Monthly Earner Product Architecture

Source Excerpt

Overview

Monthly Earner is Clark Farming Company's product line for delivering AI-driven business intelligence tools to small rural businesses. The product uses self-contained HTML dashboards as the primary deliverable format, generated from customer data through Hermes Agent processing pipelines. The target market is cash-heavy, low-digital-adoption businesses in and around Harmony, NC (ZIP 28634).

Target Market

Primary Targets (Top 5)

As of 2026-05-26, the confirmed target list is:

Previous targets Renegar and A&A were replaced in favor of larger, more digitally-receptive businesses. The full candidate list of 12 businesses (within 30 miles of Harmony) is maintained in ~/ClarkFarmingCompany/projects/Monthly Earner/Monthly_Earner_Targets.md.

Secondary Proof-of-Concept

An airplane cleaning business (RCAD RealClean Central Carolina) serves as the proof-of-concept customer. The owner has provided real business data and agreed to be a reference customer. This business has higher average ticket values (~$3,000) and provides a clearer ROI demonstration than the rural targets.

Deliverable Architecture

HTML-First Output

The decision was made to use self-contained HTML files as the standard output format for all AI-driven deliverables. Rationale:

• Immediately viewable — no special software needed; opens in any browser
• Easy to host — can be placed on a simple web server or shared as a file
• Fast to iterate — HTML is easy to generate and modify programmatically
• Professional enough — feels like a real product, not a raw report

Each deliverable is a fixed, standalone HTML file that does not depend on live data feeds or external services.

Master Style System

A centralized style system (master-styles.html) defines multiple visual themes. Each solution template pulls its styling from the master. Changing appearance for a customer becomes a one-time configuration choice rather than editing every template.

Four distinct themes:

Solution Templates

Five solution templates are planned for the first pass:

1. Customer Retention Dashboard — Shows which customers are at risk of churning, with risk levels, last service dates, and revenue impact.
2. Inventory Dashboard — Tracks stock levels, reorder points, and consumption trends.
3. Revenue Dashboard — Monthly/quarterly revenue trends with category breakdown.
4. Job Performance Report — Service job tracking with completion rates and revenue per job.
5. Lead & Pipeline Tracker — Prospects, conversion rates, and pipeline value.

Two additional designs are in progress:

• Quote & Proposal Generator (high priority — live lead exists with complex Excel quoting system)
• Weekly Business Health Report (builds on existing dashboard data)

Data Strategy

Rural-First Data Capture

The target market (Harmony, NC area) is characterized by cash-heavy transactions and low digital tool adoption. The data capture strategy emphasizes Tier 2 methods as the primary approach:

The Rural Data Capture Playbook (in project artifacts) details the full strategy. The core principle: never require new data entry processes as a prerequisite for delivering value. Start with what they already generate, then gradually layer in low-friction capture methods.

What Rural Businesses Already Track

For a typical target like Lowry's Livestock Feed:

Synthetic Data Reserves

Real data is supplemented with synthetic datasets stored in ~/ClarkFarmingCompany/library/synthetic-data/:

• Retail (800K rows, 100MB) — Transactions, customers, segments, CLV, loyalty, satisfaction
• Marketing (600K rows, 95MB) — Customers, churn risk, CLV, engagement
• E-commerce (500K rows, 78MB) — Orders, products, reviews
• SaaS (400K rows, 62MB) — Subscriptions, usage, health scores
• Airplane Cleaning — Real data from RCAD + synthetic gaps (jobs, invoices, payments)

Generation Workflow

1. Data intake — Customer provides data via Tier 1 or Tier 2 methods
2. Hermes processing — Data is analyzed, cleaned, and enriched
3. Template selection — Appropriate solution template is chosen
4. Style selection — Customer chooses from master style themes
5. HTML generation — Standalone HTML file is produced with customer data
6. Delivery — File is shared directly or hosted

Regeneration in a different style is possible by re-running the generation step.

Current Status

• Phase 1 (Data + First Tool): Complete. Customer Retention tool built and tested with RCAD airplane cleaning data. HTML dashboard generated.
• Phase 2 (Template System): In progress. Master style system created with 4 themes. Five solution templates planned.
• Phase 3 (Rural Data Strategy): In progress. Rural Data Capture Playbook created. Templates being revised for Tier 2 data alignment.
• Phase 4 (Quote Generator): Design complete. High priority due to live lead.
• Phase 5 (Weekly Health Report): Design complete. Builds on existing dashboard data.

Key Decisions

Strategic Clarification (2026-05-27)

We are not selling open-ended AI agents to small businesses. We are using AI internally to rapidly build, maintain, and improve discrete operational tools that solve specific business problems.

This is the operating paradigm:

• Not: "Install an open-ended AI agent inside the customer's business and let it figure things out."
• Instead: "Use AI to help us rapidly build, maintain, and improve small, discrete business tools that solve specific operational problems."

Examples of productized outputs:

• Python scripts that run continuously or on demand to perform specific tasks
• Self-contained HTML dashboards with customer-specific data
• Hermes agents that live on the customer's system for defined purposes
• Quote generators, reconciliation reports, health reports

This distinction matters for positioning, pricing, and delivery. Customers buy tools, not AI-as-a-service.

RCAD RealClean Proof-of-Concept Deliverables

As of 2026-05-26, seven mock HTML deliverables were created for RCAD RealClean (airplane cleaning):

All files in ~/ClarkFarmingCompany/projects/Monthly Earner/.

Time savings demonstrated:

• Weekly Health Report: 2–4 hours/week saved
• Quote Generator: 20–40 minutes per quote saved

Lowry's Livestock Feed — Primary Target

Location: Harmony, NC (0 mi from base)

Category: Feed / Country Store

Status: Discovery completed, MVP scoping in progress

Key Pain Points (from discovery call)

• Account balance visibility between the store and office is the clearest actionable business/process pain
• Employee reliability and lack of effort are frustrating but harder to address with tools
• Cash-heavy operations with low digital tool adoption

What Lowry's Already Tracks

MVP: Account Reconciliation Tool

• Feature map: lowrys-reconciliation-feature-map.md (MVP → Phase 2 → Phase 3)
• Mock deliverable: lowrys_reconciliation_report.html
• Success criteria: Owner can see account balances between store and office without manual reconciliation

Phased Plan

1. MVP: Account balance reconciliation report
2. Phase 2: Sales tracking dashboard
3. Phase 3: Inventory + seasonal planning tool

Open Questions

• How to handle recurring data updates from customers who primarily use Tier 2 methods?
• What hosting solution (if any) for customer dashboards?
• Pricing model for Monthly Earner services?
• Timeline from data intake to first deliverable?