 people fit the criteria, and I wrote to all of them, of whom 45 accepted. The final sample encompassed a wide variety of industries, ranging from software and biotech to retail, personal services, even railroad and steel.

The research design used the classic think-aloud method from expertise studies (Reference Ericsson and SimonEricsson & Simon, 1993). In this method, experts are asked to do what they typically do in their domain of expertise while concurrently thinking aloud continuously. For example, chess grand masters would be asked to play chess but continuously think aloud as they contemplated their next move. Concurrent thinking aloud accomplishes two things. First, it prevents people from making up stories retrospectively, whether about how or why they make the moves they do. Second, it allows the researcher to reconstruct step-by-step the heuristics or rules of thumb used by the experts. The term heuristics here refers to how experts process information during the task: which pieces of information they ignore, which ones they pay attention to and in what order, how they combine bits with other bits and/or transform them in which ways, and so on (Reference Kuusela and PaulKuusela & Paul, 2000).

For my study of expert entrepreneurs, I created a seventeen-page problem set of ten typical decisions that all entrepreneurs are confronted with in starting and running new ventures. By design, the experts in my sample had very little in common in terms of industry or technology experience. Yet, they had almost everything in common in terms of entrepreneurial experience. Therefore, I made entrepreneurship itself the product for the problem set used in the think-aloud task in the study. Remember, the “task” consisted of building a venture.

The seventeen-page problem set began with a brief description of the idea for a new venture based on an educational game of entrepreneurship, followed by a series of questions about potential customers, competitors, and growth prospects. I also asked what information the subjects would seek about these and how they would go about obtaining that information. I then provided information I had gathered about this venture idea and asked them to make a series of decisions, including marketing, funding, hiring, growth, crisis management, and exit.

I went out in person to each subject to gather and record the think-aloud data. Subjects spent anywhere between two and four hours on the task. During the task, I would not interrupt or ask questions, except to prompt them to continue thinking aloud if they paused for more than four seconds. After they completed the think-aloud task, I would ask a few open-ended interview questions.

Even as the recordings were being transcribed, I would listen to the protocols over and over again in airports, planes, and hotels, and in between travel times. Initially, the recordings sounded like any entrepreneur anywhere talking about building new ventures. But listening closely and repeatedly, one thing started emerging from the data, almost by the third protocol. Each expert entrepreneur expressed negative opinions about doing market research. Comments ranged from, “I don’t believe in market research” to “I would not do anything like surveys or focus groups because they have never worked for me.”

As I puzzled over why they might be averse to market research and how they made marketing decisions while ignoring market research data, I eventually discovered that their comments generalized beyond market research to encompass all predictive information.

Bottom line was that expert entrepreneurs exhibited a deep distrust of predictive information of any kind. In follow-on interviews, one subject put it this way: “I think predictions about the past are OK. But don’t believe anyone who is predicting the future, even when they are talking about what they themselves will or won’t do or want or don’t want.”

Armed with this insight, I set about scouring a variety of literatures on prediction. One of the earliest and most inspiring insights on the topic came from Reference KnightFrank Knight’s (1921) dissertation, “Risk, Uncertainty and Profit,” which was and is still in print and studied to this day. To analyze the think-aloud data, I devised a coding scheme based on Frank Knight’s distinctions between risk, uncertainty, and true unknowability (what we call Knightian uncertainty).

1. 1. Risk refers to situations characterized by a known distribution but an unknown draw. Trying to guess the color of a ball drawn from an urn containing fifty green balls and fifty red balls is an example of risk.

2. 2. Uncertainty consists of unknown draws from unknown distributions. This is akin to guessing the color of a ball drawn from an urn consisting of an unknown number of balls of an unknown number of colors. Here, we would have to make repeated draws over time to create an estimation of the distribution of balls in the urn. Thereafter, we can do calculations of odds, similar to those in the case of risk.

3. 3. Knightian uncertainty comprises situations where the distribution is not only unknown but also unknowable. We can imagine this to be an urn that consists of totally random objects, not just balls. No matter how many draws we make from it, we will not be any closer to an accurate estimation of what is in the urn.

It turned out that the lessons that expert entrepreneurs had learned through years of deliberate practice, starting and running ventures, had taught them ways to tackle this third kind of fundamental unpredictability (Reference Dew, Ramesh, Read, Sarasvathy, Virginia, Ericsson, Hoffman, Kozbelt and WilliamsDew et al., 2018; Reference Ericsson, Krampe and Tesch-RömerEricsson, Krampe, & Tesch-Römer, 1993). Additionally, they also learned to deal with ambiguities about their own goals (Reference MarchMarch, 1978; Reference SlovicSlovic, 1995) as well as ways to determine what would count as relevant and irrelevant information in the face of unknowable futures combined with ambiguous goals (Reference Chiappe and KuklaChiappe & Kukla, 1996; Reference Fodor, Adler and RipsFodor, 1983/2008).

In sum, the following three aspects characterize the effectual problem space:

1. 1. Knightian uncertainty: The future is not only unknown, but also unknowable

2. 2. Goal ambiguity: Goals are often not clearly specified, even when aspirations are present

3. 3. Isotropy: It is not obvious which information is relevant and which is not

Note that the term “isotropy” has slightly different meanings in different contexts. In engineering, for example, it refers to the property of a material that is invariant with respect to direction. In philosophy, it refers to the epistemic interconnectedness of beliefs in the sense that “Crudely, everything that the scientist knows is, in principle, relevant to determining what else he ought to believe. In principle, our botany constrains our astronomy, if only we could think of ways to make them connect” (Reference Fodor, Adler and RipsFodor, 2008, p. 903). Cognitively speaking, “ … isotropy is the principle that any fact may turn out to be (ir)relevant to the confirmation of any other” (Reference Fodor, Adler and RipsFodor, 2008, p. 905). In other words, isotropy points to the problem of informational relevance.

1.2 Effectuation: Developments Over the Past Two Decades

Over the past two decades, I have developed the findings from that first study into the five principles of effectuation that three of my coauthors and I explored empirically and enhanced theoretically (Reference SarasvathySarasvathy, 2001; Reference SarasvathySarasvathy, 2009). Nicholas Dew studied effectuation through an in-depth history of the radio frequency identification industry, which led to the development of the dynamic process model in Figure 1 (Reference Sarasvathy and DewSarasvathy & Dew, 2005). Robert Wiltbank collected data from angel investors to help develop performance implications for effectuation (Reference Wiltbank, Read, Dew and SarasvathyWiltbank et al., 2009). And Stuart Read has worked to connect both principles and process to the practice of marketing, entrepreneurial pedagogy, and strategy inside large corporations, as well as within small and new ventures (Reference Read, Song and SmitRead et al., 2009; Reference Read, Sarasvathy, Dew and WiltbankRead et al., 2016; Reference Read, Song and SmitRead, Song, & Smit, 2009). Together, the four of us developed the first survey-based measures and published over a dozen peer-reviewed articles using a variety of different methods.

