MOTA TECHLINK

🎯 Background & Goals1/13

PRD §1 · Project Background

Background & Goals

From passive exposure to active engagement — joining the conversation with genuine, valuable content the moment users signal their pain.

🚫

Traditional Acquisition Pain Points

✕Search ads keyword bidding: CPC keeps climbing in an endless bidding spiral
✕Display / feed ads: Force-feeding impressions to people who haven't expressed a need — low conversion
✕KOL campaigns: High upfront cost, unpredictable ROI
✕Core problem: Can't find people who are actively looking for a solution

✅

The MOTA Approach

✓Proactively discover real users who are publicly discussing problems and seeking help
✓Join the conversation under pain-signal posts with genuinely useful content
✓Product links appear as a natural extension, not hard-sell advertising
✓Pay only for actual results (clicks / conversions) — zero waste

Five-Step Closed-Loop Workflow (Hover for details)

📝

Onboard Product

Name + description + URL

Customer provides product name, description copy, product URL, and target market languages in the user dashboard — system auto-creates the project

🤖

AI Analysis

LLM infers pain points & keywords

Ollama/Gamma 4 analyzes the product → infers audience pain points → generates multi-platform search keywords → creates tasks and enqueues them

📡

Full-Web Scan

20+ platforms, continuous discovery

Scanner Agent cluster scans Reddit/YouTube/Google/Amazon/Quora/Twitter and 20+ other platforms 24/7 to find real posts discussing related problems

💬

Generate Reply

Value-added content + tracking link

AI generates a helpful reply + injects tracking links (mota_pid + mota_ref + UTM); optional human review before publishing

📊

Track & Bill

Pay per click / conversion

Tracking link clicked → funnel_events logged → Prepaid: real-time credit deduction / PAYG: end-of-month settlement → Admin can adjust invoices

Business Goals (Hover for details)

🎯

Product-First

Start from the customer's product/service — the system reverse-engineers target audience pain-point profiles. No research prompts required.

User only needs to provide product name, description, and URL LLM automatically infers domain tags, pain points, competitors, keywords Generated tasks cover search queries across all platforms

🌍

Full-Platform Coverage

Major English platforms + Spanish/Portuguese/Japanese/Korean/Russian local platforms — global market reach.

11 Scanner types cover 20+ platforms Ollama multilingual models support 6 languages Local platforms: Taringa!/Reclame Aqui/知恵袋/지식iN/Pikabu etc.

✨

Genuinely Valuable Replies

AI replies must contain truly useful content. Product links are a natural extension, not ads — minimizing ban risk.

Prompt rules: no direct advertising tone Quality score ≥ threshold for auto-publish Disguise detection + human review fallback

💰

Pay for Performance

Billed per actual posted reply click/conversion. Admins can apply discount adjustments to invoices.

Prepaid: top-up Credits with real-time deduction PAYG: monthly accumulation via Stripe card charge CPA: deep conversion events with weighted billing

🏗️

Three-Tier Architecture

Portal (public site), User Dashboard (project & billing management), Admin Dashboard (global operations).

Portal: SSR/SSG + SEO + registration funnel User Dashboard: Next.js + Recharts Admin Dashboard: Tanstack Table + full RBAC

8 Signal Capture Methods (Hover for details)

M1😤

Negative Search

Reddit · Google · Twitter

Find posts with a 'frustration' tone — users who've tried generic solutions but failed are prime candidates for alternatives

M2❓

Community Q&A Radar

Reddit · Quora · HN

Capture unanswered or poorly-answered questions where the product can be a better answer

M3🎥

Video Comment Mining

YouTube

Comment sections under competitor tutorials are a natural 'what's not good enough' database

M4⭐

Negative Review Targeting

Amazon · App Store · G2

Competitor negative reviews = potential target users for our product

M5⚡

Real-Time Sentiment

Twitter/X

Monitor keyword complaints in real time — be first to engage

M6🔧

Technical Q&A

Stack Overflow · SE

High-view unanswered technical questions are ideal for tool/SaaS products

M7🔗

Cross-Source Validation

Multi-platform

Same problem appearing on multiple platforms = stronger signal, higher match confidence

M8🌐

Multilingual Local

Taringa! · Reclame Aqui · 知恵袋 etc.

