Unit 2: BI as Business Support
Human decision-making is a complex process. This is because humans are complex. No matter what your business is, running it successfully requires dozens, if not hundreds, of tactical decisions every day by each staff member. These can be as simple as whether individuals prioritize incoming emails or concerted project time for the first three hours in the office. Operational decisions are more complex than these simple daily decisions as they can change how the organization functions over months or years. These can also relate to personnel in terms of process improvements, training, or recruitment. They can also relate to larger operational decisions on how the business runs or what it produces. The most complex decisions, however, are strategic decisions, as they determine the direction of the business for years to come. Decisions at all levels require thoughtful consideration and, typically, data to be made most effectively. Having lots of data is never enough, however. How it is used, understood, and then applied is the complex process that management must deal with, in addition to measuring just how effective those decisions have been on the business' short- and long-term success. This is the value that business intelligence (BI) can bring to a firm that can collect and use data effectively. Whether decisions are big or small also depends on the business itself. Opening a second location is an enormous strategic decision for a small coffee shop. For Starbucks, however, it is tactical. According to Wikipedia, as of early 2020, there were 30,000 Starbucks locations globally. If it turned out to be a mistake to open a Starbucks in a single location and it did not do well, Starbucks could easily absorb the cost. A small coffee shop may not recover from the error. At the same time, Starbucks is unlikely to make the wrong decision because it can access and use massive amounts of data to open new locations. A small shop will likely rely more on the owner's intuition or "gut".
Completing this unit should take you approximately 7 hours.
Upon successful completion of this unit, you will be able to:
- outline how business intelligence teams "define the problem" by learning more about the context of "the problem" and its relationship to other aspects of the operation;
- compare the different methods BI uses to support management teams, such as data mining, reporting, and visualization, trend and statistical analysis, and predictive analytics; and
- describe various management tasks, from policy making to performance evaluation to improving procurement strategies to identify relevant trends to understand how they benefit from using BI.
2.1: Defining the Problem
Managers often recognize a problem that needs to be addressed or an innovation that could be made to enhance business success. How they frame that problem or decide whether and how that innovation will suit their business is the most important part of the decision-making process. If the wrong questions are asked at this stage, the entire process can be flawed, and a poor decision may be made.
Watch this video to see what happens when people ask the right questions. We will look into this a bit more in section 2.1.1.
Several years ago, a market research firm did a project for a religiously-based non-profit organization that once provided death benefit insurance to a now-defunct fraternal order. The group was trying to determine how to retain relevance and market share in a dynamic financial services industry. They intended for the firm to evaluate their products to determine what new lines they should offer.
As the market research firm discussed the client's problem with their team, they discovered that the client was not sure to whom they would sell these new products because their traditional customer base had disappeared. The firm planned the project to determine not the best product but who the best customers would be for their existing products, which they already knew well. Fortunately, this client was open-minded enough to recognize that the original question they had asked was not the one they needed answered.
2.1.1: Framing Internal Client Discussions
The best way to ensure you are asking the right questions is to listen to the client, whether internal or external, as they express their needs and concerns. For internal clients, it is easier to have some contextual and cultural understanding of why this particular concern is arising at this time, which may also help with some reframing. For instance, if your firm has just made an acquisition and it is trying to decide what to do about combining two new HR employee teams and systems, knowing the concerns of your friends in HR from last week's happy hour can help you to frame the requirements around the worries they have raised.
Once you have read, think through the discussion questions after reading the example of Belgium Brewing. How does this process take place in your organization? Are the employees a big part of the decision-making process, or does management maintain most of the control? Have you been asked to support strategic, operational, or tactical decision-making? How could Belgian Brewing use BI to enhance its decisions on how to increase sustainability while growing profits in the future?
Think about Malcolm Gladwell's stories about his friend Howard Moskowitz. Imagine how much more satisfied Prego was with Howard's decision-making support than Pepsi was. It took Howard several years after the Pepsi taste tests to realize Pepsi was asking the wrong question about how much aspartame to put in its new Diet Pepsi formula. By that time, he was primed to rethink Prego's approach to gaining market share by focusing on taste and texture.
