Interview demo

Internal AI Marketing Operations Demo

Two-month MVP concept for reducing manual admin, reporting and creative workload.

Static JSON dashboardHuman review requiredNo auto-publishing

Business goal

July lead-generation pilot for service-business webinar offers

This page shows how AI PM Radar can be repositioned as an internal marketing operations dashboard for service businesses without adding a backend or exposing client data.

Client type

Course instructorsHair salonsImage consultantsCoaches and service professionals

Funnel metrics

Campaign funnel snapshot

Lightweight operations metrics for admin follow-up, response quality, and attendance tracking.

Reach24,800

Clicks1,680

Group joins412

Messages received236

Messages replied189

Registrations121

Attendance74

Spend$960.00

Derived KPIs

Operational health indicators

Interview-safe examples of the KPI layer a PM could define before any full system build-out.

Group join rate24.5%

Message reply rate80.1%

Registration rate51.3%

Attendance rate61.2%

Cost per lead$7.93

Cost per attendee$12.97

Flow view

Funnel handoff map

Clear enough for an interview demo, still honest about the static MVP boundary.

Reach24,800

Clicks1,680

Group joins412

Messages received236

Messages replied189

Registrations121

Attendance74

Spend$960.00

Creative review

AI-assisted creative review board

Sample analysis only. Suggestions are drafted outputs that still require human approval before reuse.

CR-01

Short-form video

Hook: Why are leads watching but not booking?

Caption: Three fixes service businesses can test this week before paying for more traffic.

Visual: Presenter talking to camera with a clean metric overlay and webinar promise banner.

First 3 seconds: Fast cut from ad spend screenshot to close-up question headline.

Performance summary: High click-through rate, but weak group-join conversion after the first landing step.

Possible issue: The promise is clear, but the transition from click to join may feel too operational and not outcome-led.

AI suggestion: Test a second variant that shows a real customer scenario before the admin checklist.

Next action: Draft a more benefit-first opening and compare join rate over the next 3 days.

CR-02

Static image carousel

Hook: Can one checklist improve message reply consistency?

Caption: A simple AI-assisted reply framework for salons, coaches, and course businesses.

Visual: Three-panel carousel with inbox screenshots, checklist cards, and CTA card.

First 3 seconds: First card shows unread messages and response delay pain point.

Performance summary: Lower clicks than video, but stronger message quality once prospects reach the admin team.

Possible issue: The CTA card may be too text-heavy on mobile and could reduce swipe completion.

AI suggestion: Shorten CTA copy and add a clearer before-versus-after proof frame.

Next action: Create a lighter final panel and review completion rate in the next carousel batch.

AI output preview

Weekly report draft

Traffic quality is acceptable, but the main drop-off still happens between click and group join.
Reply discipline is relatively strong once prospects message the team, suggesting admin workflow support is more urgent than more traffic volume.
Next weekly experiment should compare outcome-led creative framing against process-led framing.

Ad copy ideas

Your leads are clicking. The problem starts after the click. Here is the admin workflow fix.
One AI-assisted reply checklist for coaches, salons, and instructors who need faster follow-up without messy automation.
Before buying more traffic, fix the join-to-message handoff.

Deck and follow-up support

PPT outline

Campaign objective and audience segments
Funnel performance snapshot and derived KPIs
Creative findings and likely bottlenecks
Recommended next experiments and human review gates

Follow-up suggestions

Review message templates with a human operator before reuse.
Log which creative variants improve group join rate, not only click-through rate.
Keep personally identifiable data out of prompt examples and dashboards.

Risk controls

Human review required before any copy is published or sent.
No sensitive personal data is stored in this demo.
No auto-publishing or auto-sending actions are enabled.
All sample data is anonymised for interview-safe discussion.
API keys are not committed to the repository.
Role-based access is recommended if this concept is rebuilt for production.

Demo boundaries

Portfolio demo only, not a production SaaS platform.
Uses anonymised sample data stored in local JSON.
No backend, no login, no live client data, and no auto-publishing.