Howard Dernehl documents common perceptual biases and offers suggestions for how to cultivate awareness and perceiving without bias.
Howard Dernehl: Awareness and Perceiving Without Bias
This is a guest post by Howard Dernehl, MBA. I met Howard at the first Bootstrappers Breakfast we hosted in Sunnyvale in October, 2006. Francis Adanza had met him at several entrepreneurial networking events earlier that year and invited him. He was a senior product manager at NetApp by day and a startup entrepreneur in the early morning, evening, and on the weekend. He researched this article over the last year and I asked if I could publish it on the SKMurphy blog.
Key Insights and Strategic Moves
- Successful entrepreneurs, bootstrappers and startups strive to understand the world as it actually exists. However, biases distort our ability to discern objective truths and truly understand the world.
- The mere awareness of a handful of cognitive biases and the common research biases found in investigations seeking to identify cause-and-effects enables the serious entrepreneur to avoid invalid or erroneous conclusions.
- Avoiding bias in a business setting increases the likelihood of a successful bootstrap or startup because decisions and actions are based on making reliable assessments of constraints and opportunities rather than faulty conclusions.
- Cognitive biases affect our personal understanding of things and events. In human thinking every person is susceptible to unconsciously following rules-of-thumb or making judgments that are not based on fully vetted facts or on grounded, provable assessments.
- Research biases affect the results of investigations or studies designed to better understand things and events. They affect any aspect of the business factors (pricing, quality, market penetration, customer satisfaction, etc.) that alters or skews the results or conclusions away from a truly realistic understanding of what exists or has occurred.
- Survivorship Bias: in a research study, what is observable as a sample of things or events for which a part of the representative sample is mistakenly not included, often because of an unintended sequential filter.
- Sampling Bias: in a research study, some individual members representing the full set or universe of a population are systematically more likely to be chosen in a sample than others.
- Selection Bias: in a research study, the chosen individual members representing the full set or universe of a population are systematically different from the actual population intended to be investigated. This is often caused by a confounding factor or condition that impacts the variables measuring both what is presumed to be the cause, as well as the resulting effect.
- Avoiding bias can make startups and bootstrappers more successful because with a better understanding of the world the entrepreneur’s actions can be more accurately configured to bring forth desired situations and intended accomplishments.
- This ready-to-use information to enhance your investigative and analytical skills will enable you to quickly spot common characteristics of worldview distortions and take corrective action.
Cognitive Bias and Research Bias
We are observers and actors in the world, and understanding the world about us is an interpretation, a perception, of what really exists and happens. Our interpretation is both personal and shared. We want to be aware of what is important and to accurately perceive things and events so that our interpretations are based on assessments and conclusions that are highly correlated to what truly exists. Bias is a distortion in understanding of the world as it truly is.
- Cognitive bias is a personal, unconscious observing and thinking that misshapes the correlation between what is and what we conclude.
- Research bias is an objective and methodical distortion in understanding caused by inadvertently misusing data collection methods, analytical tools and/or logic. It results in a misrepresentation of what could be sound conclusions and understanding.
If a business effort is really a hobby or a lifestyle business then the lack of competitive alternatives in the marketplace are not as impactful as if the goal is to eventually create a growing enterprise that contends against rivals. A competitive, growing startup requires an ever-increasing accumulation and balance of empowering capitalized assets working for the business: intellectual, social and financial capital. Avoiding bias can make us more successful because with a better understanding of the world our actions can be more accurately configured to bring forth the capitalized resources, situations and outcomes we pursue.
Bias is an error in thinking that usually leads to false or misleading assumptions or conclusions and therefore can be disastrous for accomplishing our goals and objectives. Avoiding bias is essential for grounding thoughts, analysis and conclusions into the workings of the real world, thereby avoiding the hazards associated of wishful thinking that exaggerates the likelihood of desired outcomes, or is overly pessimistic.