Figure 1 The effectual process

After the first few of our articles appeared, more researchers became interested, and now over 700 articles have been published, including about 100 in top-ranked journals (Reference Alsos, Clausen, Mauer, Read and SarasvathyAlsos et al., 2019). However, most of the important work remains to be done. Extant works have largely focused on demonstrating existence. These studies found evidence for the use of effectuation in several different business domains, such as family business (Reference Bloemen-Bekx, Lambrechts and Van GilsBloemen-Bekx, Lambrechts, & Van Gils, 2021), international business (Reference Harms and SchieleHarms & Schiele, 2012; Reference Karami, Wooliscroft and McNeillKarami, Wooliscroft, & McNeill, 2020; Reference Mainela and PuhakkaMainela & Puhakka, 2009), social enterprise (Reference Corner and HoCorner & Ho, 2010; Reference Servantie and RispalServantie & Rispal, 2018), R&D departments (Reference Brettel, Mauer, Engelen and KüpperBrettel et al., 2012; Reference Fischer, Greven, Tornow and BrettelFischer et al., 2021; Reference Szambelan and JiangSzambelan & Jiang, 2020) and social media (Reference Fischer and ReuberFischer & Reuber, 2011). Data have been gathered from over fifty countries using a variety of methods ranging from surveys and case studies to mathematical models and conjoint analyses. Several studies have also examined the relationship between differential or combinatorial uses of predictive and effectual approaches in these different settings, relating them to different types of entrepreneurial and innovative activities (Reference Alsos, Clausen, Hytti and SolvollAlsos et al., 2016; Reference Frese, Geiger and DostFrese, Geiger, & Dost, 2020; Reference Gabrielsson and PolitisGabrielsson & Politis, 2011; Reference Reymen, Berends, Oudehand and StultiënsReymen et al., 2017; Reference Smolka, Verheul, Burmeister-Lamp and HeugensSmolka et al., 2018).

However, as I will outline in this Element, in-depth examinations of the radical revisions of fundamental concepts in the social sciences that the logic of effectuation calls for have barely begun to be touched upon. Before I get to that in Section 2 of this Element, let me provide a brief outline of the principles, process, and performance implications of effectuation.

1.3 Effectuation: Principles, Process, and Performance Implications

The original study yielded five heuristics. At first glance, these could be seen as unconnected, but a deeper look shows that they are logically consistent with each other. Each reduces, and even minimizes, the use of predictive information. Each also offers systematic criteria for acting in the face of a truly unknowable future. Together, they form a logic of non-predictive control.

This might come as a surprise to those imbued with the notion that prediction and control are tightly related, even inseparable. In other words, either prediction is already equal to control, or any distance between them is purely a matter of missing information. Both of these conceptions rule out the possibility of prediction being orthogonal to control.

Traditionally, the relationship between prediction and control has been dominated by the positivist paradigm in science, where control is impossible without prediction. In this view, there is a congruence between prediction and control (Reference SellarsSellars, 1939). Even in situations where we do not have access to predictive information in full, or where there exists some distance between prediction and control, it is still prediction that leads to control. In other words, we need predictive information and strategies based on that to achieve control. A simple example may serve to clarify these relationships between prediction and control. Predicting whether it will rain or not can help us avoid getting wet. Moreover, accurate predictions of rain can help us decide whether we need to carry an umbrella or even choose whether to go out at all or not.

However, by simply carrying an umbrella, we can ignore predictions of rain. In this case, by doing something within our control, namely carrying an umbrella, we obviate the need to make predictions, let alone worry about the accuracy of those predictions. It is in this sense that the five heuristics derived from the study of expert entrepreneurs challenge this tight link between prediction and control. Instead, effectuation posits not only distance but also an orthogonal relationship between prediction and control.

In sum, traditional (causal) views are based on the logic: To the extent we can predict the future, we can achieve control over it. The effectual view inverts this logic: To the extent we can work with things within our control, we don’t need to predict the future. Note that both prediction and control can be more or less accurate, more or less error-prone. All the same, the two logics, causal and effectual, point to different sets of possible actions under uncertainty.

In other words, effectuation exposes control itself as a strategy to shape and cocreate new futures. In making new futures, predictions about those futures are neither necessary nor valid. Knowledge about the past and present is still valid and useful, even when such knowledge uses prediction as the touchstone for validity, such as in scientific knowledge. But predictions about what we can and cannot do with such knowledge are often useless and irrelevant. Even more unpredictable are the issues of what other people will and will not do with extant knowledge or how they might react to what we do with it. Yet, people can commit to using things within their control in particular ways that obviate the need to predict what they might otherwise do or not.

Through the actions and interactions of venture building, expert entrepreneurs learn the following five principles of effectuationFootnote ¹:

1. 1. Bird-in-hand: Instead of clearly defined goals or predicted opportunities, they begin with who they are, what they know, and whom they know.

2. 2. Affordable loss: Instead of calculating risk-adjusted expected return for making investments, they invest no more than they can afford to lose.

3. 3. Crazy quilt: Instead of targeting particular potential stakeholders for particular resources, they work with anyone and everyone who commits real resources to the venture, allowing those who commit to self-select into the venture as actual stakeholders.

4. 4. Lemonade: Instead of avoiding surprises, they leverage both positive and negative contingencies.

5. 5. Pilot-in-the-plane: Instead of viewing the future as driven by trends outside their control, they learn how human beings shape and cocreate new possibilities.

The process model in Figure 1 shows how the five principles of effectuaion work together iteratively over time to fabricate a variety of artifacts, including new products, ventures, markets, opportunities, sociopolitical institutions, and other elements of new futures. Note that the process can begin anywhere in the diagram. For ease of comprehension, let us begin with the “bird-in-hand” box.

Effectual entrepreneurs begin with what is already within their control, or more accurately, what they perceive or deem to be within their control. These consist in elements of their identity (who they are), knowledge (what they know), and network (whom they know). Based on means within their control, they come up with possible actions. These actions are usually mundane, even routine parts of the way they live and act, adding up to venture ideas that are not likely brilliant, nor even incrementally innovative.

Note that there is no assumption here that mere possession of something implies any prediction of what one could or should do with it. Any effectuator’s knowledge of possible actions based on what they deem within their control may be incomplete or nonexistent, or simply wrong. Yet, by simply taking inventory of what they possess, they can come up with actions they could take, irrespective of whether those actions thereafter lead to any particular effects they may or may not predict. In other words, there is no need to assume a tight coupling of cause and effect in seeing something as within one’s control.

Let’s consider again the example of an umbrella. It could be argued that when one sees an umbrella, one sees its predicted effect in protecting us against rain and sun. Yet, when asked what can be done with an umbrella, people come up with a number of other uses for it such as a walking stick, basket, weapon, art, and so on. In the context of a venture, an umbrella could evoke ideas such as insurance (Travelers Insurance), golfing (Arnold Palmer Brands), banking (Citibank), and so on. The point here is that control of means is not limited to one’s predictions of effects arising from its use. Instead, simply by approaching means as drivers of actions, it becomes possible to produce a one-to-many mapping from them to a variety of viable and valuable actions, even when the ensuing effects of those actions are not or cannot be predicted.

Effectual entrepreneurs choose between these actions/venture ideas based on what they can afford and are willing to lose. This allows them to keep the downside within their control while allowing upsides to be forged through action rather than predicted ahead of time. Using affordable loss criteria also focuses attention on things worth doing, even if the venture eventually fails. Traversing the first two boxes in Figure 1, consisting of bird-in-hand and affordable loss, can be rather quick, doable on Day 1 as it were. These can also be entirely subjective and idiosyncratic, especially at the beginning of the process.

On Day 2, effectuators start interacting with other people, some of them targeted, but most contingent on their current lives and circumstances. Each interaction offers the possibility of stakeholder self-selection. Parties to each interaction can choose to invest what they can afford to lose, or not.

When an interaction results in commitments from both or all sides, new means start to become available, kick-starting an expanding cycle of resources. At the same time, goals of the new venture also get reshaped, resulting in a converging cycle of constraints. Taken together, the two cycles, expanding and contracting, eventually coagulate into new possibilities.

Along the way, things completely outside the control of all current stakeholders can impact the process, positively or negatively. Effectual stakeholders transform these contingencies into new means and new constraints that keep the process moving forward toward the cocreation of new futures, key elements of which may involve new products, ventures, markets, opportunities, and institutions.

The process does not guarantee success or even completion. Everyone involved always has the option to quit when their affordable loss levels are reached or when an external contingency destroys the venture. However, when success does occur, there is a high probability that the artifact cocreated through that success will be innovative. Similarly, when failure happens, chances are good that its costs would be low since no one loses more than they can afford to lose.

