In the business world, decisions are rarely made based on complete information. There are many different factors that are necessary for decision-making, such as market and competitor trends, customer needs, and internal company situations, and these are constantly changing. Therefore, it is extremely important for business leaders, especially managers and department heads, to be able to make rational and effective decisions even with incomplete information.

What is the difference between people who can make optimal decisions with limited information and people who cannot? It's not about having advanced analytical skills or special expertise. Successful business leaders have one thing in common: quick understanding of information and keen insight.

In this article, we will introduce you to some key points for making rational decisions with incomplete information. The points introduced here are in no particular order or priority. By learning just one of these skills, you will be able to make more accurate decisions in business.

first

Making business decisions based on incomplete information is difficult. You may rely on your intuition or make decisions without checking the quality or credibility of the data. This can lead to biased and incorrect decisions. This chapter explores the illusions surrounding the use of data in business, the biases associated with intuitive decisions, and the biases associated with using data to make decisions.

Illusions surrounding data utilization in business

Here are two typical examples of illusions that are often held when it comes to data utilization.

one,"You have to be a math specialist to make decisions using data.” is an illusion. Many people believe that using data effectively in business requires advanced math skills or a degree. Additionally, some people may find it difficult to understand and explain data to others. However, you don't need to be a math specialist to make decisions using data.

the other one is,"By using big data, we should be able to make "perfect" decisions." This is an illusion. But this is also false. In fact, there are many risks and points to note when making decisions using big data. For example, big data often contains missing data or unrelated data, and data cleansing and processing are required before it can be used. In addition, there is a risk that data may be biased or misused, so a certain level of literacy and knowledge about privacy and security are required to utilize it. The modern challenge in making informed decisions is not a lack of information, but the ability to use it wisely.

These illusions surrounding the use of data in business are hindering the use of data to make rational and optimal decisions. The first step to overcoming this is to recognize that data is not a magic tool, but a tool that can be useful if used correctly. And understand that this doesn't require advanced math skills, but rather a certain level of literacy and knowledge about privacy and security.

Bias due to intuitive judgment

Intuitive bias is a thinking error that results from relying too much on feelings, instincts, or rules of thumb. This method is also called KKD, which is an acronym for the Japanese words for ``experience'' (KEIKEN), ``intuition'' (KAN), and ``courage'' (DOKYOU). This may result in bias.

Biases caused by using data to make decisions

Data bias is an error in analysis caused by bias or lack of data. These biases can affect the analysis of your data and the interpretation and conclusions you draw from your results.

Avoiding these biases and making the best business decisions based on incomplete information requires balancing intuition with data, logic and creativity. In this article, we will introduce the skills and processes necessary for this purpose.

ability to ask questions

As we grow up, we are often taught that the answer is more important than the question. And even in work, the value of asking questions is often underestimated. Once an answer or solution is found, few people may decide to ask further questions.

However, formulating questions is an important skill not only for decision-making, but also for communication and learning. Of course, simply formulating questions is meaningless. Questions with unclear purpose and direction are a waste of time and may not yield useful information.

A well-known method of asking questions is the Socratic question-and-answer method. This is a method of stimulating tastes through dialogue, proposed by the ancient Greek philosopher Socrates, but it can also be used in business.

There are many different types of questions that can be asked in the Socratic method, but we will introduce some examples of how to use it in business.

• Clarification question:What is the specific goal of this problem? What problem are you trying to solve?
• Prerequisite questions:What do we not know about this issue? What assumptions are made that this proposal will be successful?
• Reasoning question:Why do you think the answer you have now is factual? Is there any data or evidence to support this proposal?
• origin question:Where did you get your current thinking from? Who first pointed out this problem?
• Point of view or point of view questions:How would others respond to this problem? How do you think your competitors would respond if they were in the same situation?
• Hypothetical question:What are the possible answers other than the current answer? How will the plan change if things don't work out as expected?
• Implications and Consequence Questions:What does this mean and what effect will it have if I try it? What risks are involved in implementing this proposal?

By asking these kinds of questions about a problem before making a decision, you can clarify the problem and gain a more comprehensive understanding of the situation. As a result, you can formulate a more optimal hypothesis for your problem.

The ability to ask questions not only prevents information bias, but also stimulates critical thinking and opens up new perspectives and possibilities.

Problem setting ability

"If I had just one hour to save the world, I would spend the first 1 minutes finding out what the problem is and the remaining 55 minutes solving it."

albert einstein

Before checking data and information, it is extremely important to correctly understand problems and issues and prioritize solutions when making decisions to solve problems. However, determining the essence of a problem or issue is not easy. This is where the skill of problem-setting ability comes in handy. Ability to set goals is the ability to figure out what to do on your own. By acquiring problem-setting skills, you will be able to get to the essence of problems and issues and find effective solutions.

