| Course | BU 351 Introduction to Business Analytics |
|---|---|
| Unit | Unit 1 |
| Paper type | Analytics types and framing |
| Length | About 910 words, 6 pages |
| Format | APA 7 student paper |
| School | Herzing University |
| Program | BS in Business Analytics |
| Updated | September 2026 |
Free sample paper for BU 351 Unit 1
"Memberships Are Slipping": Sorting a Climbing Gym Owner's Request Into Four Kinds of Analytics and Rewriting It as Questions Data Can Answer
Student Name
Business, Herzing University
BU 351: Introduction to Business Analytics
Instructor Name
January 21, 2026
"Memberships Are Slipping": Sorting a Climbing Gym Owner's Request Into Four Kinds of Analytics and Rewriting It as Questions Data Can Answer
The Request as It Was Made
In early January 2026, the owner of Basecamp Climbing, a composite chain of seven indoor climbing gyms along Colorado's Front Range built for this course, sent the analytics intern a two-line message: "Memberships are slipping. Find out why and tell me what to do about it." The owner has a membership database covering January 2023 through December 2025, with one record per member showing the gym, plan, join date, cancellation date if any, age band, recent visits, autopay failures, whether the member took the introductory class and whether they were referred by a friend. This paper does not analyze that data. It does the step that comes before analysis: it sorts what the owner is asking into the four kinds of analytics, shows that the message contains several different questions and rewrites each as something a specific analysis could answer.
Four Kinds of Analytics, Defined by What They Cannot Do
Business analytics is usually divided into four types. Descriptive analytics summarizes what happened: how many members joined and left, and when. Diagnostic analytics looks for why it happened, by comparing groups and conditions. Predictive analytics estimates what is likely to happen next, such as which current members are likely to cancel. Prescriptive analytics recommends what to do, weighing actions against goals and limits. Davenport and Harris (2007) describe organizations that compete on analytics as moving along this sequence, from reporting what happened toward using models to decide what to do.
Each type is more useful when defined by what it cannot establish. A descriptive summary cannot explain; it can show that cancellations rose in May but not why. A diagnostic finding cannot forecast; it can show that members who skipped the introductory class left more often, but not which specific members will leave next month. A prediction cannot choose; it can rank members by risk, but it cannot say whether spending money to keep them is worthwhile. And a prescriptive answer is only as good as the costs, effects and limits fed into it. Provost and Fawcett (2013) make a similar point when they argue that the first task of any analytics project is understanding the business problem well enough to know which kind of answer would actually help.
The Questions Hiding in Two Sentences
The owner's message contains at least four separate questions. The word "slipping" hides a descriptive question: is the number of members actually falling, or is something else, such as the rate of cancellations or the pace of new sign-ups, getting worse? The request to "find out why" is diagnostic: which kinds of members, gyms or circumstances are associated with leaving? The request to "tell me what to do" is prescriptive, but it quietly depends on a predictive question: which current members are most likely to leave soon, so that any action can be aimed at them rather than spread across everyone?
These questions have to be answered in order. Recommending an action before establishing whether memberships are really falling risks solving a problem that does not exist. Explaining why members leave before measuring how many leave and when risks explaining a pattern that is not there.
A useful test for any framing is to ask what answer would change the owner's next decision. If memberships turn out to be growing, the owner does not need a rescue plan; if a particular gym accounts for most of the losses, the plan should be local rather than chainwide; if payment problems drive cancellations, the fix may be in billing rather than in marketing. Framing the questions first makes it possible to see which answers matter before any time is spent computing them.
Rewriting Each Question So It Can Be Answered
Each question is rewritten below with the measure, the level and the period that would make it answerable from the data the owner has.
| Type | Original wording | Answerable version |
|---|---|---|
| Descriptive | Memberships are slipping | How did active memberships, new joins and the monthly cancellation rate change from 2024 to 2025, overall and by gym? |
| Diagnostic | Find out why | Among members who joined from June 2024 through September 2025, how does the share who canceled within 90 days differ by introductory class, plan, referral, autopay failures and gym? |
| Predictive | (implied) | Using information known on June 30, 2025, how well can we rank active members by their chance of canceling in the next 60 days? |
| Prescriptive | Tell me what to do | With a monthly retention budget of $5,000 and limited staff time, how should spending be divided among available retention actions to keep the most members? |
What the Data Can and Cannot Support
The database can support the descriptive and diagnostic questions directly, since it records joins, cancellations and several member characteristics. It can support a predictive model, with one caution: some fields, such as recent visits, are recorded as of the member's last date rather than on a fixed date, so a model must use only information that would have been known at the time of prediction. The database cannot answer the prescriptive question alone. It does not record what retention actions cost or how well they work, so those figures must come from the owner and staff and be labeled as assumptions. It also cannot say why members leave in their own words; no exit survey exists. Wang and Strong (1996) argued that data quality depends on fitness for the consumer's purpose, not only accuracy, and these limits are a matter of fitness: the data are good for some questions and silent on others.
The Framing Handed to the Next Stage
The analysis that follows in this course will proceed in the order set out above. It will first check the quality of the membership data, then answer the descriptive question, explore the diagnostic one, build a simple predictive ranking and finally formulate the prescriptive question as an allocation problem. The owner's two sentences will be answered, but as four questions rather than one. The gyms, the owner and the member file are fictional, invented for this course.