When conducting market or competitive intelligence analysis, an external client typically arrives with a question they need to be answered to increase their market share or otherwise increase their competitive advantage(s). Such questions are called intelligence requirements. In business intelligence, which is focused internally, the client will come from within the business itself. This can make the process of ensuring the client is asking the right questions more or less difficult than telling an external client that they don't need to know what they think they need to know.
When an organization has fully embraced the value of BI, its culture will allow the analyst to "speak truth to power" and inform management that they are asking the wrong question without issue. When a business only thinks it is using BI, it may be more difficult for the analyst to speak up and challenge prevailing assumptions. This culture is the first thing for the analyst to recognize.
If the environment is permissive, there will be no obstacles to the analyst suggesting changes to the requirement. In a less permissive business environment, the analyst will have to learn how to most effectively communicate with their decision-makers by first identifying their individual motivations and influences and how to best communicate their findings, whether through written reports or in-person briefings.
The analyst will also have to learn how formal or informal these communications should be. This is fodder for an entire course as the possible intentions and communication strategies are infinite, but the article on negotiation in 2.1.3. may be of some use as you develop your strategies for communicating with your decision-makers.
2.1.2: Drafting the Terms of Reference (TOR)
Once the analyst or team has been given a requirement, the first thing to do is to draft a Terms of Reference (TOR) document. The first purpose of a TOR is to ensure the assigned analyst or team and the decision-maker fully agree with what is expected.
Once both sides agree to the TOR, it serves as a contract that keeps both sides informed. The decision-maker has agreed that the analyst or team understands what is expected, when, in what format, and that it will stay within defined resource constraints. For the analyst or team, the TOR serves as a checklist for the team to ensure everyone knows their roles and responsibilities, and including timelines helps ensure the project will meet its deadlines.
Without a TOR, analysts and teams are likely to go off and get lost "in the weeds", chasing interesting tangential leads that may not have ultimate value to the project, thus expending limited resources in both time and budget without gaining relevant insights and moving the project forward.
This additional reading provides some process ideas, from brainstorming to mind mapping, that may help you think about ensuring all needed elements end up in the TOR. Here are two models from environmental protection and education that can help you create the TOR template that is right for your business.
We will further investigate refining the scope of your project in the next lesson once the main parameters of the project are defined in the TOR.
2.1.3: Negotiating the Project Scope
Once it is clear whether this decision is tactical, operational, or strategic, the business can determine the effort that should be spent on it. Strategic decisions should always be better resourced than tactical ones, for instance, as their relative impact on the business will be much greater. Once the analyst or team has developed the TOR and the decision-maker has approved it, the analyst or team must further define its project scope and work plan.
This article will help you to understand the importance of setting parameters and defining your scope to be sure everyone stays on target and all the moving parts and pieces can be harnessed for a positive project outcome.
Sometimes there are competing priorities in an analyst team. We will discuss analytic team dynamics later in the course. For now, it will be useful to understand that defining the scope of your project beyond the TOR will require some negotiation within the team. To do this effectively without causing disagreements within a team that must continue to work together for a shared goal. Read this article on negotiation to understand how to manage competing priorities, etc.
Have you participated in negotiations at work? Have they followed the flow of the model from the article? Or have they been contentious or messy without a collaborative or satisfactory outcome? When everyone comes to the table with a positive attitude and is prepared with all of the information needed to develop an effective work plan, the experience can be very positive.
The article mentions the need for managers involved in the negotiation to be mindful of delegation. Part of the negotiation at this stage should be establishing the roles and responsibilities of each team member. It will be important to determine who has the needed skills and place them in the appropriate roles with the necessary authorities so they can participate in the project as meaningfully as possible. Once roles and responsibilities are determined, deliverables, project milestones, and timelines can be established.
2.2: The Art and Science of Decision-Making
Intelligence analysis and decision-making can be said to have elements of both art and science. This is because standard approaches and tested techniques can codify and help make the processes orderly and productive. However, these methods are only as good as the individuals who implement them. Humans can be unpredictable, and even the best forecaster can be hampered by the outcome of a project, even when the analysis has been perfect.