With this succinct summary of effectuation in hand, in Section 2, as we examine its implications for revising a series of taken-for-granted assumptions in the social sciences, we will also delve deeper into the principles and process, as well as flesh out a variety of nuances.

2.1 Prediction and Control

It is conventional to think of prediction as leading to control. In other words, the better we can predict, the more control we have, at least over our own outcomes, if not over the phenomena themselves. Hark back to the umbrella example earlier. This relationship between prediction and control is depicted as a point on the left-hand side of Figure 2. Contrast this with the right-hand side where this point expands into the two dimensions of a finite space. Another way to see the difference is that in the conventional view, the space between prediction and control vanishes into a dimensionless point.

Figure 2 The prediction-control space

It is important to note that the prediction-control space is not infinite. Uncertainty bounds the space at any given point in time. Yet, the space can be transformed not only through exogenous forces but also through endogenous actions and interactions, reshaping the uncertainty. In other words, the space itself evolves through time, both through the accumulation of knowledge as in science, but also through the cocreation of new ends worth achieving, as in effectual entrepreneurship.

Simply adding the caveat that predictions be empirically testable and falsifiable has led to an astonishing development of knowledge over the course of recent human history. This is why prediction is conventionally considered the cornerstone of science (Reference FeiblemanFeibleman, 1972; Reference RychlakRychlak, 1964). We will delve more into this later in Section 3 when we return to reversed relationships between prediction and control as the core of differentiating the scientific from the entrepreneurial method.

Here, I would like to point out that several developments in business research and education embrace the idea that prediction leads to control. Why do we do market research? So that we can estimate future demand and select between alternate products to cater to it (Reference BemmaorBemmaor, 1995). Prediction can help map out the potential market space so we can apply strategic analyses such as Porter’s five forces (Reference PorterPorter, 2008) or Strengths Weaknesses Opportunities Threats (Reference AndrewsAndrews, 1997) to figure out sustainable competitive advantages. In financial decisions, we make estimates of future cash flows and potential risks in order to calculate the net present value (NPV) of risk-adjusted expected returns (Reference Brealey and MyersBrealey & Myers, 1984). The project with the higher NPV is the optimal candidate for investment. Even in human resources management, we seek to predict which of any two or more candidates is more likely to perform better on the job (Reference Sackett, Gruys and EllingsonSackett, Gruys, & Ellingson, 1998; Reference SalgadoSalgado, 2003). And so on.

Business books are replete with ways to predict the future. This can also include predicting the behavior of human beings in the face of predicted and uncertain futures (Reference Harris, Lee, Thompson and KrantonHarris et al., 2016). Sometimes, these behavioral predictions are inputs into system-level predictions, and at other times, the predicted behaviors indicate outcome scenarios.

Expert entrepreneurs learn the hard way that predictions often don’t work, and even when they do, they do not automatically or even easily lead to control. Most times, they do not even tell us how to decide between alternate courses of possible actions, not only in the longer run but also in the immediate future.

Yet, so much of the world is organized around predictive models that potential stakeholders routinely ask for predictions; for example, investors. Most banks require entrepreneurs to calculate and submit financial projections, which in turn rely on predictions of demand and technological trajectories, as well as the future behaviors of actual and potential competitors. Often, macroeconomic forecasts and political trends also feed into these projections. Formal and informal advice related to fundraising urge predictive data collection of one kind or another. Everything from in-depth ethnographic interviews with potential customers (Reference LiedtkaLiedtka, 2014) to big data analyses (Reference SainiSaini, 2018) are offered as tools for building ventures more likely to succeed.

However, these models do not always highlight the fundamental assumptions that they take for granted and build on. For example, what if one is wrong in choosing which people to talk to and gather data from in the first place? What if the current set of possible markets does not include new markets that unexpectedly come to be? Or the biggest assumption of all, namely, that there exists “the” future which can be predicted at all.

Of course, predictive techniques are most useful in the face of risk. When dealing with an unknown draw from a known distribution, we can indeed make accurate predictions that result in better control of outcomes, provided, of course, we have clear preferences on the set of possible outcomes, as in the case of wanting to avoid getting wet in the rain. Prediction can also be useful in the face of unknown draws from unknown distributions. Assuming we have the time and resources to make multiple draws and learn through sophisticated estimation and experimentation techniques, even in the face of uncertainty, better prediction can lead to better outcomes, albeit at a higher cost (namely, the cost of estimation and experimentation).

However, in the face of true or Knightian uncertainty, prediction simply does not and cannot work. Entrepreneurs often face Knightian uncertainty and begin losing faith in prediction. As we will see a little later, they might even overlearn this lesson and choose not to invest in predictive techniques, even when faced with risk and uncertainty, rather than Knightian uncertainty. But they offer reasons as to why they choose to make this error.

In brief, expert entrepreneurs learn the following lesson: To the extent you are willing to work with things already within your control, you don’t need to predict the future. Nor do you need to chase things outside your control. This allows control to develop into a dimension of its own. One implication of this lesson is that prediction and control can be conceptualized as orthogonal axes, as depicted on the right-hand side of Figure 2. Now, expanding the vanishing point on the left-hand side into a space of four possible strategies for action on the right gives us the CAVE framework:

1. 1. Causal: High in prediction, but low in control (top left)

2. 2. Adaptive: Low in both prediction and control (bottom left)

3. 3. Visionary: High in both prediction and control (top right)

4. 4. Effectual: Low in prediction and high in control (bottom right)

Note that ex ante, these have to do with the perception of what and how much is predictable and/or controllable, not what actually turns out to be so ex post. In other words, expert entrepreneurs learn to perceive the future as unpredictable and rely on things perceived to be within their control as more reliable drivers of what to do next. We will return later to an examination of errors in perceptions as to which quadrant one finds oneself in. For now, let us analyze each of the four quadrants as viable strategies.

The heroic visionary has always been the stuff of legend, intrepidly leading humanity out of caves and on to the conquest of space and time. This is true of entrepreneurship as well. The rags-to-riches myths of Horatio Alger are a case in point (Reference WeissWeiss, 1969). So is the award given to honor perseverance, integrity, and excellence by the Horatio Alger Association of Distinguished Americans. Popular books also celebrate notions of visionaries who can see around the corner, who skate to where the puck will be rather than where it currently is (Reference Christensen, Raynor and VerlindenChristensen, Raynor, & Verlinden, 2001). A similar celebration of vision permeated knowledge development before the scientific method became widely taught (Reference Westfall, Lindberg and NumbersWestfall, 2020). Divine revelation and the ability to “read the signs of nature” were the hallmarks of ability and achievement in the arts and craftsmen’s guilds, as opposed to “tinkerers” who chanced upon inventions that, even when useful, could not rise to the theoretical understandings of philosophers and scholastics.

The rise of science arguably eroded the heroic pedestal, slowly replacing it with a more careful and systematic machinery of method. All the same, notions of scientific “genius” continue to spotlight the theoretical accomplishments of men such as Einstein, Turing, and Darwin, while also celebrating the plodding data collection and experimental work of a growing body of thousands of scientists, such as those working on the Hadron Collider or the Human Genome project (Reference SimontonSimonton, 1988).

Predictions count not only in the natural sciences but also in the social sciences (Reference HindmanHindman, 2015; Reference MirowskiMirowski, 1991). The touchstone for theoretical achievement in any science is the testability of a theory’s predictions and the subsequent discoveries arising out of efforts to falsify those predictions. Predictions that stand unfalsified get incorporated into technologies and policies. These rely on the predictions to build useful products and institutions that improve human life and society.

