Files
super-search-business/docs/super-search-business-spec.md
T
root 9af2d3ab62 feat: Super Search for Business — project inception
Product spec, architecture diagram, README, CHANGELOG.
Multi-tenant competitive intelligence SaaS built on Super Search v2.
Pricing: Starter 99/mo, Pro 99/mo, Enterprise ,499+/mo.
2026-07-25 14:13:47 -04:00

8.2 KiB

Super Search for Business — Product Specification

Status: Draft v1.0 · July 25, 2026 Author: Sho'Nuff (Hermes Agent) for Germaine Brown / IT Pro Partner


1. Product Overview

Super Search for Business is a multi-tenant competitive intelligence and OSINT search platform sold as a SaaS subscription. Companies subscribe to gain continuous visibility into their competitive landscape: who their competitors are, what customers say about them, what the market is looking for, and actionable intelligence from public records.

What it is: A business search engine that does the work of a competitive intelligence analyst — for $199-1,499/mo instead of $5k+/mo.

What it isn't: SEO keyword research (Semrush), web traffic estimation (SimilarWeb), or a generic search tool. This is investigative business intelligence.

Core value proposition

"Know everything the public internet knows about your competitors, your reputation, and your market — continuously, automatically, and without hiring an analyst."


2. Target Market

Primary

  • SMB service businesses (legal, medical, trades, agencies) who can't afford a dedicated analyst
  • Small consulting firms doing competitive research for clients
  • Debt recovery / skip tracing firms (existing DRE workflow)

Secondary

  • Mid-market companies wanting continuous market monitoring
  • Franchise operators tracking local competitive dynamics
  • Law firms doing due diligence research

Buyer persona

"Mike" — Operations director at a 50-person HVAC company. He knows there are 4 competitors in his market but has no systematic way to track their pricing, reviews, hiring, or expansion. He'd pay $199/mo to get a monthly report instead of spending 4 hours Googling.


3. Feature Matrix by Tier

Feature Starter ($199) Pro ($499) Enterprise ($1,499+)
Competitive search queries 50/mo Unlimited Unlimited
Web extraction (full pages) 25/mo 100/mo Unlimited
News monitoring
Sentiment analysis (reviews) Basic Advanced Custom models
Saved searches 5 25 Unlimited
Scheduled reports Monthly Weekly Daily + real-time alerts
Report export PDF PDF + CSV PDF + CSV + API
OSINT person/business dossier 5/mo Unlimited
Court records search
Skip tracing
White-label domain
API access
Custom dashboard
Multi-user accounts 1 3 Unlimited
Dedicated analyst summary

4. Technical Architecture

See super-search-business-architecture.html for the visual diagram.

Stack

Layer Technology Status
Search engine Super Search v2 (FastMCP, 7 providers) Built
OSINT enrichment osint-person MCP server Built
API gateway Caddy reverse proxy Running
Portal frontend React + TypeScript (new) 🔨 To build
REST API FastAPI wrapping MCP tools 🔨 To build
Multi-tenant DB PostgreSQL (new schema) 🔨 To build
Auth OAuth/SSO (existing pattern from Ops Portal) 🔨 Adapt
Billing Stripe subscriptions 🔨 Integrate
Hosting app1 (Hetzner) or netcup VPS 🔨 Deploy
Reports Markdown → PDF pipeline (pandoc/weasyprint) 🔨 To build
Scheduler Hermes cron → report generation 🔨 Configure

What's already built (80%)

Super Search v2 already has:

  • 7 providers with circuit breakers and health checks
  • Web extraction (single + batch)
  • Racing search (fastest provider wins)
  • Premium search (Exa API)
  • Academic, news, image search
  • Web lookups (DuckDuckGo Instant + Wikipedia)
  • Screenshot capture (browserless Chrome)
  • OSINT person search with identity clustering, source scoring, dossier reports
  • All accessible via MCP protocol on :8899

What needs building (20%)

