Skip to content

PRODUCT.md

Latest commit

 

History

History
298 lines (215 loc) · 7.83 KB

PRODUCT.md

File metadata and controls

298 lines (215 loc) · 7.83 KB

Retenu - Product Documentation (Honest Version)

What This Is

Retenu is a tool I'm building to help agencies find money they're losing to billing mistakes. It's currently a working demo/MVP, not a production SaaS yet.

Current Status: Demo with local data storage. Core detection logic works. Some integrations started but not complete.

The Problem (This Part Is Real)

Service businesses lose money in ways they don't notice:

1. Unbilled Hours

Team logs 40 hours for Client X. Invoice goes out for 35 hours. Nobody notices the gap.

Why it happens:

  • Time gets logged but forgotten when invoicing
  • "Quick calls" and emails never tracked
  • Confusion about what's billable

2. Scope Creep

Contract says 10 hours/month. Client actually uses 14 hours. Agency eats the difference to "keep the relationship."

Why it happens:

  • No automated tracking against limits
  • Awkward to push back on clients
  • "It's just a few extra hours"

3. Late Payments

Invoices sit at 45, 60, 90 days overdue. Nobody follows up systematically.

Why it happens:

  • No automated reminders
  • Embarrassment about chasing money
  • Invoices get lost in email

4. Missing Invoices

Project finishes. Team moves to next thing. Final invoice never sent.

Why it happens:

  • No system to check "did we bill for this?"
  • Handoff problems between project and finance

How Much Money? I Don't Know Yet.

I've read that agencies lose 4-10% of revenue to these issues. But I can't prove that number. I don't have customer data yet to validate it.

What I can say: If you've ever forgotten to bill for something, or had a client go months without paying, you know the problem is real.

Who This Is For

Ideal user:

  • Digital/creative agency or consultancy
  • Bills clients by the hour or on retainer
  • Has 5-50 active clients
  • Uses some kind of time tracking (even spreadsheets)
  • Doesn't have a full-time finance person

Not for:

  • Product companies (no hourly billing)
  • Agencies that only do fixed-price projects
  • Large enterprises with finance teams

What Retenu Actually Does (Today)

Working Features

1. Dashboard

  • Shows total "revenue at risk" (sum of detected issues)
  • Breaks down by issue type
  • Lists active alerts

2. Client Management

  • Add clients with retainer amounts, hourly rates, hour limits
  • Track client status (active/inactive)
  • Set per-client financial settings

3. Time Entry Tracking

  • Manual entry
  • CSV upload (works)
  • Toggl import (partially built - API connection exists)

4. Invoice Tracking

  • Create and track invoices
  • Mark as draft/sent/paid/overdue
  • See which clients have outstanding balances

5. Detection Engine This is NOT AI. It's conditional logic that compares numbers:

Unbilled Hours Alert:
IF (hours logged for client) > (hours invoiced for client) by more than X%
THEN create alert with $ amount = gap × hourly rate

Scope Creep Alert:
IF (hours worked this month) > (client's hour limit) by more than X%
THEN create alert

Late Payment Alert:
IF (invoice due date) was more than X days ago AND status != paid
THEN create alert

Missing Invoice Alert:
IF (client has billable hours this month) AND (no invoice exists for this month)
THEN create alert

That's it. It's math, not machine learning. But it still finds real problems.

6. Stripe Integration

  • OAuth connection flow exists
  • Can sync payment status from Stripe
  • Webhook handling for payment events

What's NOT Built Yet

  • QuickBooks integration (not started)
  • Xero integration (not started)
  • FreshBooks integration (not started)
  • Harvest integration (not started)
  • Clockify integration (not started)
  • Slack notifications (not started)
  • Email alerts (not started)
  • SMS alerts (not started)
  • One-click invoice generation from alerts (not started)
  • Team member management (basic only)
  • Reports/exports (not started)
  • Mobile app (not started)

How It Works (Technical)

Data Flow

User enters data (or imports CSV/Toggl)
         ↓
Data stored in browser (demo mode) or Supabase (production)
         ↓
Detection engine runs (on page load or manual refresh)
         ↓
Compares: time entries vs invoices vs client settings
         ↓
Creates alerts for detected issues
         ↓
Dashboard displays alerts and totals

Tech Stack (Real)

  • Frontend: Next.js 14, React, Tailwind CSS
  • Backend: Next.js API routes, Supabase
  • Database: PostgreSQL (via Supabase) or localStorage (demo)
  • Payments: Stripe (for connected accounts)
  • Hosting: Not deployed yet (runs locally)

Data Storage

Demo Mode:

  • All data in browser localStorage
  • Pre-populated with fake sample data
  • Good for trying the product

Production Mode:

  • Supabase PostgreSQL database
  • Row-level security for multi-tenant
  • Requires Supabase account setup

What Users Need to Provide

Minimum to Use

  1. Clients - Name, monthly retainer (if any), hourly rate
  2. Time entries - Client, hours, date, billable flag
  3. Invoices - Client, amount, dates, status

How to Get Data In

Method Status Notes
Manual entry Works Tedious but functional
CSV upload Works For time entries
Toggl API Partial Connection exists, sync is buggy
Other integrations Not built On roadmap

Security & Compliance

What I Can Honestly Say

  • Passwords hashed with bcrypt (via Supabase Auth)
  • HTTPS in production
  • Data stored in Supabase (their security, not mine)
  • No SOC 2 certification
  • No ISO 27001
  • No security audit completed
  • I'm a solo developer, not an enterprise vendor

What This Means

If you're an agency handling client financial data, you should know:

  • This is early-stage software
  • I haven't had a professional security review
  • For now, use demo mode with fake data to evaluate
  • Production use is at your own risk until I mature the product

Pricing

Current: Free (it's a demo)

Future plans (not implemented):

  • Probably freemium model
  • Free tier for small usage
  • Paid tier for more clients/integrations
  • Haven't figured out exact pricing yet

What I'm Trying to Validate

Before building more, I want to know:

  1. Is the core problem real? Do agencies actually lose significant money this way?
  2. Does the detection logic catch real issues? When pointed at real data, does it find things?
  3. Would agencies pay for this? And how much?
  4. What integrations matter most? Toggl? QuickBooks? Something else?

If you're an agency owner testing this, I'd love your feedback:

  • Did it find real Retenus in your data?
  • What features are missing that you'd need?
  • Would you pay for this? How much?

Roadmap (Honest)

Near Term (Next 1-2 Months)

  • Fix Toggl integration bugs
  • Add Clockify integration (seems popular)
  • Email notifications for alerts
  • Basic reporting/export

Medium Term (3-6 Months)

  • QuickBooks integration
  • Stripe payment sync improvements
  • One-click recovery actions
  • Team member tracking

Long Term (Maybe)

  • Mobile app
  • More accounting integrations
  • Actual AI/ML for pattern detection
  • Enterprise features

Might Never Build

  • Things that don't get requested
  • Features that add complexity without value

Glossary (No BS)

Term What It Actually Means
Retenuage Money you should have collected but didn't
Detection Engine If/then logic comparing your numbers
Alert "Hey, this looks wrong" notification
Scope Creep Client getting more than they paid for
Recovery Actually collecting the money you're owed

Contact

This is built by a solo developer.

  • GitHub: [project repo]
  • Email: [your email]

I read every message. Feedback helps me build something actually useful.

Last Updated: March 2024 Status: Demo/MVP - Not Production Ready