The sciences are careful and cognizant of the fact that predictions do not automatically imply control. In fact, the second half of the twentieth century thoroughly shook up any complacent expectations that scientific knowledge would inevitably lead to complete control of the forces of nature. Developments in quantum physics had already raised doubts that were only reinforced through the mathematics of chaos theory and rising attention to complex and self-organizing phenomena such as weather patterns and viruses (Reference WerndlWerndl, 2009). Not to mention the continual contradictions of sociopolitical institutions and the human psyche failing to keep up with technological advances (Reference Gazzaley and RosenGazzaley & Rosen, 2016; Reference MosesMoses, 2007; Reference VoltiVolti, 2005).

For now, I call the high prediction-low control (top left) quadrant of Figure 2 “Causal” to signify the fact that predictions offer insights into cause–effect relationships that we can then use to obtain specific effects we seek to obtain.

Given that predictions do not always lead to control and that the increasing speed of technological advances can destabilize markets and societies in unpredictable ways, adaptive approaches remain important in modernity as much, if not more, as in primitive times. Silicon Valley relies on catchphrases such as “agile” and “lean,” seeking to shorten the distance between knowledge and action as they both rapidly coevolve to produce novelty and financial value. So much so that the very word “entrepreneurial” is sometimes deemed synonymous with adaptive.

Both causal and adaptive evoke the possibility of sidestepping the unexpected, even when explicitly acknowledging its inexorability. Based on increasingly sophisticated ways to gather and analyze more data, the causal approach assumes that it is possible to carefully plan ways to avoid surprises. The adaptive approach advocates nimbleness. By quickly reacting to incoming surprises, the entrepreneur can both avoid the consequences of unpleasant surprises and jump on new possibilities offered by pleasant ones.

In contrast to these three approaches, effectual entrepreneurs learn systematic ways to minimize the use of prediction. While control is a desirable outcome in all four quadrants, control becomes strategy in the effectual quadrant. This makes sense because prediction is ruled out in the effectual case. Non-predictive control, also known as effectuation, involves using things within one’s control to shape and cocreate rather than predict the future.

2.2 Means and Ends

It is conventional to begin with a goal. In most decision-theoretic and even behavioral models, goals are given and outside the scope of the analysis. Given a goal, we can examine how to achieve it, tracing out which causes lead up to it.

Thus, goal-driven action, as depicted on the left-hand side of Figure 3, begins with a goal and then seeks means that can bring it about. Since, in many cases in life and almost all situations in entrepreneurship, not all means necessary to the achievement of the goal will be within the control of founding entrepreneurs, they will have to engage in finding those means that are outside their control. When goal-driven, resource seeking becomes an important aspect of building new ventures. This then sets in motion the search for particular resource providers. Thus, seeking out and convincing targeted resource providers drives subgoals in new venture creation.

Figure 3 Means and ends

An effectual approach begins only with means already within the entrepreneurs’ control. Entrepreneurs imagine several different possible proximate subgoals that are achievable with means already at hand. The means consist of their identity (who I am), knowledge (what I know), and network (whom I know). The ends arise from examining the question, “Given who I am, what I know and whom I know, what kinds of actions are doable and worth doing?” or “What can we do, given who we are, what we know and whom we know?” Notice that this can work with individual entrepreneurs as well as founding teams.

It is also possible for firms, and even communities and countries, to use this heuristic as a starting point for action. An example of this can be found in a comparative study of South Korea and Taiwan that shows how and why Taiwan specialized in higher-quality intermediate inputs, such as vehicle parts, whereas Korea built on its large, vertically integrated business groups known as “chaebol’s” to produce and export higher-quality final goods, such as cars (Reference Feenstra and HamiltonFeenstra & Hamilton, 2006).

Additionally, the initial set of means, as described by expert entrepreneurs, does not emphasize “What I/we have.” Instead, the very notion of a resource, or something owned or possessed, is seen as a function of identity, knowledge, and network. Why is this important? Because seeing resources as a creative combination of who I am, what I know, and whom I know is a way to generate a larger, more open set of possible ends. Conceptualizing resources in the more familiar sense of money or time immediately brings attention to constraints, to things that could be lost, a subject better addressed in the next section where we will examine the affordable loss principle. In contrast, identity, knowledge, and network are not depletable in the sense in which money and time may be.

This view of resources coheres much better with the way Reference PenrosePenrose (1959) conceptualized resources than more recent views such as those found in the resource-based (Reference BarneyBarney, 1991) or resource-dependency (Reference Pfeffer and SalancikPffeffer & Salancik, 1978) views in strategic management. Resources in the Penrosean sense are versatile in a variety of different ways. For example, existing resources may be underutilized, creating slack within the system (Reference Cyert and MarchCyert & March, 1963/1992). Additionally, resources adapted to one use may be exaptable to other and new uses (Reference Andriani and CattaniAndreani & Cattani, 2016; Reference Gould and VrbaGould & Vrba, 1982). This versatility pervades effectuators’ views of means-driven action. Additionally, as we will see in more depth later, this view reframes even waste and negative contingencies into positive possibilities through the affordable loss and lemonade principles respectively.

The conventional view of formulating action as goal-driven is a way of narrowing down outcome choices while broadening the need for resources. It consists of a many-to-one mapping, where all roads leading to a predetermined goal are in focus rather than the choice of the goal itself. The effectual view is means-driven and can be seen either as a many-to-many mapping, or when we consider identity, knowledge, and network as combined into a single set of means, a one-to-many mapping.

Means-driven action is narrower and more constrained in terms of resources, but it can open up and enlarge the set of possible outcomes. Since an end is not specified ex ante, unpredictability in outcomes is not seen as a problem in a means-driven approach. It is already inherent in the process. In fact, as explicated later, goal ambiguity itself becomes a resource to be leveraged through the crazy quilt principle to allow stakeholders to self-select into the effectual process. In the goal-driven case, both ends and means are constrained, even when entrepreneurs may get very creative with the way they combine their readily available means, as in the case of bricolage (Reference Baker and NelsonBaker & Nelson, 2005; Reference Levi-StraussLevi-Strauss, 1966).

We need to be careful about separating means from ends in too simplistic a manner. As Simon argued, means and ends exist in hierarchies (Reference SimonSimon, 1964). Today’s goals achieved can become tomorrow’s means for achieving larger or newer goals. Similarly, today’s failures can constrain tomorrow’s means. Furthermore, existing knowledge of means–ends relationships constrains what is even perceived as means or as possible ends. As Reference GigerenzerGigerenzer (1991) showed through a historical analysis, tools often drive new theories, defying conventional wisdom of movement in the other direction. In any case, developments of tools and theories, means and ends, often go hand-in-hand or in iterative reflexive cycles.

Note that it is not that effectuators don’t have goals or should not begin with clear goals. It is just that goals subserve means-driven action. When push comes to shove, effectuators are open to reshaping their goals rather than predicting and pursuing means not within their control. Put differently, effectuators are more tightly tethered to means than to goals, and more attentive to what is possible within their control than what may only be probable due to things outside their control.

Yet, it matters whether one approaches entrepreneurship from a goal-driven or a means-driven perspective. When entrepreneurs are closely tied to a predetermined goal, as in the case of visionary or causal approaches to venturing, their focus immediately turns to which means are necessary but not currently available. In other words, goal-driven action spotlights missing resources or the lack of needed resources, thereby reinforcing an attitude of scarcity. This awareness of scarcity further obscures ends that could be achieved with means already at hand.

2.3 Risk and Return

Figure 4 depicts the standard conceptualization of the risk curve with Return on the vertical axis and Risk (usually measured as volatility in finance or, more generally, variance in statistics) on the horizontal axis.

Figure 4 Risk and return

In this view, most theories begin with the predicted linear relationship along the 45-degree angle, wherein higher risks are associated with higher returns and lower returns are associated with lower risks (Reference Chang and PinegarChang & Pinegar, 1988; Reference MarkowitzMarkowitz, 1991). The intuition here is that returns compensate for risks taken. This not only has the advantage of serving to incentivize necessary risk-taking, without which we cannot survive as a species or an economy, but also has the virtue of earners of higher returns “deserving” the returns they earn, as justified compensation for risks they dared to take.

