Only about 31 percent of Copperline's cancellations are within a churn model's reach, and MT362's Unit 2 assessment weighs that share against the vendor fee, staff hours and discount cost. Searches like "mt 362 unit 2 assignment example", "mt362 unit 2 sample" and "mt362 unit 2 example" land here.
What a finished MT362 Unit 2 use case assessment looks like
A six-page assessment in four parts: problem, options, costs and a recommendation. The problem section works from the chain's own exit survey, answered by 2,860 cancelers over the past year: 31 percent said food was piling up, 19 percent that a pet had died or been rehomed, 17 percent price, 14 percent a switch to another retailer, 11 percent a vet-prescribed diet and 8 percent other. Three responses are then compared in one table: a schedule calculator that sets delivery by bag size and pet weight, a rule offering a skip whenever a subscriber edits the schedule, and a vendor-scored churn model triggering tailored offers a month before the likely cancellation. Each row carries cost, data needed, staff time and the share of cancellations it could plausibly reach.
How a MT362 Unit 2 example is structured
Problem precedes technology throughout. The paper states the business outcome, fewer cancellations net of discount cost, before any tool appears, and it tests every option against the reasons customers actually gave. Sorting those reasons by what could change them does most of the analytical work: a schedule problem answers to a process fix, a pet's death answers to nothing a marketer should attempt, and only price and competitor switching sit where a prediction might help. Costs are laid out for each option on the same basis, first-year dollars and analyst hours, so the comparison stays honest. The model is neither dismissed nor assumed; its reach is capped at roughly 31 percent of cancellations and its value estimated inside that cap. A closing recommendation sequences the options, cheapest first, and carries the model forward only as a candidate for a controlled test.
Six reasons, sorted by what changes them
Survey reasons are grouped into three bins: fixable by process, outside marketing's reach, and responsive to an offer. Only price and competitor switching, 31 percent together, land in the third bin, which sets the ceiling for any model.
Food piling up is a scheduling problem
A 15-pound bag for a 12-pound terrier lasts far longer than the default four-week cadence assumes. The paper proposes a calculator using bag size and pet weight, built by the web team in about three weeks with no vendor involved.
Pet loss is excluded from every option
Subscribers citing a pet's death receive no save offer under any option. This is framed as a design rule, with the note that a discount sent after that message would cost goodwill no retention figure could recover.
What the model would cost
A vendor quote of $48,000 a year, about 220 analyst hours to connect order and schedule data, and discount spend on subscribers who might have stayed anyway. Each figure is labeled as quoted, estimated or assumed.
Cheapest first, model under test
The recommendation sequences the work: calculator and pet-loss suppression now, skip rule next, and the churn model only as one arm of a controlled pilot, where its lift over the simpler rule can be measured rather than promised.
Where marks go in MT362 Unit 2
Tool-first proposals are the pattern MT362 graders most often mark down in this unit: an assessment that opens with a product and then locates a problem for it has skipped the question the course is built around. Grounding the problem in evidence the business already holds, an exit survey, cancellation logs or complaint themes, draws credit, and so does sorting causes by what could change them. Weaker papers compare a model against doing nothing rather than against a simpler rule, which flatters the model. License fees alone understate cost, leaving out staff time and the discounts a model would spend on customers who were never leaving. A recommendation that carries the AI option forward as a testable candidate, with its ceiling stated, reads as managerial judgment rather than enthusiasm.
Get a MT362 Unit 2 example written to your instructions
Tell us which business problem your Unit 2 scenario names and any figures it supplies, such as survey results, volumes or costs. Along with the assignment sheet and rubric, that lets the free first custom assessment compare an AI option against simpler responses on equal terms, costs included; turnaround is 24-48h.
MT362 Unit 2 questions, answered
What if the AI option is clearly the best answer?
Then the assessment should show why the simpler options fall short, not skip them. Graders tend to trust a recommendation more when cheaper alternatives were weighed and found wanting on stated grounds. Even a strong case usually benefits from naming the model's limits, such as the share of the problem it can reach, and from proposing a test before full adoption.
Where do I find cost figures for an AI tool?
Vendor pricing pages, analyst reports and case materials in the course are common sources, and many prompts supply figures directly. Where nothing is available, a clearly labeled assumption with its reasoning is acceptable in most sections. Include staff time and data preparation alongside license fees, since those often exceed the subscription price in the first year.
How technical should the assessment be?
Technical enough to name the system's prediction target, its data needs and how its output would reach a customer, and no further. The unit is weighing a business decision. Detail about algorithms rarely earns credit here, while a clear account of costs, data readiness and the share of the problem addressed usually does.