PRAO 3017 · Welcome. You're in the right place.
PR, digital marketing, and a stint as a Washington DC correspondent. I've pitched gatekeepers, and I've watched the gatekeepers stop mattering.
Health and risk communication, digital media, and AI in communication, mostly how communities stay resilient when information gets messy. PhD in mass communication.
CVA 928 · Mondays 14:00–17:00 · lolawxie@hkbu.edu.hk
Not an editor. Not a journalist. Not even, mostly, your followers. This course is about working both sides of that.
A new account, one platform, eight weeks live. You own the audience relationship, and you find out exactly what that costs in time, consistency, and nerve.
Vet creators against a real client brief with a fixed budget. Recommend three. Measure what actually happened. Handle it when it goes wrong.
You will run an account for real. Creators from Hong Kong and elsewhere will come in and tell you what nobody puts in the highlight reel.
You're graded on how well you reason about your own data, not on how fast you grow. A small account analyzed rigorously beats a large account analyzed lazily. That's not a consolation prize; it's the whole design.
The next seven slides go through each of these. The account is the biggest single piece and the one that runs longest.
Most sessions end with a short studio lab you submit in the room. That submission is your attendance. It can't be done remotely or afterwards.
Your positioning brief at Session 3 is pass/revise, not graded, but passing it is part of this component, and your account can't launch until it does.
Complete one native platform analytics certification of your choosing. Guidance on which ones count is on the site after Session 3.
Twice in the semester you produce content with generative AI and hand in the whole process, not just the result.
Every prompt, including the ones that didn't work.
Exactly what the model gave you.
What you actually published, or would.
Where the model was wrong for your audience, and how you knew.
The critique is what's graded. The other three are evidence that the critique is honest. Lab 1: assigned S4, due S5. Lab 2: assigned S8, due S9.
Two pages. Name your platform and defend it. Subject and audience. A weekly output commitment. Three or four content pillars. Success criteria that are not follower count. An ethics and risk plan.
Eight entries, 250 words each, each with a dated screenshot of your native analytics. Same four questions every week. Next slide.
1,500 words on what the data taught you, including two decisions you'd reverse. Plus a dated evidence appendix.
The site has a drafting tool with exactly these four boxes. Draft there, submit on Moodle.
Evidential rigor. Dated, legible, complete. Every claim traceable to a screenshot.
Analytical quality. Do your changes follow from your data, or from your mood?
Conceptual integration. Can you name the theory that explains what you saw?
Professional presentation. Would a client read this without wincing?
Evidence isn't optional. Purchased engagement and fabricated or misrepresented metrics are academic integrity violations, not grading matters.
One client brief for the whole class. One fixed budget. You and a partner write 2,000 words plus an appendix answering one question: who should this client pay, and why?
who plausibly fit the brief, with the authenticity and brand-safety signals you checked.
with the budget allocated between them and each rate benchmarked.
against the ones you didn't pick. That's where the marks are.
Assigned Session 8 · due at the start of Session 10
The lab is your attendance. It can't be made up afterwards.
Real posts, real numbers, real decisions. Small and true beats big and bought.
Dated screenshots every week, not a scramble in November.
Office hours, the Q&A box, the clinic. Early questions beat late consequences.
A hypothesis that fails and is understood scores higher than one that succeeds and isn't.
AI use, paid help, borrowed ideas. Disclosure is never penalized. Silence is.
The account clinic opens each session. Bring your problem, get an answer in front of everyone who has the same one.
Book through the site. Confirmation and reminder come automatically.
Anonymous questions that help more than one person get answered on the site within a few days.
Local and international creators dropping in to say what actually happened, schedules permitting.
Inside the ten-day window, no questions asked. After that: 10% immediately, 10% per further day.
Wait 24 hours, then email within two weeks. I'll re-read before we talk.
Three parts. Work alone, write it down, be honest. Nobody is judging your taste. The point is to find the thing you'd actually keep doing for eight weeks.
Three things you are most likely to become an influencer of. Specific beats broad: "budget izakaya in Kowloon" is a niche; "food" is not.