In terms of incentives, this intuition is embodied in the “efficient frontier” (Reference Gollinger and MorganGollinger & Morgan, 1993) typically shown in the risk–return space as the curve along the northwest in the top-left quadrant of Figure 4. Both investment managers and managers of companies they invest in seek to get onto that efficient frontier. Portfolios that fall below that curve are considered suboptimal. Unintended behavioral consequences ensue, as they almost always do in the face of optimal goals. Investors often interpret risk-taking as a necessary, even desirable course of action. Venture capitalists get comfortable with nine out of ten investments failing, even though the failure rate of non-venture capital funded ventures is almost the inverse of this number (Reference Nielsen, Sarasvathy, Turcan and FraserNielsen & Sarasvathy, 2018). We will look into this in a bit more detail in Section 2.5.

In a think-aloud experimental study comparing bankers with entrepreneurs, my coauthors Lester Lave and Herbert Simon and I found a set of interesting differences in how these two groups perceived and managed both financial and nonfinancial risks (Reference Sarasvathy, Simon and LaveSarasvathy, Simon, & Lave, 1998). I have summarized the results from that study as the two straight, perpendicular arrows in Figure 4. We found that bankers began with a desired level of expected return and then suggested ways to reduce risk (horizontal arrow going from right to left) through strategies such as diversification and staging of investments into smaller chunks, something akin to real options. In contrast, entrepreneurs accepted a minimal level of risk beyond which they did not want to go and then came up with creative ideas to increase return (vertical arrow going from bottom to top).

It is easy to see why bankers would feel no control over increasing returns but nonetheless believe they could control risk. Since this study preceded my dissertation that led to the discovery of effectual reasoning, it was not easy to see why entrepreneurs would feel they could change/increase the expected return on low-risk projects. It was also not easy to see why they would choose low-risk projects in the first place.

With the toolbox of effectuation, both of these results are explained: the former in terms of the pilot-in-the-plane principle, and the latter in terms of affordable loss. In other words, entrepreneurs have a worldview in which new futures are cocreated rather than predicted. That opens up ways to control and shape returns. At the same time, this also sensitizes them to the inefficacy of predictions and hence the inadvisability of placing large bets or being risk-takers in the conventional sense of the term. This explains their preference for lower risk.

At the time we completed the study of bankers and entrepreneurs, dominant theories in entrepreneurship and finance were arguing that entrepreneurs were risk-takers (Reference BrockhausBrockhaus, 1980; Reference MacCrimmon and WehrungMacCrimmon & Wehrung, 1990). On average, entrepreneurs were hypothesized to score higher on measures of risk propensity when compared with the general, non-entrepreneurial population. Bankers, according to conventional wisdom, were likely risk-averse, even more than non-bankers. Our findings contradicted conventional wisdom on both sides.

Once we move these findings into the prediction-control space, however, these findings are no longer surprising. Bankers have developed a belief in predictive control and some degree of competence in risk prevention strategies. Entrepreneurs have learned non-predictive (effectual) control, refusing to trust expected return calculations but building competence in shaping higher returns without taking much risk. That is why entrepreneurs tackling the structured tasks in the study began by choosing projects with the lowest level of risk.

Interestingly, in our study of bankers and entrepreneurs, these attitudes and approaches to risk and return carried through to nonfinancial risks as well. Without belaboring the point, effectual entrepreneurs challenge the traditional relationship illustrated through the 45-degree line in Figure 4. While venture capitalists may only make supranormal returns by taking large risks across multiple ventures, most of which tend to fail, entrepreneurs cannot invest in multiple ventures. Instead, they reduce the risks they take to their affordable loss and “make” higher returns through the effectual process.

The entrepreneurs’ only alternative to effectual cocreation of returns is to hope that their single bet (or very few bets) would somehow hit the jackpot. Expert entrepreneurs reject the latter. As one of them put it, “If you want to place a bet, go to Vegas. Starting a company is nothing like rolling the dice.”

This rearrangement of the risk–return relationship has wide-ranging implications not only in entrepreneurship and finance but also in the lives of individuals and societies. At first glance, it may seem too good to be true to think that we can systematically combine low risk with high return. To see how this can be accomplished, we need to unpack both concepts. Let us begin with risk.

The notion of risk as deviations from expectations in predicted distributions of returns is a very narrow conceptualization. In the lives of individuals and societies, risks are often multifaceted and interact in multiple ways (Reference Bromiley, Curley and YatesBromiley & Curley, 1992; Reference March and ShapiraMarch & Shapira, 1987; Reference YatesYates, 1992). For example, the same person who is risk-averse when it comes to finance may be a risk-taker in sporting activities or other behaviors such as smoking or dating. Risk-taking also changes with life stage and in response to life events.

Risk-taking varies over the lifetime of the same individual. Single young people may take risks that their parents may not in later life. Yet parents may take risks that young people ignore. In fact, just the decision to have a child entails important risks, such as hazards to the mother’s health and a nonzero probability of children with severe physical and mental disabilities. At the other end of the spectrum, empty nesters may grow more risk-averse over time. Alternatively, they could throw caution to the winds as they age and take on risky projects in retirement. So too with societies. Democracies are often born in revolutions, requiring a high level of risk-taking to overthrow tyranny. But more mature societies may become risk-averse, resulting in complacency and even apathy, leading to declines in democracy itself.

Criticisms of gross oversimplification notwithstanding, the point here is that the conceptualization of risk as volatility in expected return may be too thin a concept for picturing real-world spaces for managing risk and return. This point is important in conceptualizations of return as well. Returns too are multifaceted and multidimensional, changing over life and circumstance. In fact, a careful examination of the most impactful advances in knowledge or human well-being shows that not only is it difficult to specify returns or beneficial consequences in advance, it may also be impossible to even imagine, let alone predict them ex ante (Reference Crisp and ZaltaCrisp, 2021; Reference Langer, Snyder and LopezLanger, 2005).

History is strewn with examples of unpredictability, both in risks and returns (Reference Beatty and CarreraBeatty & Carrera, 2011). This unpredictability often takes the form of Knightian uncertainty, thwarting the very attempt to conceptualize a risk–return space in the first place.

Yet, what we need here is not to simply throw away risk–return calculations. These calculations are still useful in predictable situations. What we need is to deepen our understanding of a more pluralistic conceptualization of multiple layers of risks and returns, and how they interact with control and changing goals. Additionally, there are ecological and behavioral aspects at work in the risk–return space.

2.4 Competition and Commitment/Collusion

The concept of competition is central to understandings of free markets, whether in the neoclassical sense of perfect competition or more nuanced versions that include innovation and imperfect information. The canonical illustration of competition in strategic management is encapsulated in Porter’s five forces framework, depicted on the right-hand side of Figure 5 (Reference PorterPorter, 2008). In this framework, an industry or market is characterized not only by competitive rivalry from within but is also subject to external threats from all four sides: new entrants and substitutes, as well as the differential bargaining power of suppliers and customers. Note that this framework does not and cannot explain how markets come to be in the first place.

Figure 5 Competition and commitment

The effectual framework precedes Porter by offering a view of how new markets are cocreated in the first place by entrepreneurs and their stakeholders. There are two important points to note here. First, entrepreneurial ventures are not merely captured in the term “new entrants” as in the Porter framework. Second, the particular role of a stakeholder, whether a supplier or a customer or something else, is not given or assumed a priori in the effectual framework. Instead, the effectual process transforms a variety of readily available means into a new market through people self-selecting into stakeholder roles in the process through specific commitments. These commitments not only determine the resources coming into the new market but also explain and cocreate all five forces of the Porter framework. In this sense, the effectual process offers a framework of self-selective collusion rather than competition.