Example: The Human Element in Analysis
An analyst team was tasked to write a handbook using the Structured Analysis of Competing Hypotheses (SACH) analytic method. The protagonist analyst in the book was a hapless weatherman who had failed in a recent weather prediction. It had rained when he said it would not, and his boss' wife's family reunion picnic was ruined. This poor weatherman forgot his lunch the next day and, desperate to avoid his boss, conducted a SACH matrix to determine where his boss would most likely eat lunch so he could go elsewhere.
We will look at this SACH method in Unit 5, but suffice it to say, he used all the evidence at his disposal of his boss' heart disease and efforts to lose weight and to please his wife and determined the boss would go to the nearby Subway instead of Burger King for lunch. The analysis was conducted perfectly, but to his horror, when the weatherman saw his boss enter the Burger King where he had sought refuge, he heard the boss tell the cashier, "I had planned on going to Subway, but the smell of your fries was too appealing and I changed my mind".
The analysis was not wrong, but sometimes things happen in the human mind that are out of the analyst's control. This human element is discussed in the next two articles. You may be participating in either of these processes as an analyst or a manager. In developing the requirement, careful problem analysis is needed, which takes place long before any implementable decisions are made.
It is valuable to note the various decision-making styles briefly described: psychological, cognitive, and normative. Understanding which of these your management uses will go a long way to helping you most effectively negotiate requirements, project scope, and communicate your analytical findings. Another useful takeaway from the article is the description of the three main approaches to decision-making: avoiding, problem-solving, and problem-seeking. Sometimes the right decision is no decision, so avoiding it is not necessarily bad. However, if your manager consistently avoids decisions claiming there is not enough information despite your best efforts to provide it, you may need to find new ways to communicate your findings to allow the manager to have more confidence in making decisions.
The challenge of problem-seeking when it sends the team back to the proverbial drawing board can be overcome with a robust TOR development process and a fully informed and formal scope negotiation process, both of which will help alleviate concerns that the project is not progressing as needed. It will be quite valuable for you to know the individual approach your manager or management team is likely to use and that of the overall organizational culture.
The article mentions Daniel Kahneman without explaining much about him. We will get to know him much better in the next section when we start to think about thinking.
Sometimes the most well-structured decision-making processes go awry, not because of the process itself, but because of the participants or the environment. This article shows us what can go wrong and how to get things back on track to meet
your TOR deliverables.The article discusses obstacles to improved decision-making, including "cognitive limitations, heuristics and biases and individual inclinations". Heuristics are mental shortcuts individuals use to solve problems. These have great use, for instance, telling humans to run when they see a saber-toothed tiger without thinking too much about the decision. The choice of which cat to adopt from a shelter today may require less use of heuristics and more cognitive exercise.
2.2.1: Thinking about Thinking
This insight has profound implications for economics, public policy, and our self-awareness, which can help BI analysts avoid bias.
Kahneman and his long-time thinking partner, Amos Tversky, famously changed how humans think about their own cognition and are known for popularizing the concepts associated with behavioral economics and our awareness and understanding of our thought processes, or metacognition.
Kahneman's examples may not seem relevant to business or management, but a decision-maker's memory of the outcome of previous decisions may inappropriately cloud their recall of the process or certain aspects of the implementation. This may result in a manager believing deeply that a particular analytic method or other approach is useless because "it did not work last time". This natural human failure to correctly recall an experience and rely on memory, what Tversky and Kahneman call simply "the story we tell ourselves about the experience", can result in faulty decision-making.
In this exercise, you will create a two-column journal and describe what the text is saying and how that information makes you feel or think. This gives us insights into our thinking processes and pathways. Read the first excerpt, then make your own basic evidence log. Consider what the text says and means to you, and practice using your log for the next examples.
2.2.2: Use Analysis, or "Go with Your Gut"?
People use various methods to solve problems every day. The best ones require some complex thinking, but thinking is hard, and some people do not have a lot of experience putting heavy lifting into their brain power and look for easier ways out. These ways have varying degrees of effectiveness. Read this article to learn about several common problem-solving techniques.
