The product I would launch in 30 days: a data coach that reads Instagram and tells creators what to repeat, what to change and what to drop.
- Status
- Proposal. Brand and specs ready, development not started. There's no demo yet.
The problem
More and more creators and businesses need to be on social media, but nobody teaches them to read the data: save rate, reach, like rate and retention stay meaningless numbers.
Existing tools collect the data and explain the metrics. None interprets them along the path of a single profile, against its goals and its niche.
What it does
Who it's for: Italian creators on Instagram, from 10 thousand to 1 million followers, who have the numbers but don't know what to do with them. Later, agencies and small businesses with an in-house social media manager.
The creator picks 2 or 3 goals and up to 3 niches. Every metric is compared with the niche, every post gets two scores and at least two pieces of advice, and a 30-day roadmap unlocks step by step. The tone is a coach's: informal, explains the why before the advice, never promises results.
How I built it
In 30 days the MVP would ship: sign-up and Instagram connection, home with the latest posts, a dashboard with one box per metric, a content ranking with advice and the 30-day roadmap. Instagram only, web app in Italian.
Out of the MVP, on purpose: TikTok, agency area, automatic publishing, payments and a native app.
Technical choices: the official Instagram API, which doesn't need a Facebook Page; Next.js and Supabase with Row Level Security; encrypted tokens; daily jobs. The code computes the numbers and Claude explains them without doing maths, so the advice stays verifiable.
Stack
- Claude Opus 5.5
- Claude Code
- Claude API
- Next.js
- TypeScript
- Supabase
- Row Level Security
- Vercel
- Instagram API
The 4-week plan
Each week has a goal I can verify, and one thing I cut if I'm late.
Week1
- What I build
- Sign-up, Instagram connection with official login, goals and niches, post import
- How I verify
- Real data from the test accounts is in the database
- If I'm late
- Niches picked from a fixed list, no suggestions
Week2
- What I build
- Home with the last 3 posts and a box dashboard: views, reach, save rate, like rate, followers, with niche comparison
- How I verify
- Testers see their own metrics, each with a definition and a comparison
- If I'm late
- Numbers instead of charts
Week3
- What I build
- Content ranking with two scores and advice written by Claude
- How I verify
- Advice quality reviewed on 20 real posts before opening the beta
- If I'm late
- No repeatable format, advice only
Week4
- What I build
- 30-day roadmap, landing page with waitlist and beta for the first creators
- How I verify
- At least 10 real creators connected
- If I'm late
- Roadmap generated once, without step-by-step unlocks
Planned plans
The yearly plan costs 10 months out of 12.
| Plan | What's included |
|---|---|
| Free | 7-day trial of Light |
| Light | 1 profile, advice on boxes and on 5 posts a month, overall benchmark score, 30-day roadmap |
| Pro | Instagram and TikTok, everything included, formats and 30/60/90 roadmap |
| Agency | 10, 15 or 20 clients, each with Instagram and TikTok |
| Enterprise | More than 20 clients, with a dedicated contact |
Gallery