Effectuation could thus be seen as a framework of self-organization. It offers a view of how all kinds of people living their various lives can, at times, coalesce into new ventures and markets. This can happen intentionally, with entrepreneurs kick-starting the process with their bird-in-hand. But it can also occur contingently, with the process beginning anywhere in the effectual process diagram. Entrepreneurship history is rife with stories of chance encounters between people leading to ideas and actions for new ventures (Reference DewDew, 2009; Reference KoehnKoehn, 2001; Reference Shah and TripsasShah & Tripsas, 2007). Take the case of Microsoft, for instance. Although Gates and Allen were writing code in their own venture, the deal with IBM that helped them “make the big time” happened because their competitor, Gary Kildall, had blown off his meeting with IBM (Reference AllenAllen, 2012).

Whether intentionally forged or contingently cocreated, each effectual commitment resolves a variety of components, such as resources, relationships, roles, and responsibilities, that feed into new markets emerging from interactions between self-selecting stakeholders. Entrepreneurs, in this view, consist of one kind of stakeholder. The specific commitments in the effectual process result in the formulation of new entrants and substitutes, as depicted in the Porter framework. The commitments also specify roles for incoming stakeholders and relational structures between stakeholders, such as suppliers and customers. Finally, commitments determine who colludes with whom and who does not. It is, therefore, possible to undertake an effectual analysis of how and why specific industries emerge into competitive spaces of the kind we see within the five forces framework.

Traditionally, a slew of benefits have been attributed to competitive forces. Classical economists debated and discussed the benefits of the “invisible hand,” namely doing away with the need for a central authority to manage the economy (Reference SmithSmith, 1863). Under the school of neoclassical economics, these benefits were formalized through equilibrium (Reference Arrow and DebreuArrow & Debreu, 1954; Reference MarshallMarshall, 1890 [2009]) and optimality conditions, showing that given certain reasonable assumptions, free competition could ensure at least that no one got worse off (Reference Pareto, Schwier and PagePareto, Schwier, & Page, 1971). In the twentieth century, Reference Hayek, Nishiyama and LeubeHayek (1984) and Reference SchumpeterSchumpeter (1975[1942]) moved beyond neoclassical conceptualizations of perfectly competitive markets, to argue for the innovative potential of competitive markets, even when not entirely perfect.

How does a collusive or even a cocreative framework play into and alter these benefits? First, it should be noted that competition is a characteristic of markets and may be theorized in terms of the behavior of participants in existing markets, such as suppliers and buyers. This does not necessarily mean that competitive behavior is ubiquitous or even necessary in the origin of new markets. Nor does this suggest or assume any reasons for why people should behave competitively. Most importantly, competitive and cooperative behaviors interact in interesting ways even within existing markets (Reference Brandenburger and NalebuffBrandenburger & Nalebuff, 1996).

Adam Smith himself argued for a more nuanced understanding of human behavior and interaction as the basis for exchanges in market economies:

Current and potential competitors can become self-selecting stakeholders in the effectual process, colluding to cocreate new markets. As we showed in great detail and depth in Reference Sarasvathy and DewSarasvathy and Dew (2005), effectual commitments are crucial in the face of Knightian uncertainty. And by definition, it is not possible to compete in a market that does not yet exist and whose shape and boundaries cannot be specified ex ante. However, as the new market begins to take shape and coagulate into more predictable boundaries, competitive entry increases to manifest the rivalry that we see in most extant markets and in the Porter framework.

A telling example of the reality of cocreation masked by a theory of competition can be found in Brian Loasby’s historical examination of Schumpeter’s example of creative destruction (Reference LoasbyLoasby, 1999; Reference Loasby2001). In his famous chapter on this idea, Schumpeter used the example of railroads replacing horse carriages as evidence for his argument supporting the creative–destructive force of innovation. When Loasby studied the actual history of these two industries, he found that the manufacturers of horse carriages simply retooled their facilities to make railroad carriages. While it appeared that technological innovation had destroyed firms in the horse carriage industry, in reality, several of the very same firms from the old industry cocreated the new one, learning from each other to transform resources already within their control.

Why are commitments crucial to action in the face of Knightian uncertainty? The answer is that by simply agreeing to act in predictable ways, we can impose predictability. In other words, effectual commitments can sometimes transform uncertainties into reliable predictions. A simple example from daily life may suffice. For those of us who grew up and learned to drive in places like Mumbai, the daily commute can be a matter of unpredictability, even bordering on Knightian uncertainty. Yet, as I was astonished to learn during my first visit outside India to Germany, it is possible to create predictability through traffic lights, but only to the extent that people commit to heeding them.

The larger lesson here is that in several domains of human action, actual commitments between stakeholders can cocreate valuable structures that obviate the necessity to predict the future, at costs that can be affordable.

In the case of effectual entrepreneurship, this can be even more beneficial since the process is more like shaping something out of clay than traversing environments largely given and outside the control of individual commuters. Spatial metaphors, such as fields of discovery, can benefit both from commitment frameworks as well as frameworks of competition. But in the face of Knightian uncertainty, transformative metaphors involving shaping and making something new are more useful than motifs of discovery.

In fact, as I have argued elsewhere and will again in Section 2.6, markets are not naturally occurring arenas within which competition happens. Markets are artifacts shaped and formed through human action and interaction. The benefits of competition depend in important ways on humans’ ability to collude and cocreate through non-predictive (effectual) commitments. Let’s turn to an examination of how exactly these commitments come to be.

2.5 Success and Failure

In practice, entrepreneurs will sometimes assert that “failure is not an option.” This is often countered in academia with the snarky, seemingly empirical rejoinder, “most firms fail.” It turns out that both of these statements are misleading in an eerily similar way. The first is misleading because what expert entrepreneurs mean when they say “failure is not an option” is that success–failure is not a 0-1 variable. The second is misleading when it confounds the success/failure of the entrepreneur with the success/failure of the firm.

As I argued in the previous section, expert entrepreneurs learn that failure is always an option, even when they may claim in public pronouncements to the contrary. One example of reality, as opposed to retrospective claims, comes from the history of eBay:

Figure 6 contrasts the view from effectuation on the right-hand side with the standard one of success–failure as a 0-1 option. The standard depiction on the left-hand side incorporates the probability of success and the probability of failure, adding up to 1. This means that failure may be defined as not succeeding, and success can be considered not failing. While this conceptualization may be useful for certain kinds of formal modeling of performance, there is something clearly missing from it from an experiential point of view. Not merely entrepreneurial experience, but human experience more generally, suggests a more nuanced view (Reference PetroskiPetroski, 1985; Reference RobertsRoberts, 1989). Hence, our engagement with psychological variables such as persistence, grit, resilience, and so on.

Figure 6 Success and failure

Ordinary life is filled with failures, opening doors to eventual success. Instances also abound of success leading to negative consequences, ranging from simple bad behavior to debilitating relational failures. It could be argued that we need to keep definitions simple and measurable. Therefore, confounding relational and other types of psychosocial aspects of performance with simple firm performance obscures the point. While simple firm performance statistics may be the starting point (even that is not that “simple” as we will see later), several other variables are crucial for theorizing as well as teaching. Moreover, entrepreneurs cannot afford to concern themselves only with firm performance. How they think and feel about it, and relate to others through those cognitions and emotions, has a real impact on whether they succeed or fail. Hence, a deeper dive is warranted.

In order to develop an understanding of the two effectual models on the right-hand side of Figure 6, one relating to firm performance and the other to the careers of entrepreneurs, we will take a much deeper dive into the very notion of success and failure. Over the past decade, several studies in entrepreneurship have begun mapping out the landscape of entrepreneurial performance (Reference Audretsch, Santarelli and VivarelliAudretsch, Santarelli, & Vivarelli, 1999; OECD, 2015, 2016). Consider the statement alluded to earlier, “most firms fail.” In one sense, this statement is correct; most firms do not exist forever. Even in the short term, about half the firms that get started will disappear. However, that does not always mean failure. The term “exit” is more appropriate. Consider, for example, that of the 50 percent or so of firms that exit, a third are profitable at the time of exit (Reference HeaddHeadd, 2003).