Identifying the issue starts with the purpose (what you want to achieve)

Many companies have large amounts of data and believe that there must be something valuable in it. However, even if you blindly analyze such data, you often end up in a situation where you do not know how to use the analysis results.

It is important to clarify what you want to make a decision about, that is, the purpose of data analysis. If the purpose is not clear, the scope of the data analysis cannot be clarified, and there are many cases where the analysis ends up yielding no useful suggestions despite spending a great deal of effort and money.

Once you have clarified your purpose, there are two main approaches to identifying the issues that need to be solved to achieve that purpose.

① If you have past data and can predict what the problem is: Identify from actual data

For example, let's assume that your objective is to maximize sales.

Sales can be broken down into "new sales" and "existing sales." By further breaking down each element, you can identify the elements that make up sales.

Once you have identified the elements as shown in the diagram above, start comparing the data starting from the top of the diagram.

For example, if you see the results shown in the diagram above, if you compare "new sales" and "existing sales" that make up sales, you will find that "existing sales" are stable while "new sales" are decreasing. You can see.

Next, we will compare "CVR", which constitutes "new sales", and "inflow".

You can see that while "CVR" is stable, "inflow" is decreasing.

Similarly, if you compare "display ads (inflows from)" and "listing ads (inflows from)" which make up the "inflows" that are decreasing, you can see that the number of inflows from listing ads has decreased. You can see that it is.

② If there is no past data and it is not possible to predict what the issues will be: Create a future hypothesis and identify possible future issues.

If there is no past data and it is not possible to predict what the issue is, start by creating a hypothesis for the future (Figure 1 above). Creating a future hypothesis means specifically imagining the future situation (vision) that your company is aiming for. For example, when creating a new business, we think about what kind of products and services we want to provide, what kind of customers we want to serve, and what kind of value we want to create. By creating a hypothesis for the future, you can clarify your vision and direction.

Next, check to see if there are any past facts that can be used as reference in situations or examples similar to the future hypothesis (Figure 2 above). For example, examples of companies that have succeeded or failed in the same industry or market, or similar products or services may be helpful.

Then, refer to case studies to infer the causal relationship between past facts and results (Figure 3). For example, consider what factors influenced success or failure, and what kind of reactions there were. The issues identified here are the anticipated issues for your company, and the issues that have a greater impact can be considered higher priority issues.

This is how to formulate future hypotheses and identify challenges. As a concrete example, consider the following case.

• Future hypothesis (vision)
  • In 2025, we will provide Japan's most popular online English conversation service.
• Past facts (data) close to the hypothesis
  • For example, refer to examples of success and failure of online language training services in countries other than Japan.
• Inferred causal relationship
  • Possible success factors: teacher skill and diversity, course flexibility and pricing, customer support, etc.
  • Possible reasons for failure include: technical glitches and delays, teacher shortages and turnover, and low customer satisfaction and retention rates.
• Identifying issues
  • Of the causal relationships inferred above, it was found that ``course flexibility and fees'' had the greatest influence. (=It becomes a high priority issue)

In this way, identifying issues by creating future hypotheses is a useful method when it is not clear what the issue will be.

Of course, this method isn't perfect. The future is often unpredictable, and the causal relationship between past information and results is not necessarily correct. However, using this method to clarify your ideas and hypotheses, identify possible issues and risks, and consider solutions and verification methods is important for making rational decisions with incomplete information. This is a step.

The above two approaches to identifying issues are located in Step 2 of the 8 steps that form the basis of data analysis in business, which our company XICA focuses on.

If you would like to learn more about the 8 steps of parallel data analysis, we have prepared a knowledge sharing document for you.

Free downloads of related materials

The 8 Essential Steps for Turning Data Analysis into Business Results
~What are the important points to keep in mind when analyzing data that can be used in business settings? ～

Download Now

Next, I'd like to introduce a slightly similar, but complementary, approach to the two approaches I described earlier. This approach is useful for identifying what information is ideal for making rational decisions.

IWIK, a framework useful for setting issues

There are various frameworks for finding problems (As is / To be, 5-Why analysis, etc.) to improve problem-setting skills, but the IWIK ("I Wish I Knew") process is also effective. IWIK is a problem-setting method based on the simple question, "What do I need to know to make the best decision for my purpose?"

IWIK is led by Christopher Frank (vice president of the global advertising and brand management team at American Express), Paul Magnon (head of global alliances at Google, previously at Deloitte and IBM), and Oded Netzer (Columbia University Business This is the problem setting process proposed by Professor and Associate Director of Research at the University of California, Columbia Data Science Institute.