Cover Spanish/Portuguese/Japanese/Korean/Russian local communities to expand acquisition geography

1 / 13

Background & Goals

From passive exposure to active engagement — joining the conversation with genuine, valuable content the moment users signal their pain.

🚫

Traditional Acquisition Pain Points

✕Search ads keyword bidding: CPC keeps climbing in an endless bidding spiral
✕Display / feed ads: Force-feeding impressions to people who haven't expressed a need — low conversion
✕KOL campaigns: High upfront cost, unpredictable ROI
✕Core problem: Can't find people who are actively looking for a solution

✅

The MOTA Approach

✓Proactively discover real users who are publicly discussing problems and seeking help
✓Join the conversation under pain-signal posts with genuinely useful content
✓Product links appear as a natural extension, not hard-sell advertising
✓Pay only for actual results (clicks / conversions) — zero waste

Five-Step Closed-Loop Workflow (Hover for details)

📝

Onboard Product

Name + description + URL

Customer provides product name, description copy, product URL, and target market languages in the user dashboard — system auto-creates the project

🤖

AI Analysis

LLM infers pain points & keywords

Ollama/Gamma 4 analyzes the product → infers audience pain points → generates multi-platform search keywords → creates tasks and enqueues them

📡

Full-Web Scan

20+ platforms, continuous discovery

Scanner Agent cluster scans Reddit/YouTube/Google/Amazon/Quora/Twitter and 20+ other platforms 24/7 to find real posts discussing related problems

💬

Generate Reply

Value-added content + tracking link

AI generates a helpful reply + injects tracking links (mota_pid + mota_ref + UTM); optional human review before publishing

📊

Track & Bill

Pay per click / conversion

Tracking link clicked → funnel_events logged → Prepaid: real-time credit deduction / PAYG: end-of-month settlement → Admin can adjust invoices

Business Goals (Hover for details)

🎯

Product-First

Start from the customer's product/service — the system reverse-engineers target audience pain-point profiles. No research prompts required.

User only needs to provide product name, description, and URL LLM automatically infers domain tags, pain points, competitors, keywords Generated tasks cover search queries across all platforms

🌍

Full-Platform Coverage

Major English platforms + Spanish/Portuguese/Japanese/Korean/Russian local platforms — global market reach.

11 Scanner types cover 20+ platforms Ollama multilingual models support 6 languages Local platforms: Taringa!/Reclame Aqui/知恵袋/지식iN/Pikabu etc.

✨

Genuinely Valuable Replies

AI replies must contain truly useful content. Product links are a natural extension, not ads — minimizing ban risk.

Prompt rules: no direct advertising tone Quality score ≥ threshold for auto-publish Disguise detection + human review fallback

💰

Pay for Performance

Billed per actual posted reply click/conversion. Admins can apply discount adjustments to invoices.

Prepaid: top-up Credits with real-time deduction PAYG: monthly accumulation via Stripe card charge CPA: deep conversion events with weighted billing

🏗️

Three-Tier Architecture

Portal (public site), User Dashboard (project & billing management), Admin Dashboard (global operations).

Portal: SSR/SSG + SEO + registration funnel User Dashboard: Next.js + Recharts Admin Dashboard: Tanstack Table + full RBAC

8 Signal Capture Methods (Hover for details)

M1😤

Negative Search

Reddit · Google · Twitter

Find posts with a 'frustration' tone — users who've tried generic solutions but failed are prime candidates for alternatives

M2❓

Community Q&A Radar

Reddit · Quora · HN

Capture unanswered or poorly-answered questions where the product can be a better answer

M3🎥

Video Comment Mining

YouTube

Comment sections under competitor tutorials are a natural 'what's not good enough' database

M4⭐

Negative Review Targeting

Amazon · App Store · G2

Competitor negative reviews = potential target users for our product

M5⚡

Real-Time Sentiment

Twitter/X

Monitor keyword complaints in real time — be first to engage

M6🔧

Technical Q&A

Stack Overflow · SE

High-view unanswered technical questions are ideal for tool/SaaS products

M7🔗

Cross-Source Validation

Multi-platform

Same problem appearing on multiple platforms = stronger signal, higher match confidence

M8🌐

Multilingual Local

Taringa! · Reclame Aqui · 知恵袋 etc.

Cover Spanish/Portuguese/Japanese/Korean/Russian local communities to expand acquisition geography