Firms exit for a variety of voluntary and semi-voluntary reasons, some professional and others personal. On the personal front, these can range from the life stage of founders to job opportunities that pay more than the venture under certain macroeconomic conditions (Reference Luzzi and SassonLuzzi & Sasson, 2016) to life events such as death, marriage, divorce, and parenting (Reference Wennberg, Wiklund, DeTienne and CardonWennberg et al., 2010). Conflicts and clashes between founding team members form one of the most prevalent reasons for the closing down of profitable firms (Reference WassermanWasserman, 2012). Sometimes the firm gets restarted in another name with another configuration of some of the founders engaging in the restart. The original firm still gets counted as an exit in the data. Additionally, sometimes investors can force firm exit when they determine the need to pull out capital or break up a profitable firm for liquidity reasons (Reference WassermanWasserman, 2003).

That still leaves two kinds of involuntary exits. Sometimes firms simply abort because of a lack of demand and/or run out of cash. Alternatively, firms can go bankrupt with money still owed to creditors (Reference Lawless and WarrenLawless & Warren, 2005). In addition to a paucity of demand or capital, bankruptcies can ensue internally through mismanagement or externally due to shocks such as a new disruptive technology or regulatory changes.

In sum, the popular press overemphasizes the patently false claim that nine out of ten ventures fail. This number is true for firms funded by venture capitalists. However, venture capital funds less than one-hundredth of 1 percent of firms in the US and even lower fractions in other countries. Strictly speaking, the reverse is true when it comes to the statistics of firm failure (Reference Lawless and WarrenLawless & Warren, 2005). Only one out of ten firms is clearly a failure, with bankruptcies and money still owed to creditors.

In terms of exits more generally, voluntary and involuntary exits together account for about half of all firms in the economy. Add to that the fact that about a third of exiting firms are profitable at the time of exit, and the landscape of entrepreneurial performance looks very different from received wisdom (Reference Gompers, Kovner, Lerner and ScharfsteinGompers et al., 2006).

Now that we have a clearer picture of success/failure rates of firms, we need to examine an additional confusion that plagues the domain of entrepreneurship. This confusion occurs when the performance of ventures is routinely assumed to be the same as the performance of entrepreneurs. Effectuation consists of lessons entrepreneurs learn through at least ten or more years of immersive experience starting and running multiple firms. Expert entrepreneurs, therefore, learn to take an instrumental view of firm performance. They realize that no matter what the true rate of firm failure may be, they can control and create their own success simply by starting more than one firm. Additionally, as they learn to invest no more than what they can afford to lose in any given venture, they are able to start several. Finally, by making sure they learn and accumulate lessons for each venture and strengthen relationships with stakeholders through multiple projects and ventures, they can cocreate success that leverages failures along the way (Reference NahataNahata, 2019; Reference Nielsen and SarasvathyNielsen & Sarasvathy, 2016; Reference ParkerParker, 2013). This is a story not of persistence per se but of building and employing techniques of outliving small failures and cumulating small successes over time.

The right-hand side of Figure 6 illustrates the relationship between the success of a venture and the lessons learned by expert entrepreneurs over time and experience, including multiple ventures. Novice entrepreneurs may start with any mix of prediction and control in their decision and action logics. Preferences for effectual action could be innate in some (rare) cases, but lack of resources may drive people to use effectuation principles such as “bird-in-hand” and “affordable loss,” even when their natural inclinations are visionary, adaptive, or causal. However, through multiple ventures or projects within ventures, some of which may fail or succeed, entrepreneurs who continue through to expertise learn the principles of effectuation.

In contrast, most ventures that survive and start growing require some mix and match of both predictive and non-predictive strategies. A large proportion of entrepreneurs either do not want to or are incapable of mixing and matching strategies. Hence, they either quit the venture as it grows beyond a certain stage (mostly to start other new ventures or pursue other careers) or get fired by venture capitalists.

This explains both how expert entrepreneurs become more effectual and why large companies often do not continue to be run by their founders.

A final consideration for understanding the performance implications of effectuation involves the issue of involuntary exit due to external shocks. Both predictive and effectual approaches may be impacted by external shocks entirely out of the control of entrepreneurs. In the predictive case, the predictions could simply be wrong or falsified through external shocks. Similarly, in the effectual case, the venture could run out of the affordable loss limits of its stakeholder commitments, or these commitments could be destroyed through external shocks. In both cases, the venture could be severely or fatally damaged.

To see the overall implications for success and failure in effectuation, we revisit the effectual process in Figure 7.

Figure 7 The effectual process with performance implications (within dashed circles)

The important consequences for performance in the effectual process are more interesting than the obvious ones, increasing the probability of success or reducing the probability of failure. Instead, the costs of failure are likely to be lower in the effectual case since no one in the process invests more than they can afford to lose. In other words, irrespective of whether effectuation alters any objective probability of failure that may exist, it reduces the cost of failure nonetheless. Similarly, if and when success does occur, effectuation increases the probability of innovation, especially through the incorporation of self-selected stakeholders, none of whom needs to be prescient or brilliant in their innovative proclivities.

Most importantly, both failures and successes are inputs into the effectual process, not zero-sum outcomes. In terms of outcomes, the process either aborts due to conscious quitting (due to running out of affordable loss resources or conflicts between particular stakeholders) or it results in the cocreation of new markets and other innovative artifacts.

2.6 Finding and Making

Making is very different from finding. And making effectually is very different from making ecologically. The difference hinges on human agency. There is a tendency and a temptation to keep agentic human action out of our theories and models. That is the appeal of the exogenously given scissors of demand and supply in economics. That too is the seduction of some evolutionary and ecological models that attribute innovations to blind variation and natural selection, without involving the messiness of human motivations and intersubjective interactions such as conversations, commitments, and conflicts. An effectual approach offers practicable heuristics to move about and reshape the prediction-control space while unapologetically embracing William James’ (1916) insight, “The trail of the human serpent is over everything.”

Traditionally, both ordinary and innovative decisions and behaviors have been modeled in terms of search and selection processes. Whether considered in terms of optimal search, evolutionary models, or complexity landscapes, strategic management abounds with theories involving search and selection (Reference AngusAngus, 2019; Reference Makadok and WalkerMakadok & Walker, 1996; Reference Posen, Keil, Kim and MeissnerPosen et al., 2018). In entrepreneurship research, search models are complemented with or sometimes contrasted with models of creation. Recent discussions about the discovery and creation of opportunities offer a case in point (Reference Alvarez and BarneyAlvarez & Barney, 2007; Reference DavidssonDavidsson, 2023; Reference KleinKlein, 2008). These also hark back to Schumpeter’s ideas about creative destruction and Hayek’s arguments for competition as a discovery procedure. Given that the words “discovery” and “creation” have not only been imbued with multiple and muddled meanings but also evoke specters of debates between science and religion, I wish to use the simpler, more mundane notions of finding and making in this section.

Finding takes the world as given and is largely exogenous to action. Making explicitly emphasizes the endogeneity of the environment to action. Both finding and making rely on prior knowledge. However, whereas finding points to the feasibility of prediction and estimation about the future, making highlights ignorance and unpredictability. Additionally, while finding assumes a sharp separation between the environment and the processes of search, making emphasizes actionable levers for transforming environments and shaping new futures.

