| Course | MT236 Introduction to Supply Chain Management |
|---|---|
| Unit | Unit 4 |
| Paper type | Demand forecast exercise |
| Length | About 900 words, 6 pages double-spaced |
| Format | APA 7 student paper |
| School | Purdue University Global |
| Program | BS in Business Administration |
| Updated | October 2026 |
Free sample paper for MT236 Unit 4
An Opening Order of 238 Cases Is Not a Monthly Rate: A Unit 4 Demand Forecast Exercise for a Hot Sauce Maker
[Student Name]
Purdue University Global
MT236: Introduction to Supply Chain Management
Unit 4 Demand Forecast Exercise
[Instructor Name]
[Date]
Alamo Ridge Pepper Co., its shipment history, the new customer and every figure are composite and illustrative; bracketed figures stand where shipment records would supply the numbers; the sources cited are real.
The Data
Alamo Ridge Pepper Co. shipped [23,400] cases of five-ounce sauce in year one and [24,870] in year two, about 6.3 percent growth. Shipments follow a clear pattern: steady through spring and summer at about [1,700 to 1,900] cases a month, then a rise from September through November, peaking near [2,700] cases in October as grocery chains stock for football season and holiday cooking, and easing in December.
To test forecasting methods fairly, the first eighteen months were used to build each forecast and the last six, July through December of year two, were held back as a test. Each method forecast one month ahead, and the misses were averaged without regard to sign, the mean absolute deviation, so a 100-case shortfall and a 100-case overshoot count the same.
Method One: Last Month's Actual
The simplest forecast says next month will equal this month. For July through December, its forecasts were [1,860], [1,940], [1,930], [2,410], [2,670] and [2,720] cases, against actual shipments of [1,940], [1,930], [2,410], [2,670], [2,720] and [2,305]. The mean absolute deviation was about [216] cases. The method lags every turn: it underforecast the September rise by 480 cases and overforecast December by 415.
Method Two: Three-Month Moving Average
A three-month average smooths noise but lags even more when demand changes direction. Its forecasts for the same months ran from about [1,880] to [2,610], and its mean absolute deviation was about [307] cases, the worst of the three. It was still averaging summer months when the fall rise began.
Method Three: Last Year's Month, Scaled for Growth
The third method takes the same month last year and scales it by the 6.3 percent growth shown between the two years' January to June totals. Its forecasts were [1,913], [1,966], [2,392], [2,657], [2,764] and [2,285] cases. Its mean absolute deviation was about [26] cases, far better than the other two, because it builds in the seasonal pattern. Hyndman and Athanasopoulos (2021) recommend comparing methods on held-back data and note that a simple seasonal method is often hard to beat when the pattern repeats each year.
The New Customer
In January, Alamo Ridge begins shipping to a Southeast grocery chain with about [92] stores, served from a distribution center near Charlotte. The chain's opening order is [238] cases, enough to fill shelves in every store and stock its warehouse.
Repeating 238 cases every month would be a mistake. The opening order fills the pipeline once: shelves, backrooms and the distribution center. After that, the chain will reorder only what shoppers buy. Lee et al. (1997) showed how one-time orders and batch ordering can make a customer's orders look far more variable than its real demand, and treating a pipeline fill as a monthly rate is a simple version of that error.
The rate comes from comparable stores. In Alamo Ridge's existing Texas chain, stores of similar size sell about [1.6] cases of five-ounce sauce a month in their first year, before shoppers know the brand. Applied to [92] stores, that gives about [149] cases a month, or about 1,790 a year. The opening order of 238 is about 60 percent above that rate.
After January, the new chain will reorder what shoppers buy, not what it took to fill the shelves.
Why Not a Fancier Method
A regression on time with monthly dummy variables, or an exponential smoothing model with trend and season, would fit the history slightly better. With only two years of data, though, those models have little to learn from, and their extra precision would mostly be fitting noise. The scaled seasonal method is easy for the owner to understand and explain to the grocery buyer, and it can be replaced by a fuller model once three or four years of history exist. For the new chain, no historical method is possible at all, which is why its forecast starts from store counts and sales rates.
The Combined Forecast
For the coming year, the forecast for the existing business uses method three, last year's months scaled by recent growth. The new chain is added on top: 238 cases in January, then about [149] a month, rising by about 6 percent a quarter if early sales match the Texas stores, and scaled up for the fall season using the same monthly pattern. The combined forecast for October, the peak month, is about [2,870] cases for the existing business plus about [215] for the new chain.
How the Forecast Will Be Checked
Chopra and Meindl (2019) remind planners that every forecast misses, so each one should travel with its error and a schedule for revising it. Alamo Ridge will compare each month's forecast with actual shipments, track the mean absolute deviation for both accounts separately and revise the new chain's rate after three months of real orders. If the new chain's reorders run more than 25 percent above or below 149 cases for two straight months, the rate will be reset from its actual sales data.
References
Chopra, S., & Meindl, P. (2019). Supply chain management: Strategy, planning, and operation (7th ed.). Pearson.
