| Course | MT358 Social Media Marketing and AI Optimization |
|---|---|
| Unit | Unit 6 |
| Paper type | Seminar reflection |
| Length | About 920 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 MT358 Unit 6
Who Switched It On for Complaints? An MT358 Unit 6 Seminar Reflection on a Bakery Chatbot's Coupon Reply
[Student Name]
Purdue University Global
MT358: Social Media Marketing and AI Optimization
Unit 6 Seminar Reflection
[Instructor Name]
[Date]
The seminar and the bakery case are composite; the sources cited are real.
How the Seminar Worked
The instructor released the case in four slices, and at each one the group decided what should happen next. First came the customer's public post: their child had reacted to a pastry labeled nut-free. Second came the bakery's automated reply, posted within a minute: an apology and a 20 percent coupon. Third came the screenshots, shared widely within hours. Fourth came the chain's statement the next morning, blaming a software error.
My First Position
My notes from the first slice read: "The vendor sold a broken product. A chatbot should never answer an allergy complaint with a coupon." I argued that the chain should hold the software company responsible.
The Question That Moved Me
Dana, who manages patient messages for a family medicine practice, asked the question that turned the discussion: who switched the assistant on for public complaints? At her practice, she said, the messaging system can draft replies, but staff decide which kinds of messages it may answer, and anything mentioning symptoms goes to a nurse. Someone at the bakery had decided that the chatbot could answer public complaints, including ones about safety. The software did what it was configured to do. Glikson and Woolley (2020), reviewing research on how people come to trust AI systems, found that trust in such systems rests heavily on how reliably they behave in the situations people actually meet, and the chain had never checked how its assistant behaved with a safety complaint.
Then the instructor added a question of her own: who at the chain was accountable at nine on a Saturday night, when the coupon posted? The answer was nobody. No one was watching the account, which the group agreed was the real failure, and that changed where I located it. The chatbot's reply was bad, but it was the predictable result of a scope decision made by people. The software could not know that an allergy complaint was different from a complaint about a late order unless someone had told it, and no one had.
Human Accountability for Automated Output
The course chapter on automated customer care argues that organizations remain accountable for what their automated systems say, because people choose where to deploy them. The trade article we read on chatbot escalation recommended that safety, health and legal complaints always route to a person. Coombs (2007), whose work on crisis communication we read earlier, argues that an organization's response should match how much responsibility the public attributes to it. Blaming the software in the next morning's statement asked the public to see the chain as a victim, when the public saw it as the party that put the chatbot in charge.
The chatbot did not decide to answer an allergy complaint; someone decided it could answer every complaint.
Where I Still Disagree
I have not abandoned my first position entirely. Vendors of these tools do oversell them, and the bakery's vendor advertised its assistant as able to handle customer care from start to finish. A buyer who takes that promise at face value has been misled to some degree. But the seminar persuaded me that being misled does not transfer accountability. The chain chose to believe the promise, chose not to test it on a complaint about safety and chose to leave the account unwatched on a weekend evening. The vendor's overselling explains the decision; it does not excuse it.
What Would Have Been Better
At the first slice, a person should have answered: acknowledging the reaction, asking whether the child was safe, offering to speak directly and saying the chain would investigate the labeling. No coupon. At the fourth slice, the statement should have said that the chain had allowed an automated system to answer a safety complaint and had now stopped it from doing so.
The Yarn Studio
At the yarn studio, we use an automated tool to suggest replies to common questions in our Instagram messages, such as shipping times. After the seminar, I checked its scope. It was set to suggest replies for every message, including complaints about dye bleeding onto skin and possible allergic reactions to wool. I changed the setting so that any message mentioning skin, allergy, rash, bleed or refund goes to me or one of the owners, and the tool suggests nothing. I am also proposing two rules for the protocol we will write in Unit 8: automated replies are barred from any public thread that contains a complaint, and every scheduled post pauses as soon as a complaint begins to spread, so a cheerful colorway post never appears beneath an angry one. Castelo et al. (2019) found that people are less accepting of algorithms for tasks they see as requiring human judgment, and a customer reporting a reaction is exactly that kind of task.
The Lesson I Am Keeping
My notes from the first slice blamed a vendor for a broken product. My notes from the last slice blame no one in particular, because the failure was a decision that nobody remembered making. The most important choice about an automated tool is where it is allowed to speak, and that choice always belongs to people, who should write it down and review it.