If we want to move strategically and perform powerfully and boldly for business and personal success we will need to minimize the misunderstandings and unsound assessments induced by cognitive and research biases. We need to being perceiving without bias.
Avoiding Bias for Successful Entrepreneurship
Inspiration, speculation and out-of-the-box creativity are essential for the success of an innovative startup. However, manifesting and realizing the imagined outcome is dependent on rational and realistic understanding of the relevant features and attributes of the world that is to be addressed by the business objective. Pipe dreams don’t materialize, regardless of the expended effort.
Bias is a factor that skews or misrepresents our effort to understand what truly occurs or exists, leading to distortions and omissions of relevant constraints and opportunities. Crucial to designing a product or service that will generate entrepreneurial success is the ability and practice of being aware of relevant phenomena, facts and situations; designing testable assertions to enable useful assessments; making logical inferences and deductions; and preparing compelling story examples and anecdotes. Core to these investigative, analytical and communication skills is avoiding cognitive and research biases as they distort and misrepresent what is real.
Reducing or eliminating cognitive errors clarifies and informs. Humans are inherently biased to be biased, as it is well-known that we see patterns in random configurations of data, whether through our senses or our thinking. Further, we often mistake correlation for causation.
In the face of competition from current and future vendors an enterprise effectively pursuing sales and revenue will necessarily overcome challenging market forces by producing and displaying customer value. Generating product or service value requires matching its designs, operational practices, performance-in-use and buyer’s satisfaction to a factual and grounded understanding of what truly fulfills, satisfies or supplies target market wants, needs and/or demands.
A successful entrepreneurial path is in alignment with the facts that provide the foundation for understanding what is really and truly happening and emerging in the target market for the planned product or service. Significantly reducing perceptual, experiential and analytical bias will lead to designing more applicable and valued intellectual capital, more engaging and active social capital and the means to attract or self-generate additional financial capital.
Biased personal and shared social truths (groupthink) block the understanding of objective truths which are necessarily grounded in fact, logic and reasoning. Fate rewards those who carefully observe and act in accordance with the world as it is.
Cognitive bias is personal, unconscious observing and thinking that misshapes the correlation between what is and what we conclude. Bias is a blinder or inhibitor, the lack of a neutral and through sensing and interpretation of what is really happening with the phenomena that we are evaluating.
Cognitive biases arise from the human brain’s capability for making conclusions and decisions quickly with little or incomplete information. This is of special benefit in urgent or emergent circumstances where we need to act quickly and a simple reaction or the following a rule-of-thumb produces a satisfactory result. However, where data-gathering, reasoning and logic are necessary to properly assess or handle a complex situation the lack of rigor and accuracy resulting from cognitive bias will often produce erroneous conclusions and actions.
There is a large class of cognitive bias types in human thinking, including the confirmation bias for which we unconsciously seek information that confirms what we already believe: we live in a self-reinforcing belief bubble. Interestingly, we more easily see cognitive biases in others than ourselves. Recognizing cognitive biases in oneself, business associates, information sources and competitor behaviors can help mitigate the negative effects of distorted or skewed conclusions and behaviors. Some cognitive biases include:
- Confirmation bias: we respond more favorably to information that supports our current beliefs than information that undermines them.
- Optimism bias: we unrealistically presume that we will be more successful that what is likely.
- False consensus bias: we think that many others are in agreement with us than is actually so.
- Anchoring bias: we are dominated by the early information we receive on a subject and de-emphasize additional or contrary information that we later acquire.
- Ego bias: we tend to credit ourselves when desirable outcomes occur while we fault external forces when we achieve poor results.
- Functional bias: we overlook alternative uses for the presumed utility of tools and technologies and thereby miss opportunities to repurpose ideas, concepts or objects to produce new capabilities and consequences.
- False expertise bias: we mistakenly believe ourselves to have expertise in a domain that is outside of an area for which we have achieved certification, recognition or professional success.