References
Davenport, T. H., & Harris, J. G. (2007). Competing on analytics: The new science of winning. Harvard Business School Press.
Provost, F., & Fawcett, T. (2013). Data science for business: What you need to know about data mining and data-analytic thinking. O'Reilly Media.
Wang, R. Y., & Strong, D. M. (1996). Beyond accuracy: What data quality means to data consumers. Journal of Management Information Systems, 12(4), 5-33. https://doi.org/10.1080/07421222.1996.11518099
The Herzing BU 351 Unit 1 instructions
The first BU 351 unit asks students to take a loosely worded business request and frame it as questions analytics can answer. Most versions want the four types of analytics explained, the request broken into the separate questions it contains, each question classified by type and rewritten so that a specific analysis could answer it. Many rubrics also ask what data would be needed and what the available data cannot support. Instructors reward framing that is specific about measures, time periods and levels of analysis. The case may be supplied or chosen. Plan on a brief paper built around a table of rewritten questions, citing the course text and any analytics sources in APA 7.
How the BU 351 Unit 1 example is put together
The sample quotes the owner's request exactly and describes the membership data available. A definitions section explains the four types of analytics, each with a sentence on what it cannot do, supported by two sources. The next section shows that the two-sentence request hides four questions and explains why they must be answered in order. A table rewrites each question with a measure, a level and a time period, for example turning slipping into a comparison of active members, joins and cancellation rates across two years. A data-fitness section identifies what the file can support and what it cannot, and the paper ends by handing the framed questions to the next unit.
BU 351 Unit 1 rubric: what full points look like
Herzing's criteria for the framing unit usually weigh understanding of the analytics types, the decomposition of the request, the quality of the rewritten questions and awareness of data limits. Understanding earns full points when each type is defined accurately and its limits are clear. Decomposition is judged by finding all the distinct questions hidden in the request, including implied ones. Rewritten questions score highest when each names what will be measured, for whom and over what period, as the sample's table does. Awareness of data limits rewards a frank statement of what the data cannot answer. Tidy organization and accurate citations take the last points, and instructors notice when the framing is carried into later units.
BU 351 Unit 1 help: mistakes that cost points
The most common Unit 1 mistake is jumping straight to an answer, such as recommending a discount, without asking what the request actually means. Another is classifying the whole request as one type when it contains several. Students often rewrite questions without measures or time periods, leaving them as vague as the original. Some papers define the four types correctly but never apply them to the case. Others ignore the limits of the data, promising diagnostic answers from a file that cannot support them. Finally, keep the original wording in view; showing the gap between how a manager speaks and what an analysis can answer is the point of the exercise.
Get a BU 351 Unit 1 example written to your instructions
Give us the Unit 1 instructions, the BU 351 rubric from your classroom, and the business request or scenario the assignment supplies. We write a custom example that keeps the original wording intact, defines each type by its limits, splits the request into separate questions and closes on a rewrite carrying measure and period. First custom sample free, back in 24 to 48 hours.
More BU 351 papers and business analytics samples
- BU 351 Unit 2: Data quality assessment
- BU 351 Unit 3: Descriptive analysis
- BU 351 Unit 4: Exploratory analysis
- BU 351 Unit 5: Predictive exercise
- BU 351 Unit 6: Prescriptive or optimization exercise
- BU 351 Unit 7: Reporting to a business audience
- BU 351 Unit 8: End-to-end analytics case
- BU 451 Unit 1: Environment and setup exercise
- BU 453 Unit 1: Requirements to entities exercise
- BU 454 Unit 1: Encoding principles exercise
- BU 455 Unit 1: Scale assessment
BU 351 Unit 1 questions, answered
Is diagnostic analytics a separate type or part of descriptive?
Sections differ, so follow whatever your instructions and reading set out. Where four types are taught, diagnostic sits between summarizing and forecasting and answers why a pattern appeared. Where three are taught, that work is usually folded into the descriptive stage. Either framing is defensible in the paper as long as you say which one you are using and stay with it throughout.
Can one question belong to two types at once?
A question that seems to usually turns out to be two questions, and separating them is exactly what earns credit here. Asking which customers will leave and what should be done about it is a prediction followed by a decision, and the two need different data and different methods. Split it, label each half, and say which one has to be answered first.
How much detail should the data section carry?
Enough to show the question is answerable in principle. Name the fields each question would need, the level the records would have to sit at, and how far back the history would have to run. You are not sourcing anything yet, so avoid naming systems or vendors; describe the data by what it contains rather than by where it might live.
Where can I find a free BU 351 Unit 1 sample paper?
This page shows the full BU 351 Unit 1 framing paper at no cost, including the table of rewritten questions. If your case begins with a different request, send the prompt and the wording you were given, and a free custom framing paper will follow, often within two days.
Can one question belong to two analytics types in BU 351 Unit 1?
A single sentence often contains more than one question, and the sample splits the owner's request into four. Each rewritten question should belong to one type, since the type decides which analysis answers it. If a question seems to straddle two, it usually needs splitting.