| Course | GF585 Strategic Financial Planning and Forecasting Analytics |
|---|---|
| Unit | Unit 3 |
| Paper type | Driver model build |
| Length | About 1,000 words, 6 pages double-spaced |
| Format | APA 7 student paper |
| School | Purdue University Global |
| Program | MS in Finance |
| Updated | October 2026 |
Free sample paper for GF585 Unit 3
Nineteen Shaded Cells: A Unit 3 Driver Model Build for a Commercial Linen Service
[Student Name]
Purdue University Global
GF585: Strategic Financial Planning and Forecasting Analytics
Unit 3 Driver Model Build
[Instructor Name]
[Date]
The linen service, its accounts, rates, costs and test results are composite and illustrative; the articles and book cited are real.
The Driver Tree, Drawn Before Any Cell
The tree starts from EBITDA and splits twice. Revenue divides into two segments, hospitality and healthcare, and each segment into three operating quantities: accounts served, pounds per account per week and price per pound. Cost divides by what makes it move. Plant labor, gas and utilities move with pounds washed. Linen replacement moves with revenue, because worn sheets and gowns are replaced in proportion to what is billed. Delivery cost moves with the number of routes, and the remainder, rent, management and insurance, stays fixed within the year.
Drawing the tree first exposed one early mistake. A first sketch tied delivery cost to pounds, but a truck costs the same whether it carries 3,000 or 3,800 pounds, so the cost belongs to routes. The model therefore steps up only when a new route is added (Pidd, 2010).
The Input Sheet
Every value a manager might want to change lives on one sheet, shaded yellow. Formulas elsewhere are plain and never hold a typed number. The nineteen inputs, with their sources, are:
1. Hospitality accounts, 140, from the billing system's active list.
2. Hospitality pounds per account per week, 2,600, from route sheets.
3. Hospitality price, 62 cents a pound, from current contracts.
4. Healthcare accounts, 85, from the billing system.
5. Healthcare pounds per account per week, 1,150, from route sheets.
6. Healthcare price, 88 cents a pound, from current contracts.
7. Weeks billed per year, 52.
8. Plant hours per thousand pounds, 9.5, from payroll and production logs.
9. Hourly wage, $19.40, from the posted wage scale.
10. Burden factor for taxes and benefits, 1.28, from payroll.
11. Gas use, two hundredths of a therm per pound, from utility bills.
12. Gas price, 90 cents a therm, from the supply contract.
13. Water, power and chemicals, 5.2 cents a pound, from utility bills and chemical invoices.
14. Linen replacement, 11 percent of revenue, from three years of purchase records.
15. Routes, 22.
16. Cost per route, $118,000 a year for driver, truck and fuel.
17. Fixed cost, $2.15 million.
18. One-shift plant capacity, 520,000 pounds a week, from the plant manager.
19. Contract price escalator, zero for last year, used in later units.
The Calculation Sheet
The calculation sheet multiplies down the tree. Weekly hospitality pounds are 364,000 and healthcare pounds 97,750, so the plant washed 461,750 pounds a week, or about 24.0 million for the year. Hospitality revenue is $11,735,360 and healthcare revenue $4,473,040, for a total of $16,208,400.
Plant labor follows from pounds: 24,011 thousand pounds at 9.5 hours each, $19.40 an hour and a 1.28 burden comes to about $5.66 million. Gas is about $432,000, water, power and chemicals about $1.25 million and linen replacement about $1.78 million. Routes cost $2.596 million and fixed cost $2.15 million. EBITDA is about $2.33 million, a margin of 14.4 percent.
A model earns the right to forecast next year only after it returns a year already known, so every input on the yellow sheet was first set to last year's actual values and the revenue line matched the general ledger to the dollar.