Human thinking runs on sensing, awareness, perceiving, reasoning, logic, emotion, moods, judging, rationalization and bias, among other factors. Startups and bootstrappers are particularly susceptible to cognitive and research biases first, because of the founder’s inclination to create technologies that are highly innovative, impactful and disruptive, and second, because the new-product validation procedures typical in larger enterprises are minimal in small firms. Cognitive bias with its inadequate scrutiny and over-simplified reasoning seldom contributes to clear, effective thinking and ultimate success. Rather, by avoiding biases we remove a vulnerability to startup success. To better understand the world, strive be self-aware of your own biases and look for the biases with whom you interact.
Research bias is any factor or approach in the process or practice of investigation or evaluation that alters or skews the results or conclusions away from a truly realistic understanding of what exists or has occurred.
In statistics and research, to determine the factors enabling business success, there are many types of biases that can distort a true understanding of cause and effect. Biases often sneak in due to the complexities of sampling, measuring, identifying correlations of variables and possible casualty. Additionally, our interpretations are shaped by our natural striving to make a novel conclusion or justifying an action. There is often a complexity of biases and distortions affecting the gathering of information and making subsequent conclusions. The common research biases are sometimes mixed and layered within an investigative effort, and there is not a broadly accepted and definitive taxonomy. Nevertheless, understanding the key concepts is straightforward and will produce significant business value for entrepreneurs.
In an unbiased sample of things or events the individual members of the population being studied are randomly chosen from the full population set, and none of the individuals selected fall outside the full set intended to be studied. As a result the average value and degree of distribution between the sample and the entire population that they represent results only from chance. Further, with a sufficiently large sample, depending on the size of the entire population, the precision and accuracy of the measurement can be closely estimated so that the sample properly represents the whole population.
Survivorship bias, sampling bias and selection bias are explained and methods for avoiding their hazards, including false conclusions, are provided.
Illustrations of Survivorship, Sampling and Selection Biases
Survivorship, sampling and selection biases are categories of research biases to avoid or mitigate. Dodging the biases and mistaken conclusions will bolster high-impact accomplishments by accurately interpreting real world phenomena pertaining to market segments, buyers’ propensities to purchase, target market size and other important factors and attributes. These biases can impact informal business-oriented inquiries or rigorous research studies:
- Survivorship Bias: Survivorship bias is caused by looking at what is observable as a sample of things or events in which a part of the representative sample is mistakenly not included, the dropouts often filtered by time.
- Sampling Bias: some individual members representing the full set or universe of a population are, for any reason, systematically more likely to be chosen in a sample than others.
- Selection Bias: the chosen individual members representing the full set or universe of a population are systematically different from the actual population. This is often caused by a confounding factor or condition that impacts variables for both what is presumed to be the cause as well as the effect.
A remarkable illustration of survivorship bias and its correction during WWII shows how the plans for reducing attacks on American airplanes was achieved by a reversal of the original recommendation.
In order to reduce shoot-downs of bombers US Army personnel compiled the locations of bullet holes of all the aircraft that they were able to be examine, those that were able to return to their airbase and thereby be available for study. The compiled data were plotted on a diagram representing the shape of the airplane to see where the damage most often occurred and this composite was used to indicate that the bullet-hole locations, mainly on the wings and tail, should be reinforced to reduce casualties. However, an insightful researcher, Abraham Wald, a mathematician and economist, pointed out that the planes examined were those that had returned to base and therefore survived the enemy bullets, whereas the planes that crashed were not studied and therefore must have been hit in the parts of the planes that were not recorded—mainly, as it turned out, in the area of the engine. The ultimate conclusion was that the planes should be reinforced in the area of the engine rather than the wings and tail—not where the surviving planes were hit.