How would you describe the three main problem-solving techniques discussed in the article: trial-and-error, algorithm, and heuristics?
Trial-and-error is like a reboot when your Wifi is not working. Sometimes we try different things for more basic problems, like when our Diet Pepsi does not emerge as expected from a vending machine. We may press the button several more times, then try banging on the machine, and finally press the refund button. When none of these work after several tries, we may find our big friend, Hank from maintenance, and ask him to pick up and shake the machine, or we may call the support number or put in some more money for a new transaction. Eventually, we will give up and try another way to get our afternoon caffeine fix.
We think of algorithms as the complex codes within computer programs that make selections along decision trees to speed up our work. But humans also use algorithms to complete step-by-step tasks or work out complex problems by breaking them into smaller, more manageable parts and solving each. Algorithmic problem-solving is more systematic or process-oriented than the stop-and-start activity of trial-and-error. It is also used to come to a specific conclusion and is not as useful for open-ended questions or problems.
Heuristics are mental shortcuts people use to solve problems that are generally more complex than those for which the other two methods may be applicable. Heuristics come in handy when we are overloaded with information and must pick out the nuggets that matter or when we don't have enough information. They are also useful when we do not have much time to puzzle over a problem and need to decide quickly. As noted in the article, humans have developed some great reasoning methods for problem-solving, which can all be highly effective when appropriately applied. However, roadblocks sometimes hinder our ability to determine the best course of action or solve some other problem most effectively. The article also mentioned functional fixedness.
We often refer to using the military to solve any kind of instability in the world without understanding its root causes that could, perhaps, be more appropriately solved with other forms of intervention. This tendency is often characterized as our propensity to see everything as a hammer when every problem looks like a nail. The problem might be a humanitarian crisis, for which the Red Cross/Red Crescent, rather than the military, is needed if we can avoid this tendency toward functional fixedness. Think about a time when you encountered someone stuck in functional fixedness, and you were able to see an innovative solution to their problem.
Other mental barriers can make us reach incorrect conclusions called biases. Every intelligence analyst's goal is to strive for objectivity in their work. Still, they are human, and humans have developed other decision-making shortcuts that can impair their ability to reach accurate conclusions.
So, it is valuable to make snap decisions in the grocery store on which jam to purchase or shampoo to use, as there are so many to choose from. You will have analysis paralysis if you try to rationally and analytically decide on everything you will buy. It is also wise to make snap decisions about that saber-toothed tiger heading your way. Analysis may be the very last thing you do otherwise. These are, in fact, times when it is OK to "go with your gut" but not for business decisions.
For more complex decision conditions, such as deciding which political candidate you will support (the one with the best hair!), do not just go with your gut, but take the time to be informed and make a selection based on reasoning. For the kinds of decisions presented to business intelligence analysts, tested, objective methods are far more effective than the CEO just making off-the-cuff determinations that could negatively impact a significant number of people or a market, depending upon the size of the firm and the importance of the goods or services they provide.
The article describes a few common biases that are important to be aware of. Go through the brief quiz at the end and see how you do. If you miss any questions, go back and re-read that section. Think about how you might apply the concepts in that section so that the ideas will make sense and stick with you. Work through the "Personal Application Question:" Which types of bias do you recognize in your decision-making processes? How has this bias affected how you've made decisions in the past, and how can you use your awareness of it to improve your decisions making skills in the future?
Consider a specific time when you showed bias in coming to a conclusion about a person or event. How did you realize your thinking process was biased? How can you use this knowledge to avoid making this type of biased decision in the future?
High school teachers have developed a useful matrix to determine when students are effectively using heuristics that are useful to us. Think about how you have used some of these heuristics today or in the past few days. How useful were they in determining an optimal outcome?
2.2.3: Decision-Making Approaches
Once you have broken the requirement into specific parts, you will begin to attack the target from various sides and must identify who is responsible for what. There are various ways of defining your scope, but in the end, your team will likely produce a series of brief reports that bring you closer to understanding your target or full requirement.
