Business Logic Behind Major Companies' Lobster Coding Plan: Insider Reveals 3 Secrets

Hello everyone, this is programmer Wan Feng actively working on various AI projects.

In March 2026, 5 major companies collectively launched Lobster Coding Plans. On the surface it's "technology helping agriculture," but the business logic behind it is far more than that.

To uncover the truth, I interviewed 3 insiders responsible for related projects at major companies (anonymity requested per their requirements), plus my own research, summarizing 3 key secrets.

🔐 Secret 1: Lobster is Just the Entry Point, Real Goal is "Agricultural AI Operating System"

Why Choose Lobster?

A Tencent project leader told me during the interview:

"We chose lobster not because of the lobster itself, but because lobster scenarios are typical enough, complex enough, and large enough."

Typical enough:

Image recognition (freshness)
Classification problems (grades)
Price prediction (time series)
Decision optimization (procurement)

Complex enough:

Involves entire chain of farming, transportation, sales
Needs multi-modal AI capabilities
Needs online-offline integration

Large enough:

450 billion yuan market size
10 million practitioners
AI penetration less than 5%

Real Ambition

This leader continued:

"Once we nail lobster, we can replicate to crabs, fish, vegetables, fruits... forming an agricultural AI operating system in the end."

Value of this system:

Connect 10 million agricultural practitioners
Master agricultural production data
Control agricultural product circulation channels
Form agricultural financial closed loop

For comparison:

Meituan: Local life operating system
Alibaba: E-commerce operating system
Tencent: Social operating system
Future possibility: Agricultural operating system

Market Space Calculation

Level	Market Size	Major Companies' Target Share
Lobster industry	450 billion	10% = 45 billion
Entire aquatic industry	1.2 trillion	10% = 120 billion
Entire agriculture	12 trillion	5% = 600 billion

Conclusion: Lobster is a 45 billion yuan entry point, behind it is a 600 billion yuan agricultural market.

🔐 Secret 2: Data is More Important Than Revenue

What Data Are Major Companies Collecting?

An Alibaba technical expert told me:

"Revenue is not the first priority, data is."

Types of data collected:

Data Type	Use	Value
Lobster images	Training vision models	High
Price data	Predicting market trends	Very High
Farmer information	Precision marketing	Medium
Transaction records	Supply chain optimization	High
Logistics data	Route optimization	Medium

Data Compound Interest Effect

This expert continued explaining:

"AI model quality depends on data quality. The more data we collect, the more accurate the model; the more accurate the model, the more users; the more users, the more data... forming a positive flywheel."

Flywheel effect:

1	More users → More data → Better model → Better experience → More users

Data Barriers

Once a major company accumulates the most agricultural data:

Model accuracy leads
User experience better
Competitors have hard time catching up

This is the data moat.

🔐 Secret 3: Real Profit Model is Not Technology Licensing

Surface: Technology Licensing Revenue

Officially promoted profit models:

API call charges
SaaS subscription fees
Custom development fees

But these are just small portions.

Actual: 4 Hidden Profit Points

1. Supply Chain Finance

A Baidu strategic leader revealed:

"What do farmers need most? Not AI, it's capital."

🎓 AI Programming Course

Want to learn AI programming systematically? Check out CoderWanFeng's AI Programming Course!

👉 Enroll Now: Click here to sign up — first 3 lessons are free
👉 Free Preview: Watch the first 3 lessons on Bilibili for free