The IWIK process has four steps:

• STEP1. Ask a question:First, ask yourself and your team the IWIK question: What do I need to know to make the best decision for my purpose? For example, if you are deciding whether to launch a new product, ask the question, "What do I need to know to make the best decision about launching a new product?"
• STEP2. Brainstorm:Next, brainstorm the IWIK questions. Write down everything that comes to mind, regardless of whether you have past data or whether information is available. For example, we identify everything we can think of, such as the market size of a new product, customer needs and preferences, competitor trends, product production and sales costs, risks and opportunities.
• STEP3. Summarize:Next, compile a list of the information you need to know and organize it into categories. For example, type of information (quantitative/qualitative), source of information (internal/external), importance of information (high/low), credibility of information (high/low), difficulty of obtaining information (high/low) Possible categories are:
• STEP4. Consider:Finally, review the information you need to know and use scoring and ranking to prioritize the information you need to make a decision. At this time, we will also delete unnecessary information and merge similar information. For example, you can prioritize and organize as follows.
  • Example)
    • The most important information for making a decision: customer needs and preferences, market size, competitor analysis
    • The least important information in a decision: cost and risk analysis
    • Information that is less difficult to obtain: internal data and customer feedback
    • Information you need to investigate further: External data and competitor analysis

By following the IWIK process, you can organize problems and focus on the essential information, avoiding situations where you waste too much time and effort on analysis and delay decisions and actions.

Source: This simple question will help you make better decisions | Fast Company | https://www.fastcompany.com/90808070/this-simple-question-will-help-you-make-better-decisions

Get Better Decision-Making With The I-Wish-I-Knew Tool | Chief Executive | https://chiefexecutive.net/get-better-decision-making-with-the-i-wish-i-knew-tool

Ability to question data

Assessing the data and its reliability

Accurate and reliable data is essential for business decision making. However, the data is not always correct. Unreliable data can lead to incorrect conclusions and decisions.

For example, it can be biased, outdated, or manipulated due to collection and processing methods. Therefore, it is important to verify the reliability of data before using it. To verify reliability, you can ask questions such as:

• Where does the data come from? Is it a reliable source?
• How was the data collected? Are there any biases or errors in the collection method?
• How was the data processed? Are there any inappropriate operations or changes to the processing method?
• Is the data up to date? How long is the data valid?
• Does the data have sufficient quantity and quality? Are there missing values ​​or outliers in the data?
  

By checking these, you can objectively judge the reliability of the data. Unreliable data is likely to negatively impact decision-making and should be treated with caution or alternative data should be sought.

Give context to your data

Data does not simply exist as numbers or symbols; it has meaning within a certain context or background. To properly understand data, it is necessary to understand its context and background. Context or background includes the phenomenon or problem that the data represents, the people or organizations to which the data relates, and the factors or conditions under which the data is influenced. Understanding context and context allows you to assess the validity, importance, relevance, and impact of your data. To get context and context, you can ask questions like:

• What does the data represent? What phenomenon or problem does the data show?
• Who is the data important to? Who are the people and organizations to whom the data relates?
• How is the data affected? What factors and conditions affect the data?
• How does data make an impact? What people, organizations, phenomena, and problems does the data impact?
  

Reviewing these provides a deeper understanding of the context and background of your data. Data that is taken out of context is likely to mislead and bias decision-making, so it must be treated with caution.

Intuition for data

"It's better to be generally right than exactly wrong."

john maynard keynes

In order to make rational decisions with incomplete data, it is necessary to develop intuition for data. Data intuition is the ability to quickly judge the meaning and validity of data. By developing this ability, you will be able to utilize data effectively without being confused by incorrect data. The following methods are effective for developing this ability.

Determine the accuracy of information needed for decision making

Decision-making involves various stages, but among them, information gathering and analysis are the parts that take the most time and effort. However, not all decisions require perfect information and analysis. For many decisions, a rough approximation of information is sufficient. If you demand too much precision, you may end up wasting your time and effort.

The required precision depends on the purpose and impact of the decision. Generally, the greater the purpose and impact of a decision, the greater the level of precision required. Conversely, the smaller the purpose or impact of a decision, the lower the level of precision required. For example, if the purpose of decision-making is "determining the company's long-term strategy," the required accuracy will be high, but if the purpose of decision-making is "determining where to have lunch," the required precision will be low. Probably.

By determining the required precision, you can adjust the method and scope of information collection and analysis. The higher the precision required, the broader and deeper the methods and scope of information gathering and analysis will be. For example, if the purpose is to ``determine the company's long-term strategy'' as mentioned earlier, it may include market and competitor research, customer and product analysis, marketing analysis, financial and budget calculations, etc. If your goal is to decide on a restaurant, it is sufficient to look up the menus and prices of nearby restaurants, and look at reviews and ratings.

Use your senses to determine if the data is valid before diving into complex analysis.

As mentioned in the "Giving context to data" part, data does not exist on its own; it is always based on some context or background. Understanding the context and background of your data provides criteria for determining its meaning and validity.