Hyndman, R. J., & Athanasopoulos, G. (2021). Forecasting: Principles and practice (3rd ed.). OTexts. https://otexts.com/fpp3/
Lee, H. L., Padmanabhan, V., & Whang, S. (1997). Information distortion in a supply chain: The bullwhip effect. Management Science, 43(4), 546-558. https://doi.org/10.1287/mnsc.43.4.546
What the MT236 Unit 4 assignment asks for
Unit 4 typically asks for a short forecasting exercise on a company's demand history, comparing simple methods and recommending one. Most introductory sections expect a naive forecast, a moving average and perhaps a seasonal or weighted method, each scored on the same months. Hold back recent months the methods did not use and measure their errors with mean absolute deviation or a similar measure, so the comparison is fair. Describe any seasonal pattern in plain terms. Many prompts add a case with no history, such as a new customer or product, which must be built from parts like store counts and sales rates. Say how the forecast will be checked and updated. Keep the arithmetic visible, and label composite figures. Most submissions run about four pages, with a shipment chart and an error table.
How this MT236 Unit 4 example is built
The exercise describes the two years of shipments and their fall peak, then explains the held-back test and the error measure. Each method gets a section with its six forecasts, the actual shipments and its average miss, and a margin note explains why testing on unseen months matters. Hyndman and Athanasopoulos support the strong result for the seasonal method. A short section explains why a fancier model is not yet worth it with two years of data. The new customer section shows why the 238-case opening order fills the pipeline once, links that to Lee and colleagues' work on order distortion with a second note, and builds about 149 cases a month from comparable stores, which the highlighted line contrasts with the opening order. A combined forecast for October and a checking rule close it, citing Chopra and Meindl on revising forecasts.
MT236 Unit 4 rubric: what earns full points
Forecasting exercises in MT 236 are usually graded on fair testing, correct error measures and sensible treatment of new demand. Holding back months the methods did not see is the core of a fair comparison, and graders notice when methods are scored on data they were built from. Mean absolute deviation should be computed correctly and interpreted in cases, not only as a number. Explaining why a method fails, such as lagging the seasonal turn, shows understanding. A new customer forecast built from store counts and rates earns credit, while repeating an opening order loses it. A plan for checking and updating the forecast shows the writer knows forecasts are always wrong. Charts and tables should be labeled. The forecasting text by Hyndman and Athanasopoulos should be cited wherever a method is chosen.
MT236 Unit 4 help: mistakes that cost points
Forecast papers often lose points by testing methods on the same months used to build them. Another common error is choosing a method because it fits the past best rather than because it predicts unseen months best. Some writers ignore seasonality and wonder why every method misses the fall. Others forecast a new customer by repeating its first order. Error measures are sometimes computed without absolute values, so overs and unders cancel. Plans for updating are frequently missing. Watch units: a forecast in cases and actuals in bottles will not compare. If your company has little history, build demand from customers, stores or users and sales rates. Share your demand data and any new account details, and our writers can help test the methods fairly.
Get an MT236 Unit 4 example written to your instructions
Share the history or case data your Unit 4 prompt supplies, and say whether the forecast covers existing products, a new one or both. Include the rubric. Methods tested on held-back months and every assumption listed: that first custom sample costs nothing and comes back within 24-48h. Inputs the case lacks appear in the assumptions box, never hidden inside a formula. 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 MT236 papers and BS in Business Administration samples
- MT236 Unit 1: Discussion Board Post
- MT236 Unit 2: Supply Chain Map
- MT236 Unit 3: Supplier Evaluation
- MT236 Unit 5: Inventory Analysis
- MT236 Unit 6: Transportation Comparison
- MT236 Unit 7: Distribution Network Memo
- MT236 Unit 8: Process Improvement Proposal
- MT236 Unit 9: Seminar Reflection
- MT236 Unit 10: Supply Chain Report
- MT221 Unit 4: Complaint Response Letter
- MT220 Unit 4: Trade Policy Brief
- MT219 Unit 4: Product Analysis
- MT217 Unit 4: Time Value Problem Set
MT236 Unit 4 questions, answered
Which forecasting methods does an introductory course expect?
Usually the simple ones: a naive forecast, moving averages, weighted moving averages and basic exponential smoothing, sometimes with a seasonal adjustment. The example uses three because comparing them teaches more than perfecting one. More advanced methods rarely earn extra credit at this level unless the prompt asks, and they make the assumptions harder for a reader to check.
What is mean absolute deviation?
The average size of the forecast errors, ignoring whether each error was high or low. Subtract each forecast from the actual figure, drop the minus signs, add the results and divide by the number of periods. It comes out in the same units as the forecast, cases here, which makes it easy to explain to whoever plans production.
What if the product or customer has no history?
Then the forecast is built from parts that can be estimated: the number of stores or customers, how many items each carries, and a sales rate borrowed from something similar, adjusted for known differences. Every part is stated as an assumption. The example borrows the Texas chain's rate and cuts it by a quarter because the brand is unknown in the Southeast.
Where can I find a free MT236 Unit 4 sample paper?
This page contains a complete MT236 Unit 4 demand forecast exercise for a hot sauce maker. It tests three methods on six held-back months, measures their errors, forecasts a new grocery chain from store counts and includes margin notes.
Why does the scaled seasonal method win in the sample?
Because the shipments rise the same way every fall. Using last year's same month, scaled by recent growth, builds that pattern in, while last month's actual and a moving average both lag the September turn.