← Sticker Wall AGENT · DETAIL

B2B Procurement Price Comparison AI Agent: Auto-broadcast RFQs saving 120k/month, commission-based monthly income 15k+

Workflow: Receive daily procurement demand orders (materials, quantities, specifications) sent by enterprises. The AI first parses

AGENT

Key Fields

FIELD STAMPS

IndustryAI / LLM

RegionChina

ScaleSME

ChannelOnline

🔧 Workflow

Receive daily procurement demand orders (materials, quantities, specifications) sent by enterprises. The AI first parses them into structured RFQs, then automatically broadcasts them to matching suppliers via corporate email, WeChat Work, or WhatsApp, while tracking and urging replies. After collecting quotes, it performs specification normalization and data cleaning, outputting cost-reduction reports (recommended suppliers, quote comparisons, estimated savings), leaving the final decision to the enterprise's procurement head. After orders are successfully placed, a commission is charged based on a percentage of the cost-reduction amount.

🛠 Setup Requirements

Requires knowing how to build workflows using n8n or Dify, writing code and supplementing scripts using DeepSeek or Kimi K3, and mastering WeChat Work and Email Webhook APIs. A self-built supplier profile database is also required. With basic programming skills, an MVP can be built in about 1-2 months, and then replicated across industries.

🧰 Toolchain

🔧 n8n
🔧 Dify
🔧 DeepSeek API
🔧 Kimi K3
🔧 WeChat Work Bot
🔧 Google Sheets

💰 Revenue

① Small and medium-sized factories pay a monthly basic service fee (main revenue): factories pay a basic service fee of 2,000 RMB/month × 3-5 long-term clients = 6,000-10,000 RMB/month, accounting for about 33%-45% of monthly revenue (this ratio is calculated by multiplying the unit price per card by the number of clients, reported by the merchant themselves, without independent verification); ② Cost-reduction commission: enterprises pay a 10% commission based on the price reduction amount. A single client saves 120,000 RMB/month → commission is about 12,000 RMB/month, accounting for about 55%-67% of monthly revenue (also calculated by multiplying unit price by quantity, self-reported by the merchant, without independent verification); ③ Vertical industry quotation data subscription: monthly subscription to quotation benchmark data for factories in the same industry. Neither the subscription unit price nor the number of clients is public, and the proportion is not broken down; ④ Opportunity item - Front-loading procurement decision positioning: 68% of B2B buyers have determined their preferred supplier before officially starting the procurement process, and this supplier has an 80% win rate. Early intervention can increase the order conversion rate. There are no figures on how much revenue this route can generate (media figures, pending independent verification), and its proportion is not specified either.

💸 Cost

API usage fees for DeepSeek and others are about 200-600 RMB per month, the n8n community edition is free, and basic functions of WeChat Work and Email APIs are free, keeping the total cost under 1,000 RMB per month.

⏱ Time Investment

3-4 hours per day, mainly spent on RFQ parsing, urging supplier replies, and report reviews; once operations run smoothly, this is gradually compressed to 1-2 hours daily.

🚀 Getting Started

First, select a familiar industry (such as electronic components or packaging consumables), find 3 small and medium-sized factories to negotiate pay-for-results terms, manually run 10 price comparison tasks using Excel and email; then distill the workflow of these 10 tasks into prompts and scripts, integrate with the LLM API to automatically parse RFQs, and gradually replace manual operations.

🔑 Keys to Success

✅ Accumulation of vertical industry quotation data forming a barrier
✅ Accuracy of specification normalization and quote cleaning
✅ Human purchasers acting as the final judge to lower the trust barrier
✅ Pay-for-results allowing enterprises to try with zero risk

⚠️ 风险

⚠️ Supplier database obsolescence leading to an increased failure rate of inquiries, requiring continuous updates and verification
⚠️ Long payment collection cycles from enterprise clients; commission confirmation relies on proactive reporting by enterprises, which may shrink
⚠️ Low barriers in industry quotation information gaps; large platforms or peers cutting in with low prices make it easy to be substituted

📌 Real Cases

📌 An agricultural company used DeepSeek to write code to restructure its financial bookkeeping process, automating a workload that previously took 7 accountants 6-7 days, proving that outsourcing B-end processes to AI agents is a real trend
📌 Data from a B2B summit shows that 68% of buyers have determined their preferred supplier before officially launching the procurement process, and this supplier has an 80% win rate. Procurement price comparison agents must intervene early to have a chance to cut in
📌 A cross-border seller in Yiwu used an AI procurement price comparison agent to connect with packaging consumable suppliers, reducing packaging procurement costs by 11% within 3 months, saving over 20,000 RMB per month, and paying the agent a commission of 3,600 RMB/month based on the cost-reduction percentage

Sources