This illustration, which indicated that in this instance not just a contrary assessment but the exact opposite conclusion, represents the survivorship bias (aka survival bias) in its most dramatic form. Most other survivorship bias situations represent conditions for which a contrary or invalid conclusion is affected by the survivorship bias but not the opposite conclusion. There are many survivorship bias examples in the world of business. This bias can distort understanding of the size and behavior of an enterprise’s target markets and for properly assessing the viability of the features and attributes that comprise successful products and services. Startups and bootstrappers are particularly susceptible to cognitive and research biases because of the founder’s inclination to create technologies that are highly innovative, impactful and disruptive—and because the new-product validation and other cross-checking procedures typical in larger enterprises are minimal in small firms.
Survivorship bias is caused by looking at what is observable as a sample of things or events in which a part of the representative sample is mistakenly not included, the dropouts often filtered by time. Survivorship bias describes the propensity to see only the things or events that have passed an unrecognized or unintended selection condition. In the airplane example the crashed planes that did not return to the airbase were originally not considered for the analysis and in that instance should have been the whole of the sample, as the crashed planes represented the evidence of the problem to be solved.
For a survivorship bias example, consider that if a business studies the feedback of its customers through a survey for what they like and don’t like about the product or service and make conclusions regarding their target market they are only considering the customer sentiments of those that reported their assessments of satisfaction and are ignoring those that did not survive the full customer journey. The complete journey would transpire from (A) prospective customers in the total potential market being aware of the product or services through all the subsequent steps of the buying cycle, including (B) candidate customers considering purchase, (C) actual customers having completed the purchase, (D) implementers and users of the product or service, through the last step of (E) customers responding to the inquiry. In this example the intended sample of prospective customers in the target market is winnowed at each stage of the customer journey so that the ending sample is reduced to only those prospects that became customers, users and ultimately survey respondents. After the survivorship filtering the sample is no longer representative of the intended population representing the target market.
This survivorship bias sampling problem would apply to a startup or a bootstrapper and could adversely affect the ability to stimulate early customer growth or achieve other intended business outcomes.
Another example of survivorship bias relates to the popular injunction espoused by some motivational gurus and advisors wanting to whip up audience excitement rather than encourage practicality—give hope rather than educate in a manner to truly enhance the probability of success. A pitfall for simply following one’s passion to achieve business success by considering examples of accomplished entrepreneurs is caused by focusing on the notable successes, and ignoring the failures. A list might include Walt Disney, Oprah, Edison, Jerry Seinfeld, Bill Gates, Michael Dell, Mark Zukerberg, Ray Kroc and the McDonald brothers, J.K. Rowling and so forth. They all followed their passion to success, often eschewing risk-avoiding choices and instead shooting-for-the-moon. However, the examples do not include all the people that failed through taking the gamble of solely following their dreams, rather than grounding a practical career or probable venture based on targeting one’s combinations of real skills/knowledge, genuine interests/enthusiasm and expected business/market value. In this case we are looking at the one-in-a-thousand or one-in-million highly-observable survivors that achieved remarkable business success. Examples of exuberant and uncommon success abound in glorious occupations like athletics, modeling, professional acting, musicianship, the fine arts, famous experts, CEOs of successful consumer products companies and other celebrities.
To overcome survivorship bias, continuously consider in what ways the sample for which you consider to be representative does not so truly reflect the whole set of events or things you are striving to assess. You may be well-advised to continuously strive to avoid survivorship bias just as you would aim for continuous improvement in other personal and business actions.
Survivorship bias contributes to startup and bootstrapping missteps when the entrepreneur is not well-experienced in the discernment and management for the ideation, design, development, market validation, production, launch and scaling of products or services. This is because they are, as an outsider to the methodology and standard practices for the product/service lifecycle and the customer journey, especially prone to primarily see the successes in the world which are filtered from the unseen, mundane failures. The common blunders are hidden among the successful enterprise narratives and consist of rejected ideas, abandoned conceptual designs, discarded prototypes, failed customer tests, non-scalable production attributes, uncompetitive feature sets and unprofitable cost structures. The missteps and failures are not visible as they did not make it to a successful growth stage.