Calibration Against Last Year
With last year's inputs, the model returns revenue of $16,208,400, the figure in the audited statements. EBITDA comes within about $9,000 of the reported figure, a gap traced to a one-time legal fee booked in fixed cost. Rather than adjust an input to hide the difference, the gap is recorded on the test sheet with its cause. A model that is tuned until it matches tends to carry the tuning into the forecast, where it no longer belongs (Powell et al., 2008).
The Output Sheet
The output sheet shows a condensed income statement and one operating line: weekly pounds against the 520,000-pound limit. Last year the plant ran at about 89 percent of one-shift capacity. Showing that line beside the money matters because several later changes, such as new healthcare accounts, raise revenue and capacity use together.
The Test Log
Three changes were made one at a time, each followed through to EBITDA, then reversed.
1. Healthcare price up one cent. Revenue rises $50,830. Linen replacement rises with it, so EBITDA rises $45,239.
2. Plant hours fall from 9.5 to 9.0 per thousand pounds. Labor falls by $298,121, all of which reaches EBITDA.
3. Hospitality pounds down 3 percent. Revenue falls about $352,000, but labor, gas, utilities and linen fall too, so EBITDA falls $139,630.
The third test shows why the tree separates cost by driver. A loss of hotel pounds costs only about four tenths of the lost revenue, because pounds not washed carry no labor, gas or water. A model with a single margin percentage would have overstated the damage by more than double.
Checks Built Into the File
Two checks run on the output sheet. The first confirms that segment revenue adds to total revenue. The second compares weekly pounds with capacity and turns red above 95 percent. Spreadsheet research finds that most operational files contain errors and that review alone catches only some of them, so checks that run every time the file recalculates are worth the space (Panko, 1998).
What the Model Leaves Out
Routes step up in whole trucks, and the model asks the user to enter route count rather than computing it. Capacity is flagged, not modeled: the file does not add a second shift when pounds pass 520,000. Seasonality is absent, since this is an annual model, and quarterly patterns are added in the next unit. These limits are stated on the input sheet so no one mistakes a flag for a forecast.
References
Panko, R. R. (1998). What we know about spreadsheet errors. Journal of End User Computing, 10(2), 15-21.
Pidd, M. (2010). Why modelling and model use matter. Journal of the Operational Research Society, 61(1), 14-24. https://doi.org/10.1057/jors.2009.141
Powell, S. G., Baker, K. R., & Lawson, B. (2008). A critical review of the literature on spreadsheet errors. Decision Support Systems, 46(1), 128-138. https://doi.org/10.1016/j.dss.2008.06.001
What the GF585 Unit 3 assignment asks for
The third GF585 assignment asks for a working model and a short write-up of how it is built. Draw a driver tree from the result downward before entering any formula, splitting revenue into segments and operating quantities and splitting cost by what makes it move. Put every changeable value on one input sheet, mark inputs visually, and give a source for each. Keep formulas free of typed numbers. Calibrate the model by feeding last year's actual inputs and showing that it reproduces last year's results, explaining any gap rather than tuning it away. Add an output sheet with a summary and at least one operating measure, such as capacity use. Run a small number of test changes one at a time and interpret each. State what the model leaves out, and cite spreadsheet modeling research in APA 7.
Inside the GF585 Unit 3 example
The tree section opens the paper, splitting EBITDA into two segments and four cost behaviors, and notes how drawing it moved delivery cost from pounds to routes, citing Pidd. The input sheet is listed in nineteen numbered lines with a source beside each. The calculation sheet section walks pounds, segment revenue and each cost line down to EBITDA. The highlighted sentence sets the calibration standard, and a calibration section reports revenue matched to the ledger and a $9,000 EBITDA gap traced to a legal fee, citing Powell and colleagues. An output section shows the plant running at about 89 percent of one shift. Three tests follow with their EBITDA effects. A checks section cites Panko on spreadsheet errors, and limits on routes, capacity and seasonality close the build.