Once you understand the context and background of your data, it's time to use your senses to determine if it's within reason. A reasonable range refers to whether the data is close to expected or typical values. For example, whether the data is close to the overall mean or median, and how different it is compared to past data or other data.

When comparing data with other data, it is desirable that the data be of the same type or category. For example, sales can be compared with sales from other periods or regions. When making comparisons, use things like percentages and percentage changes to evaluate the relative size and change of numbers. For example, if a region's sales numbers are up 50% year over year, is that within a reasonable range? Compare it with other data to determine whether the numbers are out of line.

If you have no data to compare with and can only estimate whether the data is reasonable, consider the rough values ​​and ranges of the data based on your own knowledge and experience. For example, if the data is about the number of attendees of an event being held for the first time in an area, estimate the number based on the target population of the area and the expected attendance rate, and check whether the number is within a reasonable range. When making estimates, it is important to be careful of overconfidence and bias, and to have evidence and reasons for your estimates.

Know the key points in decision making

As with any decision, you should check the following before making your decision:

• What is the purpose of the decision?
• What is the deadline for making a decision?
• Who is involved in making decisions?

It is important to obtain stakeholder agreement on these matters in advance. This is because they greatly influence the outcome and speed of decisions. For example, if the purpose of a decision is vague, the criteria and direction will become unclear. Improper decision deadlines reduce speed and efficiency. When the wrong people are involved in decisions, there is a lack of responsibility and commitment.

It is also important to consider the following before making a decision:

• How long does it take to make a decision?
• How much risk is involved in the decision?
• How much confidence do decisions require?

These are contents that can affect not only the decision itself, but also the organization. For example, taking too long to make a decision can lead to missed opportunities and a loss of competitive advantage. Decisions that involve risk may result in failure or loss. If a decision requires a high level of confidence, it may be necessary to rely on stakeholder opinion or a second opinion from an expert.

Additionally, it is important to consider the following points when making your decision:

• Are decisions reversible or irreversible?
• If a decision is irreversible, can it be made reversible?

Whether a decision is reversible or irreversible refers to whether the decision can be remade once it has been made. For example, a decision to develop a new product can be changed or canceled midway through development. On the other hand, the decision to merge companies cannot be reversed once the merger is finalized. Generally speaking, reversible decisions are less risky than irreversible ones, but redoing them can be costly and time consuming, so you need to take that into account.

Also, even irreversible decisions may be made reversible in some way. For example, by limiting the scope and duration of a decision, the impact of a decision can be localized or tested. Specifically, when making large-scale capital investments, companies may choose to lease rather than purchase, and when investing in new businesses, they may invest in funds in stages or invest in small-scale projects before fully deploying them. By conducting a step-by-step evaluation such as implementing a pilot project, it becomes possible to make decisions about whether to modify or cancel the project.

Summary

In this article, we have broken down the key elements of making rational decisions based on incomplete information and introduced the following points:

• Illusions surrounding data usage in business:You don't need to be a math specialist to make decisions using data. It is also wrong to think that big data will enable us to make "perfect" decisions. The ability to use information wisely and make decisions is required.
• Biases caused by using “intuition” vs. “data” to make decisions:Relying too much on intuition or data can distort your judgment. It requires a balance between intuition and data, between logic and creativity.
• The ability to ask questions:By asking questions about a problem before making a decision, you can clarify the problem and formulate a hypothesis for the problem. It also opens up new perspectives and possibilities.
• Problem setting ability:There are methods for identifying issues from past data, methods for estimating issues by formulating future hypotheses (visions), and a process called IWIK ("I Wish I Knew") to frame problems and focus on essential information. There are methods etc.
• Ability to question data:Verifying the trustworthiness of data and understanding its context and context can inform criteria for determining the meaning and validity of data.
• Know the key points in decision making:It is important to consider various factors in decision making, such as the purpose of the decision, deadlines, people involved, time, risk, and reliability.

By understanding and utilizing these points, you will be able to make better and more optimal decisions in the field of business.

Please try the following actions.

1. Try applying these points to your own business or project. Think about how you can apply these points to your specific situation or problem.
2. If you would like to learn more deeply, please refer to our8 basic steps for data analysis
3. If you have specific problems or challenges, we recommend consulting with an expert or consultant. We can provide the best advice and solutions depending on your situation and goals.
   

Through these actions, we will further improve our ability to make rational decisions with incomplete information. Successful business leaders are tough questioners who understand information quickly, use good judgment, and have keen insight. Acquiring these skills will take you one step closer to achieving that goal.

XICA has been providing services in the data science field in marketing for over 10 years, and has a track record of supporting over 250 companies, mainly domestic enterprise companies. Our analysts and consultants have deep and deep expertise across a wide range of industries and use data science to help our clients make better decisions. If you are interested, pleaseContact us.

Author information

Marketing team of XICA Co., Ltd.