The examples from the article on defining your scope were varied, but your team is likely to produce a series of brief reports that bring you closer to understanding your target or full requirement. Once you have broken the requirement into specific parts, you will begin to attack the target from various sides and must identify who is responsible for what. As you will see from reading the article, some decision-making methods are used to determine who needs to be part of the decision-making process, depending upon their deciding authority or skills. For instance, someone from finance would need to be in on a procurement process, but maybe not someone from HR, unless it was a tool to be used by their department.
2.2.4: Structuring Decision-Making Effectively
Think about a time when you were part of an unstructured decision-making process. As noted in that article, this can lead to a breakdown in morale, with everyone confused and frustrated but uncertain at whom or what to direct it. There's no one to blame because no one is taking ownership of the decision. Even if everyone tries to be polite and collegial, nothing gets done because the information flow is haphazard, and there is no clear decision-making authority.
Like making a to-do list or outlining a set of steps before starting a complex project, it is important to just as effectively structure your business decision-making process to avoid missing crucial steps that may result in less-than-optimal results when you try to implement the wrong decision.
Read this article that goes more in-depth on the RAPID decision-making tool.
This provides a few examples to show how it can be used and why having such a tool can improve decision-making. In the RAPID example, the BI analyst or team will likely have the most influence on the "I" or "Input" part of the process. This is where additional information can be injected into the discussion.Even if RAPID does not feel like the tool for your business, having a structured and well-informed process can make all the difference, so you can always do some research and find one that suits your business culture and decision-making needs.
2.3: Using Data to Make Decisions
Technological advances have changed the practice of BI exponentially. Now the ability of sophisticated software to collect and process data from myriad sources allows so much more information to be available to analysts and managers, and it can overload them.
We will look more in-depth at dashboards in Unit 4, but this is a useful introduction. Think about any dashboards you may have encountered. Perhaps you did not think about all the underlying data that gave you the information you were trying to pull out of the more basic interface. Something as simple as an electronic catalog from which you can search for library materials could be considered a basic dashboard. You put in your search terms, which act as data filters, and the system shows you the best matches.
Going back to our lesson on asking the right questions, if you enter "Thomas Jefferson" and select "author", you will get references to the Declaration of Independence and other documents that he participated in writing, rather than that new biography, "Thomas Jefferson: New Insights", you had intended to locate. This is a fictitious book title, but the example reminds us of the importance of ensuring we have the right requirements for our intelligence project and that we ask our dashboards the right questions to get the information that will support our BI effort.
Read this article that will introduce you to the latest buzzword on how to more effectively manage all of this information: the dashboard.
2.3.1: Everyday Data
Returning to our unfortunate weatherman, we see how we can use data to make decisions on almost anything we can imagine, from the weather to relationships. Stanford University developed a platform for conducting Structured Analysis of Competing Hypotheses (SACH) and gave it to a graduate school for intelligence analysis for testing. Graduate students took to it like fish to water, using it to decide whether to break up with their partners, which job offer to accept, and to choose between two master's thesis topics.
However, when using data, it is important to be aware of any biases we may bring to our decision-making process. It is also useful to use a tested method, like a heuristic, or for more complex decisions, an analytic methodology. It is important to understand that there are an enormous number of them with varying degrees of efficacy, depending on the problem to which they are applied.
Example: How We Use Data to Make Everyday Decisions
You may use data to decide where to purchase your pomegranates. You may have received three grocery store circulars in the mail or your newspaper or have them on your handy grocery store phone app. When you see that pomegranates are $5 each at stores A and B, you wonder if you need pomegranates this week. Then, to your delight, you discover that store C has them for $2.50, and remember that they also carry your favorite brand of dangerously sharp cheddar cheese, which store B recently discontinued, and your mind is made up to do your shopping this week at C.
Then, you remember that only A gives 10 cents off each gallon of gas for every $50 in groceries you purchase, and you rethink your plan. Considering you only really need cheese and pomegranates and maybe some milk, an item you know is price-controlled and is unlikely to vary much between stores, you settle on C and head out confidently to satisfy your cravings, knowing that you have used data to make the wisest decision.