Survivorship bias distorts many of our perceptions and assessments about the world along many variables, including the classic faster-better-cheaper tradeoff which posits that as one or two of these three variables are increased the other(s) must necessarily be reduced. Given that this product management theory is generally accepted as valid, while product or service success is presumably based on strategically choosing the values of the three factors among an innumerable possible frontiers of optimizing choices, our assessment of speed, quality and cost relating to existing products or services might be distorted by survivorship bias.
For example, if we look at aged products from a past era we might suppose that they were made of a higher quality than they actually were, if only because those that we are likely to observe are those that have survived because they were made to a higher standard, while those that were made to a lesser or more normal standard of quality were thrown away or recycled.
It’s even possible that the surviving samples actually represent brands or manufacturers of lesser quality rather than higher quality, the opposite of the expected. This can be shown if we understand the quality of products and services is to be properly evaluated by not just the mean of the quality measure but also by the variance.
Regarding food and beverages, hamburgers and beer, for example, McDonalds and Budweiser are by far the best selling brands. Both are known for their consistent output. McDonald’s hamburgers are made following strict systematic processes and procedures so that they taste the same anywhere in the world. Similarly, for Budweiser, the output of the 12 breweries are continuously and rigorously monitored to ensure a consistent, undifferentiated taste and flavor among the output of the production facilities. However, no one would say that these are the best-tasting or most desired products in their categories. Rather, the average quality is mundane but the variance is extraordinarily low. For most (really all) products and services there is some expected variance from the average quality level, and the lesser the better.
For example, speaking of durable products, rather than consumables, if deviation from the mean indicates that both higher- and lesser-resilient units will be produced and distributed, those that are substandard, unfit, blemished or failing are more quickly removed from the set of those existing than those of higher quality. Thus the manufacturer or service provider with a higher variance will produce more exemplary product units or service delivery events, yet the overall outcome is of lower quality because of the mix of units of inferior quality. These surviving units or events will have positive and noteworthy outcomes signaling, incorrectly, as shown in this example, that the producer with the actual lower quality (higher variance) has the falsely-evaluated higher quality—the opposite of what is true is perceived.
An example might be a device, component or service maintenance/repair that has a typical lifespan of, say for illustration, five years. The “Vendor A” that reliably produces an item to this standard of actually lasting, then failing very near this five-year timespan produces a product that can be reliably replaced, repaired or serviced at a much lower cost than the “Vendor B” product for which the unpredictability causes half of the failures to occur at, say, four years and half at six years. Each has a lifespan of five years, with “Vendor A” providing minimal variance. The low-variance lifespan is of value and a measure of overall quality. However, at five-and-a-half years nearly all of the high quality products/services from “Vendor A” will have failed, while only half of the inferior products/services from “Vendor B” have faltered. A survey looking at the products or services extant at a point-in-time at about five-and-a-half years would only observe those from “Vendor B.” It would be erroneous to conclude that the maker of the only items that remained for five-and-a-half years produced the superior product, as the opposite is true.
Not a survival bias, but an example of a bias from over-aggregating a measure can occur in customer reviews of products or services. Regarding online reviews it is sometimes said the most dissatisfied customers are more likely to complain and perhaps to write them, which is a bias in itself. Regardless of whether customers most often speak up when they are annoyed or disappointed, most of the online product reviews seem to have a high average star-rating. A higher rating in some instances might be a false sign of quality. More telling might be to double-click on the distribution of the rating scores: For two vendors with the same rating of four stars the one with the highly concentrated around four-stars is more likely to produce satisfaction than the one with lots of five-stars and a bunch of 1-stars in the mix that also averages to a four-star rating. So, like the previous example, the broader distribution from the mean indicates a lower level of quality. Averages do not exist in the real world, they are an assessment, an abstract interpretation of something that varies: if you are in a desert sweltering in the day’s heat and freezing at nighttime, the average temperature might very well be comfortable, but it doesn’t exist.