References
Castelo, N., Bos, M. W., & Lehmann, D. R. (2019). Task-dependent algorithm aversion. Journal of Marketing Research, 56(5), 809-825. https://doi.org/10.1177/0022243719851788
Coombs, W. T. (2007). Protecting organization reputations during a crisis: The development and application of situational crisis communication theory. Corporate Reputation Review, 10(3), 163-176. https://doi.org/10.1057/palgrave.crr.1550049
Glikson, E., & Woolley, A. W. (2020). Human trust in artificial intelligence: Review of empirical research. Academy of Management Annals, 14(2), 627-660. https://doi.org/10.5465/annals.2018.0057
Reading the MT 358 Unit 6 instructions
Unit 6 asks for a reflection on a seminar discussion, written in the first person. The prompt expects a short account of how the seminar worked, the position the student held at the start, the exchange that changed or tested it and a course concept applied to the case. It closes with a practical change the student will make. Instructors look for distance traveled rather than a polished final opinion. A starting view should be recorded honestly, even if it now seems wrong. Classmates are credited by first name. A reflection is not an essay about automation, so general claims about AI should be kept to what the seminar raised. Disagreement that survives the discussion is welcome if it is explained.
Inside the MT358 Unit 6 example
Built in two halves, the sample follows the case's timeline first and the writer's changing view second. It opens by describing the four slices. The writer's first note is quoted directly. Dana's question about who switched the assistant on, and the instructor's question about nine on a Saturday night, are treated as the turning point, with Glikson and Woolley on why reliability in real situations shapes trust. A section on human accountability applies Coombs, under a highlighted line that someone decided the chatbot could answer every complaint. A section on lingering disagreement keeps the view that vendors oversell. What should have happened is set out next, followed by the setting the writer changed at the yarn studio and two rules proposed for Unit 8.
Reading the MT358 Unit 6 grading rubric
Seminar reflections tend to be graded on four qualities: an honest account of the starting position, a clear description of what changed it, application of a course concept and a practical commitment at the end. A quoted first note shows candor. The turning point should be a specific exchange, credited to the person who made it. One concept applied closely earns more than several mentioned in passing. Reasoned disagreement that survives the seminar shows independent thinking. A concrete change, such as a setting altered or a rule written, is worth more than a promise to be more careful. Writing stays in the first person, and sources follow APA 7. A clear sense of the seminar's structure helps the reader follow the shift.
Common MT358 Unit 6 mistakes, and how to avoid them
Weak reflections often retell the bakery case in full and tack on a closing opinion, which leaves out the thinking the unit asks for. Others present the student's view as having been right all along. Some name no specific exchange, so the change of mind appears out of nowhere. Another common problem is a closing promise to use AI responsibly that commits to nothing. Before drafting, find the note you wrote at the start of the seminar and the moment you stopped agreeing with it. Write about the gap between them. Then name one thing at your workplace or course brand that you will configure, check or forbid because of what you learned.
Get an MT358 Unit 6 example written to your instructions
Tell us which public misstep the Unit 6 seminar replayed, or which case the written alternative gave, and note where your own view shifted and where it held. Include the rubric. From that account a first-person reflection is written, naming classmates by first name alone, and returned within 24-48h, free as your first custom sample. 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 MT358 papers and BS in Business Administration samples
- MT358 Unit 1: Brand Community Misstep Post
- MT358 Unit 2: Community Research Brief
- MT358 Unit 3: Six-Week Social Content Plan
- MT358 Unit 4: Paid Social Targeting Memo
- MT358 Unit 5: AI Ad Optimization Exercise
- MT358 Unit 7: Influencer and Disclosure Plan
- MT358 Unit 8: Crisis Response Protocol
- MT358 Unit 9: Social Analytics Report
- MT358 Unit 10: Social Campaign Plan
- MT357 Unit 6: Campaign Measurement Seminar Reflection
- MT355 Unit 6: Published Study Seminar Reflection
- MT340 Unit 6: Committee Deadlock Seminar Reflection
- MT330 Unit 6: Market Entry Seminar Reflection
MT358 Unit 6 questions, answered
What if the seminar case was real rather than composite?
Discuss it using verifiable facts from reputable reports and the brand's own statements, and avoid speculation about individuals. Real cases often come with more detail than a reflection needs, so select only what bears on your change of view. Cite the reports you rely on, and focus on your reasoning rather than on retelling the incident.
Should a reflection take a position on AI tools?
To the extent the seminar moved you, and no further. A reflection is not an essay on automation; it records how your thinking developed. If your view on a tool changed, explain the specific exchange that changed it. If it did not change, explain what you considered and why you held your position, which can be equally valid.
How do I link the reflection to my own work?
Close with a concrete change you would make, tied to a real or course-assigned brand. A rule, a revised process or a question you now ask before acting all count. Keep it short and specific. Graders tend to credit a practical commitment more than a promise to be more thoughtful in future.
Where can I find a free MT358 Unit 6 sample paper?
You can read the whole MT358 Unit 6 reflection here; it follows the writer's shift from blaming a chatbot vendor to blaming the chain that set its scope.
What changes at the yarn studio in the MT358 Unit 6 sample?
The writer routes any message about skin, allergy, rash, bleeding or refunds to a person and proposes barring automated replies from complaint threads.