As you go about your day today, write down all the times you notice yourself using data to make decisions. Were there times you questioned the validity or relevance of the data you used? Did your data-driven decisions have the intended outcome?
2.3.2: Why Expert Judgement is No Better than Yours
Good judgment researcher Philip Tetlock wrote a book in 2005 called "Expert Political Judgement", which rocked the analytical world by showing how experts are not necessarily better at forecasting than dilettantes. He used the analogy of the Hedgehog and the Fox to say that Hedgehogs, representing experts, know how to do one big thing, roll into a prickly ball at the sign of danger. Foxes know many smaller tricks to avoid danger.
The fox vs. hedgehog construct can be likened to the "Soldier" vs. "Scout" approach to the world described in the video.
Do you think an analyst should be more like Galef's Soldier or Scout? What about a manager? How can the Scout traits be enhanced to increase analytic agility like the Fox's? Not only do they know many tricks, but they assess when and how to use them for the greatest effect. This is the same way a good analyst uses various analytic techniques to ensure the one or a combination selected is the right approach for each requirement.
Author Malcolm Gladwell popularized the "gut" approach the same year with his book, "Blink", which suggested that some people's snap decisions may be better than analysis. In combination, the two books were conflated by laypeople to suggest that there is little need for analysis, but that is an erroneous reading of both books.
Tetlock suggested experts get lazy and start thinking they know everything about their narrow field, and that's when they start making poor forecasts. Dilettantes know they don't know everything and try to update their understanding. Analysts use tested methods to analyze the situation and constantly allow the introduction of new information, whether or not it confirms or refutes their initial hypothesis, using lots of clever analytic tricks to maintain objectivity and create the most accurate estimates.Gladwell said experts can sometimes make snap decisions better than others based on their inherent subject-matter knowledge. He used the example of an ancient Greek art expert who could spot a fake statue based on a feeling when he saw it objectively because "it just didn't look right...it looked fresh" and not 2,000 years old.
The first expert's colleague, also an expert, so wanted such an object in his museum he allowed his expertise and objectivity to be set aside to believe the fake was real, thus allowing any new information to either confirm his belief or he rejected the information out of hand. This is not good analysis. The snap decision of the correct expert is also not analysis; it was a good use of his specialized art history knowledge.
2.3.3: How Forecasting can Help Decision-Making
Forecasting is how an analyst or team uses analysis to estimate what is likely to happen. This works much better in the short term than in the longer term, as conditions are unlikely to change as much in the next 6-12 months than in the next 5-7 years. However, as the COVID-19 pandemic has shown us, sometimes there are environmental shocks that make even the best short-term forecasting look unreliable in hindsight. Remember that extreme external shocks are rare; the last global pandemic, for instance, was a century ago. Thus, there is definite value in data-driven forecasting.
Watch this video for a basic case study showing how some simple forecasting can inform the optimal ways for a business to intervene in its strategy as conditions change. For instance, if you are trying to sell sneakers in the Netherlands, how could foreknowledge of the pandemic have helped you plan?
While public health officials have warned that the possibility of such an event was increasing with the rise of globalization, most businesses were probably not prepared for it unless they sold personal protective equipment (PPE). For instance, if you were the owner of a sports footwear company that happened to be the daughter of a high-ranking World Health Organization official, you might have been thinking about a pandemic.
Once it began, you were ready for the rise in outdoor activities as spring turned into summer in the Netherlands, and people were spending more time outdoors as they had time and needed to entertain their kids. The virus is believed to be spread less quickly in areas with good ventilation. Thus, even your older population with little interest in sports may become a new target market. Due to your forecasting ability, you were prepared ahead of your competition to make an intervention in your sports shoe strategy and be the first to serve this space.
Review this section to be sure you understand variable, fixed, and mixed costs.
Study Guide: Unit 2
We recommend reviewing this Study Guide before taking the Unit 2 Assessment.
Unit 2 Assessment
- Receive a grade
Take this assessment to see how well you understood this unit.
- This assessment does not count towards your grade. It is just for practice!
- You will see the correct answers when you submit your answers. Use this to help you study for the final exam!
- You can take this assessment as many times as you want, whenever you want.