Reading the GF585 Unit 3 grading rubric
The rubric for GF585 driver models gives early points for structure: a tree drawn before the file, inputs gathered on one sheet with sources, and formulas that never hold a hidden number. Calibration carries real weight, because a model that cannot return a known year has not earned a forecast, so a reconciliation to actual results is worth showing in full. Test changes count for more when each result is interpreted, as with the hotel pounds test that exposes a single margin percentage as misleading. Instructors in the MS in Finance program look for built-in checks and an honest list of limits. The write-up should read in the file's own direction of flow. Pidd, Powell and colleagues and Panko, cited in APA 7, meet the source expectation for this unit.
Common GF585 Unit 3 mistakes, and how to avoid them
Weak GF585 Unit 3 builds start typing formulas before deciding what drives the result, so costs end up tied to the wrong quantity, such as delivery tied to pounds. Inputs get scattered across sheets or typed inside formulas, which makes a later change silently incomplete. Many skip calibration, or force a match by adjusting an input until the answer looks right. Tests are run but not explained. A single margin percentage replaces cost behavior, which exaggerates the damage of lost volume. Limits go unstated, so a capacity flag is mistaken for a forecast of overtime. Whatever business your case supplies, keep the sequence: tree, sourced inputs on one sheet, a calculation sheet that multiplies down the tree, calibration to last year, an output with one operating line, interpreted tests, checks and stated limits.
Get a GF585 Unit 3 example written to your instructions
What comes back is a workbook plus write-up: inputs kept apart from formulas, a known period reproduced before any forecast, and a log of test changes followed to the closing figure. It draws on the case data, model requirements and rubric of your GF585 Unit 3 section. The first sample is free, within 24-48h. The paper above is an original model document written by our desk, not a submitted student paper and not an official Purdue University Global document.
More GF585 papers and MS in Finance samples
- GF585 Unit 1: Water Use Forecast Miss Post
- GF585 Unit 2: Healthcare Share Strategy Memo
- GF585 Unit 4: Revenue Forecast
- GF585 Unit 5: Scenario Set
- GF585 Unit 6: Seminar Reflection
- GF585 Unit 7: Sensitivity Table
- GF585 Unit 8: Trend Forecasting Exercise
- GF585 Unit 9: Variance Analysis
- GF585 Unit 10: Planning Report
- GF584 Unit 3: Cash Flow at Risk Measurement Workup
- GF583 Unit 3: Check Collection Float Analysis
- GF582 Unit 3: Lending Probability Problem Set
- GF581 Unit 3: Transaction Exposure Register Analysis
GF585 Unit 3 questions, answered
Why reproduce last year before forecasting?
Because a model that cannot return a year you already know has no claim on a year you do not. The sample feeds actual accounts, pounds and prices through the formulas and gets last year's revenue to the dollar. If your figures disagree with the actuals, the gap points to a missing driver or a formula error before either one reaches the forecast.
How many inputs should a driver model have?
As many as the decisions it must answer, and few enough to scan on one screen. The sample uses nineteen because management asks about price, volume, productivity, energy and routes. A model with sixty inputs usually buries the few that matter, while one with five usually hides cost behavior inside a single percentage of revenue.
Is a driver tree required, or just helpful?
Some sections ask for one explicitly; many simply reward it. The tree shows a reader, before any formula, which operating quantities the result depends on and how they combine. In the sample it also exposed a mistake early: delivery cost had first been tied to pounds, when routes, and therefore trucks and drivers, follow the number of accounts.
Where can I find a free GF585 Unit 3 sample paper?
You can read a full GF585 Unit 3 driver model build here, covering a commercial laundry with nineteen sourced inputs, calibration to last year's revenue, three interpreted tests and APA 7 references.
Why does a 3 percent drop in hotel pounds cost so little EBITDA?
Pounds that are never washed need no plant labor, gas, water or chemicals, and linen replacement falls with revenue, so only about four tenths of the lost revenue reaches EBITDA.