1. ________________
2. ________________
3. ________________
Not where you already are, but where this niche is actually consumed. Think about format (long, short, text, image), language, and who's there.
Instagram · TikTok · YouTube / Shorts · Xiaohongshu · Threads · LinkedIn · Bilibili · Substack · Douyin · Reddit · Discord
For each niche, two or three creators doing well. Then the harder question: what specifically makes them work? Consistency? A format nobody else uses? Access? Timing? A voice?
"They're funny" is not an answer. "They post the same format every Tuesday and it's built for saves" is.
In Section 1's orientation, three volunteers did this out loud for sixty seconds each. You'll get the same chance in the account clinic at Session 1, and you can drop your answer in the Q&A box on the site any time before then.
and why you'd be credible in it
and what it rewards
and the specific thing they do well
I do mine in the room: four niches, the platforms I'd choose, and the creators I'd learn from. It goes up here after Section 1's orientation, so nobody gets a head start on the exercise by scrolling.
Do your own three first. That's the point.
I think X will happen because Y.
Do the thing. Note what you changed.
Read the native analytics. All of it.
Decide what's next, and say why.
↩ and back to 1, with what you learned. That loop is one decision log entry. Do it eight times and you've run a real experiment on a real distribution system.
| # | Sec 1 · Tue | Sec 2 · Mon | What to expect | Due |
|---|---|---|---|---|
| · | 1 Sep | · | Orientation and platform scouting (this deck) | |
| 1 | 8 Sep | 7 Sep | Digital PR now: from gatekeepers to algorithms. PESO and the two lanes. | |
| 2 | 15 Sep | 14 Sep | Algorithmic visibility and platform logics. | Brief assigned |
| 3 | 22 Sep | 21 Sep | Audiences, publics, communities. | Brief due · accounts live |
| 4 | 29 Sep | 28 Sep | Authenticity, credibility, parasocial relations. | AI Lab 1 assigned |
| 5 | 6 Oct | 5 Oct | Content craft and AI in the creative pipeline. | AI Lab 1 due |
| 6 | 13 Oct | 12 Oct | Risk, crisis, brand safety. | |
| · | 20 Oct | Chung Yeung | Supervised studio (Sec 1) · office hours (Sec 2) | Account checkpoint |
| 7 | 27 Oct | 26 Oct | MID-TERM EXAMINATION, in class | Exam · 10% |
| # | Sec 1 · Tue | Sec 2 · Mon | What to expect | Due |
|---|---|---|---|---|
| 8 | 3 Nov | 2 Nov | Measurement I: analytics, vanity vs outcome metrics, relational measurement. | Memo + AI Lab 2 assigned |
| 9 | 10 Nov | 9 Nov | Influence and opinion leadership: two-step flow to creator vetting. | AI Lab 2 due |
| 10 | 17 Nov | 16 Nov | Money, disclosure, comparative markets: HK, mainland, US. | Memo due · accounts close |
| 11 | 24 Nov | 23 Nov | Retrospective, synthesis, final exam preparation. | |
| · | 27 Nov | 27 Nov | PORTFOLIO SUBMISSION (Friday) | Portfolio · 25% |
| · | 7–19 Dec | 7–19 Dec | FINAL EXAMINATION, University examination period | Exam · 20% |
Note the pile-up around Session 10: memo due, accounts close, and the portfolio is due ten days later.
Top of the page. Your dates go bold everywhere.
This deck, plus the four things to do before Session 1.
Both sections, every deadline, in one table.
The four boxes with a word counter. Drafts stay on your device.
Guides unlock after the session that needs them. Ask anything, anonymously.
Twenty-five minute Monday slots. One click.
The site holds the teaching material. Moodle holds the record. If it's being marked, it goes through Moodle.
In the anonymous Q&A box on the site, or in the account clinic at Session 1. Anything about your own situation: book office hours or email.
CVA 928 · Mondays 14:00–17:00 · lolawxie@hkbu.edu.hk