Component Effort Priority
Multi-tenant REST API wrapper 2-3 days P0
Customer portal (React) 3-5 days P0
PostgreSQL schema (tenants, users, history) 1 day P0
Stripe integration 2 days P1
Auth (OAuth + API keys) 1 day P1
Report generation pipeline 2 days P1
White-label domain routing 1 day P2
Scheduled search/monitoring 2 days P2

5. Implementation Phases

Phase 1: MVP (2 weeks)

  • REST API wrapping Super Search MCP tools
  • Single-tenant portal (validate the concept with 1-2 beta customers)
  • Manual billing (Stripe invoice links, no integration)
  • Basic report generation (markdown → email)

Phase 2: Multi-Tenant Launch (2 weeks)

  • PostgreSQL schema with tenant isolation
  • Stripe subscription integration
  • Customer signup flow (14-day trial)
  • White-label domain support
  • Saved searches + history

Phase 3: Advanced Features (ongoing)

  • Scheduled monitoring with alerts
  • Sentiment analysis pipeline
  • OSINT dossier automation
  • API access for Enterprise tier
  • Custom dashboard builder

6. Pricing & Unit Economics

Subscription Pricing

Tier Monthly Annual (2 mo free) Target CAC
Starter $199 $1,990 < $200
Pro $499 $4,990 < $500
Enterprise $1,499+ Custom < $1,000

Cost Structure (per customer, estimated)

Cost Driver Starter Pro Enterprise
Exa API (premium search) ~$5/mo ~$15/mo ~$50/mo
Firecrawl (web extraction) ~$3/mo ~$10/mo ~$30/mo
Server compute (netcup VPS) Shared Shared Dedicated
Report generation (LLM) ~$2/mo ~$5/mo ~$20/mo
Total marginal cost ~$10/mo ~$30/mo ~$100/mo
Gross margin 95% 94% 93%

Break-even

  • 10 Starter customers = $1,990/mo revenue vs. ~$200/mo infra cost
  • Infrastructure already paid for (Hermes, app1) — marginal cost only

7. Competitive Landscape

Competitor What they do Our advantage
Semrush SEO keyword + domain analytics We do actual competitive intelligence, not SEO metrics
SimilarWeb Traffic estimation We pull from court records, corporate filings, reviews
Crayon Enterprise competitive intel ($20k+) Same capability at 1/10th the price
Klue Competitive enablement We're self-serve, not sales-enablement
Custom analyst $5-15k/mo for one analyst We're automated, always-on, $199/mo

Positioning: "Competitive intelligence" tools exist — they're just either SEO tools mislabeled, or $20k+/yr enterprise suites. The SMB market is completely unserved.


8. Go-to-Market

Channel 1: Existing ITPP customers

  • Liberty Tire, Forefront Wireless, DRE — existing relationships
  • Offer them free beta access in exchange for testimonials

Channel 2: Content marketing

  • Publish sample competitive reports for well-known industries (HVAC, legal, dental)
  • "We analyzed the top 5 HVAC companies in Savannah — here's what we found"
  • LinkedIn + industry subreddits

Channel 3: Partnerships

  • Business coaches / consultants who resell to their clients
  • MSPs who want to offer competitive intel as a value-add

Launch sequence

  1. Build MVP (2 weeks)
  2. Beta with Liberty Tire + 1-2 others (free, 30 days)
  3. Collect testimonials + refine based on feedback
  4. Public launch with 3 tiers
  5. Target: 10 paying customers by month 3

9. Open Questions

  1. Product name: "Super Search for Business" is functional but not marketable. Options: IntelSight, MarketPulse, BizScope, CompassIntel?
  2. Hosting: Share app1 resources or deploy to a dedicated VPS?
  3. Data retention: How long do we keep search history per tenant? (GDPR/CCPA implications)
  4. Rate limiting: What's a "search" in the Starter tier? One query = one provider call, or one "session" with multiple providers?
  5. LLM for reports: Use deepseek-v4-pro (cheap) or Claude (better summaries)? Both?

10. Next Steps

  • Germaine approves spec and picks a product name
  • Architect the PostgreSQL multi-tenant schema
  • Scaffold the FastAPI REST wrapper around Super Search MCP
  • Build the React portal shell with auth
  • Deploy to staging and invite beta customers