Speaking precisely, sampling bias occurs when some individual members representing the full set or universe of a population are systematically more likely to be chosen in a sample than others. This skewing of the data due to a lack of randomization when choosing from the full set of individual occurrences will possibly cause the conclusions of the analysis to be invalid.
Survivorship bias, self-selection bias and under-coverage bias are examples of sampling bias. A proper research study, one enabling sound conclusions, is unbiased and depends on randomly choosing a sample of individual members from the whole group/population to provide the dataset for subsequent application of valid logic to produce sound assessments. Sampling bias inhibits the random determination of individuals for the dataset thereby disabling meaningful conclusions.
Self-selection bias, actually a sampling bias, occurs when the objects of the study are chosen with influence by people involved with the study. A simple example would involve the individual members to be studied based on the dependence of volunteers for the study. People or businesses with specific characteristics might more probably agree to participate in the research study. For example, if you sent a survey questionnaire to all, or a random sample of, the startups listed in Crunchbase asking them whether they were satisfied with their venture capitalist or angel investor(s) for a news article with the results to be widely published, one might presume that only those startups that were satisfied and willing to report their experience would provide information because those that expressed dissatisfaction would probably suffer consequences from their funding source—perhaps reducing the probability of further financing or having additional restrictions placed on the decision-making power of the management team.
Self-selection bias is an example of under-coverage bias because some individual members of the study are naturally excluded from the data set. However, there are additional possibilities for under-coverage bias that do not represent self-selection bias. For example, an online survey might under-represent those that have online access. Also, a survey of nascent startups listed in Crunchbase might only retrieve those that Crunchbase has discovered while some firms that are working in stealth mode are not disclosing their organization or founders, and these hidden ventures might have different characteristics from the average, such as the founders having significant past startup experience or unreported angel funding. These characteristics, taken as independent variables for predicting success, might skew the results for correlating with the dependent variables representing the measures of meaningful business success over time, such as longevity, attracting financing or exiting the startup stage through acquisition or becoming publicly traded.
Speaking precisely, selection bias occurs when the chosen individual members representing the full set or universe of a population are systematically different from the actual population. (Sampling bias might also be present, too.) Selection bias is often caused by a confounding factor or condition that impacts both what is presumed to be the cause as well as the effect, the cause being the independent variable and the effect being the dependent variable. This skewing of the data will possibly invalidate the conclusions of the analysis.
Here is a made-up example that illustrates the definition of selection bias. Let’s say that you are developing a phone app and can only readily develop it for just one platform, based on a business constraint, either for the iPhone or for Android-based phones. Further, let’s suppose this new app has features and attributes that make it of much more value and marketable to the top 20% of people according to wealth and income, and that the products or services that it promotes, because of the need to establish business relationships because of some business condition or restriction, are localized to the size of a mid-sized city, say about 50,000 to 150,000 population. This phone app that provides more value for higher income individuals might promote services like wealth management, high end durables like luxury vehicles or other consumer categories favored by those with higher than typical discretionary budgets or disposable income. Let’s say that you search on the Web and read a presumably reliable article stating that overall “The median iPhone app user earns $85,000 per year, which is 40 percent more than the median Android phone user with an annual income of $61,000.” Because development is constrained to be viable for only one platform you preliminarily decide that the iPhone should be chose as those users have the higher level of income.
Next, let’s say that for the initial launch you are considering designing, configuring and establishing retail businesses relationships among several mid-sized cities considered to have Silicon Valley is high income levels, such as Palo Alto, Mountain View and Sunnyvale. You want to verify the level of iPhone deployment compared to Android among residents to verify a relatively high level, since it might reasonably vary a bit among locales. You are able to get active monitored phone use data for the city of Palo Alto during business hours and a reliable and trusted data services firm is able to calculate the average level of iPhone use compared to Android. This data services firm is known to reliably sample data to avoid sampling bias and the number of samples is well over a thousand which is considered by experts to ensure a sufficiently randomized sample and avoid sampling bias to avoid erroneous measurements. You know that the Web article regarding income levels provides valid data and sound conclusions that can be trusted, and you expect the local test for Palo Alto to generally validate that Palo Alto has a relatively proportion of iPhone users.