2.6.7 Political Economy

By markets, I refer to an economic organization built on free exchange between buyers and sellers. By government, I mean a centralized or hierarchical political organization imbued with coercive authority to collect taxes and enforce regulations. In the ensuing discussion, I don’t see the two as entirely independent or mutually exclusive. To function well, markets require certain regulatory guardrails. Similarly, governments require freely functioning markets to fund a viable system of law and order. Yet, these two are distinct in ways that matter. Moreover, these two distinct ways of organizing human activity constrain and bound the prediction-control space. At the same time, as effectuators navigate and reshape the space, their actions also impinge on these boundaries.

Some entrepreneurial actions may explicitly seek to reshape the boundaries between markets and governments. Private military contractors in the US are a case in point. Others may produce innovations that inadvertently disturb existing boundaries, requiring new regulations to be enacted and enforced, or old regulations to be retired. Examples of these include artificial intelligence and social media. Yet others offer opportunities for productive new public–private partnerships. The development of COVID-19 vaccines provides recent examples. Additionally, “social entrepreneurship” is a rising new field of activity and research. Social ventures seek to solve large complex social problems using practices and techniques from private enterprise. Consider examples such as Grameen Bank tackling poverty or Transparency International taking on corruption. Crises such as climate change are accelerating social entrepreneurship in domains ranging from the commercialization of renewable energy to the rehabilitation of migrants and rebuilding of disaster zones.

In other words, entrepreneurs are increasingly involved not only in building familiar for-profit firms but also in building new institutions and organizations that defy conventional notions of “firms,” “markets,” and “governments.” Therefore, it would be worth exploring how an effectual view can help us rethink taken-for-granted assumptions about the relative roles of markets and governments in the political economy.

In this connection, two related concepts from economics are worth examining in more detail. The first one is “market failure” which may be more familiar than the second one, “polycentric governance,” that comes from the work of Nobel laureate Reference OstromElinor Ostrom (2010). Both terms begin with the idea that while certain governmental regulations are required for markets to function well, markets, when they do function well, provide the most efficient form of economic organization. However, when they don’t function well, we encounter “market failure.”

An iconic example of market failure is the “tragedy of the commons” (Reference HardinHardin, 1968) where some resources (such as clean air) are free to everyone in a community. Such common pool resources require the community to nurture and manage them. However, since they cannot be broken up into private property, there is no incentive for any member of the community to incur the costs of taking care of them. In these cases of market failure, governments need to step in and take action. Or some other form of “collective action” (such as trade unions) becomes necessary (Reference OlsonOlson, 1971).

Ostrom’s argument is theoretical, albeit a compelling one. Ostrom invested decades of research into the histories of several such situations of market failure to show how small groups of individuals came together to manage them without necessitating governmental action. From the vast amounts of in-depth data from around the world, she chronicled a process called the Institutional Analysis and Development (IAD) framework that explains the development of “polycentric governance” mechanisms. These mechanisms do not fit into the institutional structure of centralized governments, whether local or national. They also transcend market transactions. Instead, they rearrange the boundaries between market structures and government structures, enabling new kinds of institutions to be cocreated. In a recent article, we reanalyzed Ostrom’s data on the allocation of water rights in the Los Angeles basins in the 1930s to show exactly how an effectual process drives the heart of Ostrom’s IAD framework (Reference Sarasvathy and RameshSarasvathy & Ramesh, 2019).

In particular, the effectual process at the heart of the IAD framework explains how both clear and ambiguous goals of individuals get reshaped into the cocreation of the new polycentric governance mechanisms. This allows us to extend the cocreation of new ends worth achieving to the level of a community as a whole. Just as studies of goals in psychology have not paid attention to where goals come from in the first place, current debates about the role of markets versus governments assume developmental and environmental goals as given and even obvious (Reference MazzucatoMazzucato, 2015; Reference MoyoMoyo, 2009). In a characteristically complex yet insightful analysis titled Development as Freedom, Amartya Sen urges us not to treat the poor and unfortunate as patients (Reference SenSen, 1999). Instead, he argues that economic development should treat people as active agents shaping their own ends. His work has inspired the formulation of the Human Development Index that measures capabilities, not only outcomes such as GDP and PCI (Reference Anand and SenAnand & Sen, 1994).

If we are to build on the works of economists such as Ostrom and Sen to help entrepreneurs develop ventures that effectively reshape the boundaries between markets and governments, we need to understand the conditions under which effectuation can drive the cocreation of public–private partnerships, as well as new organizations and institutions of polycentric governance. Figure 8 provides a framework for the role of effectual entrepreneurship when market failures occur under conditions of Knightian uncertainty.

Figure 8 Markets, governments, and effectual entrepreneurship under uncertainty

Figure 8 illustrates how effectual entrepreneurs can get going under Knightian uncertainty to help cocreate new structures for navigating the boundaries between markets and governments, whether in terms of capital allocation or collective action (Reference Sarasvathy, Wennberg and SandströmSarasvathy, 2021). The left-hand sides of Figures 8a and 8b depict the traditional view of markets and governments. In this view, government action is seen as the remedy for market failures. The view from effectuation shows that entrepreneurs can and do help shape and cocreate new markets. Additionally, they can also move the boundaries between markets and governments, especially under conditions of uncertainty where both market action and government action might have trouble getting a kick start.

The right-hand sides in Figure 8 offer at least two examples of when and how an effectual approach could get the ball rolling. Figure 8a shows the case where large capital allocations, much beyond the coffers of individual investors and private syndicates, may be called for to overcome market failure. The right-hand side of Figure 8b illustrates situations requiring more than funding, such as those involving collective action where individuals lack incentives to act on their own. In further explicating the framework in Figure 8, however, I will eschew the term “market failure” since it sets up a false dichotomy between markets and governments, which I already repudiated at the beginning of this section.

The intuition for Figure 8a is that when funding requirements are very large, states will have a larger role to play, even though they may need to be cajoled into it through effectual action under uncertainty. Renewable energy companies collaborating with state investments in R&D is a case in point. The intuition for Figure 8b has to do with the argument from the literature on collective action for the need for selective incentives to induce individuals to contribute to the common good. Take the case of COVID-19 vaccination. While some people unhesitatingly volunteer to get vaccinated, others refuse or hesitate for a variety of different reasons. Yet the benefits of vaccination depend on almost everyone in the population getting vaccinated. Therefore, selective incentives have to be developed and implemented to get everyone vaccinated.

Incentives required may be large or small, depending on specific situations. Governments can provide both positive and negative incentives, but only in situations where the consequences are highly predictable, as in the case of the COVID vaccine. However, it is harder, if not impossible, for governments to enact and implement incentives in situations of Knightian uncertainty. Issues related to climate change are a case in point here. In such cases, the type of effectual collective action, as we saw in the LA water rights case, becomes necessary and useful. Furthermore, this effectual action can move from the effectual quadrant in Figure 8 toward public–private partnerships or even result in innovative new polycentric governance institutions that can straddle and reshape the boundaries between markets and governments.

Effectual actions and interactions in the prediction-control space are not limited to firms and markets. They can help remake the worlds we live in.

In sum,

Worlds can be made. Futures can be made. But they are not made from nothing. And they are not simply made as imagined. The constraints of means currently under control are real. They are not changeable through individual wishful thinking or will, however powerfully and persuasively imagined. Instead, flexibility in goals can be the most useful tool in making the world anew, especially in opening the door to self-selection by stakeholders willing to make valuable commitments. Both clear, well-ordered preferences and inchoate aspirations are ingredients facilitating the molding of these new goals that are themselves cocreated through the effectual process. Without goal ambiguity, worldmaking can become a difficult and stultifying endeavor surrendering agency to tyranny of various kinds, while also getting battered by the vagaries of Knightian uncertainty.

The fact that every one of the social sciences has, however recently and reluctantly, found a space for making brings us to the most important artifact that the effectual process can make: the cocreation of purpose itself.

“There is no end / To what a living world / Will demand of you.” Octavia E. Butler, Parable of the Sower, Earthseed Series

(Reference ButlerButler, 2012)