AI Content Marketing System
for Leo&Loona
AI researches family interests and prepares post concepts. People review the facts, create visuals, approve and publish. Once performance sources are connected, results can inform the next cycle.
RESEARCH · DRAFT · REVIEW · LEARNHow the system works
The full blueprint links local discovery, content creation, human approval, publication and measured learning. This app currently implements the first two stages and previews the analytics stage.
Run the Scan
Research verified family events, dates, local news, parent questions and useful seasonal topics over the next two or three weeks.
Open Local Intelligence ↗Rank opportunities
Compare timeliness, family relevance, usefulness, conversation potential and evidence confidence. Scores guide review.
0–100 editorial estimatesGenerate Posts
Create image directions, full captions, source references and a recommended publication day and time.
Open Proposed Posts ↗04 / PLANNEDAnalyze results
Link post metrics to inquiries and bookings, compare what worked and test better ideas in the next cycle.
Explore Analytics ↗Clear roles for AI and the team
Automation helps the team discover and draft. People remain responsible for fact checks, visual quality, approval and publication.
AI handles
- Searching local calendars and family topics
- Organizing sources and supporting signals
- Ranking editorial opportunities
- Suggesting format, hook and publication timing
- Drafting captions and creative prompts
People handle
- Checking dates, venues and claims
- Editing tone for each market
- Creating or choosing the final image
- Approving captions and artwork
- Publishing externally and logging the link
Next integration
- Connect social account results
- Record website actions and lead sources
- Compare posts by market and format
- Test hypotheses in future batches
Weekly research, with a useful daily brief
The original blueprint covers trends, competitors, local events and audience needs. The current scan emphasizes sourced local opportunities; deeper competitor and platform trend monitoring is part of the broader system design.
Trends & formats
Notice useful hooks, visual formats and conversation patterns. Take inspiration from structures without copying viral posts.
Competitors & gaps
Compare themes and offers in the local family category, then look for parent questions and useful topics being missed.
Local calendar
Find holidays, school periods, fairs, cultural events, mall programs and seasonal planning moments across all three markets.
Audience needs
Turn parent pain points, birthday planning, educational interests and weekend choices into practical content angles.
Morning intelligence snapshot
Events and dates carry source links and confidence labels.
Understand why a topic matters now and which market it serves.
Review the local items behind the editorial ranking.
Every recommendation should be usable
The blueprint calls for a complete content package. The current app produces the most useful core: a format choice, image brief, ready caption, ratings, timing and a reason for each decision.
For the right audience
Market, platform, topic, objective and whether Leo&Loona should be mentioned directly, softly or not at all.
For the maker
An image prompt, composition notes and a suggested single image, carousel or video format.
For the reviewer
Full text, hook, alternatives and source links, with no factual placeholders in live recommendations.
For the publisher
A concrete local day and time, plus a plain explanation of why that window fits the topic.
Choose a strong single image most often. Use a carousel for distinct steps or comparisons and video when movement or a human reaction makes the idea better.
Generate Posts ↗From concept to scheduled post
Review, approve and publish
The source blueprint proposes Sprout Social for scheduling. This app prepares content for review; it does not approve or publish posts automatically.
Check
Confirm dates, claims, sources, local tone and whether the post remains useful to families.
Create
Use the prompt to make or commission an image, then inspect the final artwork and text.
Approve
Select the final version and align it with the team’s publishing calendar.
Schedule
Publish through the chosen tool and retain the post URL and campaign ID for later analysis.
Measure attention, action and outcomes
The blueprint’s dashboards track reach, engagement, clicks, inquiries and bookings. The app’s Analytics & Learning page currently contains a labeled sample preview; real values require connected sources.
Who saw it?
Impressions, reach, views, watch time and follower change.
Who cared?
Saves, shares, comments, profile visits and engagement rate.
Who acted?
Clicks, messages, birthday inquiries and school group requests.
What followed?
Attributed bookings and revenue only when the source trail is available.
KPI & conversion example
These numbers illustrate the planned dashboard. They are not current Leo&Loona results.
Explore Analytics & Learning ↗A practical tool stack
The attached blueprint proposed a research agent, ChatGPT, Sheets, creative tools, and a publishing platform. The current Cloudflare app uses its server-side AI worker for research and drafting; other services are options for the later workflow.
Eight connected views of the work
The source plan proposes one shared workbook. These tabs describe a possible record structure, not an existing integration in this app.
Build the learning cycle in stages
The blueprint laid out progressive setup, research, approval, publishing and reporting. The timing depends on access to social accounts, a publishing tool, website analytics and lead records.
Set rules and markets
Agree on voice, content pillars, competitors, cultural considerations and measures of success.
Research and draft
Run scans, review sourced opportunities, generate content and improve the prompts from feedback.
Review and publish
Set an approval workflow, generate final creative, schedule posts and log live URLs.
Connect and learn
Import performance data, connect inquiries and bookings, then test the next content choices.
Research what matters locally. Then create posts worth reviewing.
Analytics remains a preview until the relevant performance sources are connected.