Given reliable reports that while the world market share for Android is about 85 percent and iPhone 15 percent, in the US the distribution is about 51 percent for Android and 48 percent for iPhone and 1 percent for others. Because the three Silicon Valley test cities are known to have higher median income than the average US user base you suppose that the higher income level, which is correlated to a relatively higher proportion of iPhone use compared to Android, will show a meaningfully higher percentage of iPhone use than the US average of 48 percent. Let’s say that the trusted data services firm produces a well-sampled report indicating 68 percent iPhone usage for Palo Alto and 64 percent in a similar study for Sunnyvale.
Next, suppose that the calculated results for Mountain View from the trusted data services firm are contrary to the measures for Palo Alto and Sunnyvale: nearly the inverse of what was reported overall on the presumable reliable Website, that the test results indicate Android users are much more common than iPhone users, the opposite of what would be expected!!! Given that there was no sampling bias, what went wrong with the local study?
The local study was composed of chosen individual phone users that were supposedly representing all the typical phone users, but they were systematically different from the actual population. The phone users in Mountain View may likely be the higher income individuals using iPhones, however, the users monitored during business hours might have disproportionately been Google employees, as Mountain View is the Googleplex headquarters for Alphabet and Google, which owns their favored Android technology.
The erroneous conclusion can be interpreted to be caused by a confounding effect that impacted both what is presumed to be the cause, high personal income level, as well as the effect, high relative usage of iPhones. The confounding condition was possibly generated many Google workers who were tracked by the local data services firm, and Google employees are typically well-paid and are loyal to their company’s Android technology. Because of the big influx of Google workers during the duration hours of the data collection the study population was not representative of the target population. In other words the participants in the research differed systematically from the population of interest and the baseline for Mountain View with the inflow of Google workers was selection biased as a result. In this case the condition caused a reversal of the proper assessment of the phone use for Mountain View residents, if only because most Google workers live in other nearby cities, and most Mountain View residents don’t work at Google. The skewing of the data by the confounding effect would cause an erroneous conclusion.
Awareness and Perceiving Without Bias for Entrepreneurial Success
Avoiding personal biases, shared cognitive biases and research biases can make your startup or bootstrap enterprise more successful, because with a better understanding of the world your actions can be more accurately configured to bring forth your intended situations. A growing startup requires an ever-increasing accumulation and balance of intellectual, social and financial assets and resources to effectively compete. This growth of capital working for your enterprise depends on success in product/service design, market acceptance, operations, customer satisfaction and other crucial business factors, which can be better understood through unbiased research and reasoning.
We are all susceptible to cognitive biases as they are part of the normal and generally beneficial human thought processes. Keep the list of cognitive biases and look for them within yourself whenever making significant and impactful evaluations, assessments or decisions. Bring a critical eye to any research or study, looking for survivorship, sampling or selection biases, and whenever you avoid instances of bias you increase your intellectual and social capital, leading to further growth of your firm’s monetary resources. See through cognitive biases and recognize research biases. Fate rewards entrepreneurs that carefully observe and act in accordance with the world as it really is!
Related Blog Posts
- David McRaney: Survivorship Bias
- Orienting, Observing, Doing Homework, and Paying Dues
- Serious and Competent People
- Failed Fast, Now What?
- Debugging Your Startup Requires Peace of Mind
- Seeing The Elephant: The Entrepreneur’s Challenge of Integrating Advice
- Constructive Pessimism
- Rules of a Scientist’s Life, Applied to Entrepreneurs
Image Credit: “Vintage Compass Rose” images licensed; © Claudia Mora