Frontier AI Marketing Lab 前沿 AI 营销实验室

Building new digital assets for the AI era
with Chinese DTC brands going global. 与中国 DTC 出海品牌，
一起构建 AI 时代的新数字资产。

How it works 运作方式
01

Limited Access限量开放

Every application has a quota. Quality over quantity.每个申请有名额限制。质量优先。

02

Completely Free完全免费

No fees during co-construction. We don't make money from tools.共建期间零费用。我们不靠工具赚钱。

03

Your Voice Shapes It你的反馈塑造它

Real tools get better from real use. Tell us what worked, what didn't — that's how we iterate.真正好用的工具，都来自真实的使用反馈。哪里好用、哪里难用——告诉我们，我们才能迭代。

"Free to use, limited access, built together.
We don't sell tools. We polish them with you." 「免费使用，限量开放，一起共建。
我们不卖工具，我们和你一起打磨。」

Vision 愿景

Crossing growth cycles
with Chinese DTC brands going global. 与中国 DTC 出海品牌，
一起跨越增长周期。

Going global isn't a one-season hit — it's a decade-long marathon. Cold start, scaling, multi-category, multi-market — every stage tests a brand's ability to understand data and wield technology. 出海不是一季爆款——是十年的马拉松。冷启动、规模化、多品类、多市场——每个阶段都考验品牌对数据的理解和运用技术的能力。

AI doesn't replace people. It amplifies a brand's own judgment. AI 不是替代人，是放大品牌的判断力。

Lab Tools 实验室工具

100x.pics
Active Creative Engine 主动创意引擎

The world's first active creative engine for DTC brands. AI decodes your brand DNA, then proactively generates ad creatives across all platforms — no prompt engineering required. One product photo in, full-platform assets out. 全球首个 DTC 品牌主动创意引擎。AI 主动解码品牌 DNA，看一眼产品图就知道该做什么，自动生成全平台广告素材。不需要你会写 prompt。

100x.pics →
dtc.doctor
User Journey Microscope 用户旅程显微镜

Pinpoint exactly where your visitors drop off, hesitate, or convert. A diagnostic microscope for DTC landing pages — identify leaks, optimise flows, and turn traffic into revenue with data, not guesswork. 精准定位访客在哪个环节流失、犹豫、转化。DTC 落地页的诊断显微镜——用数据而非直觉找出流量漏洞，把访客变成收入。

dtc.doctor →

Every brand will eventually own its Agent 每个品牌终将拥有自己的 Agent

What we deliver isn't a tool.
It's a brand's contextual assets in the AI era. 最终交付的不是工具。
是品牌在 AI 时代的上下文资产。
01 FRONTIER APPS Explore AI frontier 探索AI前沿 02 LIMITED CO-BUILD Co-build with brands 与品牌共建 03 REAL-WORLD POLISH Deploy in market 真实市场打磨 BRAND AGENT Contextual Assets 上下文资产 Every brand's AI-era competitive moat 每个品牌的AI时代竞争护城河
AEO · Agent Engine Optimization

Structured for Agents 为 AI Agent 结构化商品

Raw product data → structured knowledge → AI Agent discovers & recommends.
This is how your products get found in the agent era. 原始商品数据 → 结构化知识 → AI Agent 发现并推荐。
让你的商品在 Agent 时代被找到。

📦
Raw Product 原始商品
title·desc·image 标题·描述·图片
→
🔍
Entity Extract 实体抽取
attr·spec·scene 属性·规格·场景
→
📐
Structure 结构化
JSON-LD Schema
→
📡
Agent Feed Agent Feed
product graph API 知识图谱 API
→
🤖
AI Recommends AI 推荐
match → rec 匹配→推荐
1

AI Entity Extraction AI 实体抽取

LLM auto-extracts 30+ dimensions from raw product data: brand, category, use case, specs, target audience, skill level — filling gaps agents need for accurate matching. LLM 从原始商品数据中自动抽取 30+ 维度：品牌、品类、使用场景、规格、目标人群、难度级别——补齐 Agent 精准匹配所需字段。

2

Schema.org JSON-LD Schema.org 结构化

Auto-generate Google Merchant Center & Schema.org compliant Product markup, plus OpenAI/Anthropic Structured Outputs format — one product, multi-format export. 自动生成符合 Google Merchant Center 和 Schema.org 规范的 Product 标记，同时输出 OpenAI/Anthropic Structured Outputs 格式——一个商品，多格式分发。

3

Agent-Ready Feed Agent 就绪的 Feed

Output llms.txt / products.json / GraphQL feeds. AI agents query via Function Calling or MCP protocol — millisecond retrieval, always up-to-date. 输出 llms.txt / products.json / GraphQL 等产品 API，AI Agent 通过 Function Calling 或 MCP 协议实时查询——毫秒级返回，始终最新。

4

Multi-Agent Distribution 多 Agent 分发

One structured product record → simultaneously distributed to ChatGPT, Gemini, Perplexity, Kimi, Doubao — every major AI platform your customers use. 一份结构化商品记录 → 同时分发给 ChatGPT、Gemini、Perplexity、Kimi、豆包——你客户使用的每一个主流 AI 平台。

// Schema.org Product — agent-readable structured data {
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Backfire G2 Electric Skateboard",
  "brand": { "@type": "Brand", "name": "Backfire" },
  "offers": { "price": "349.00", "priceCurrency": "USD" },
  "aggregateRating": { "ratingValue": "4.7", "reviewCount": "876" },
  "additionalProperty": [
    { "name": "useCase", "value": "commuting, cruising" },
    { "name": "skillLevel", "value": "beginner" },
    { "name": "maxSpeed", "value": "38 km/h" },
    { "name": "range", "value": "20 km" }
  ]
}

↑ User asks ChatGPT "I want buy a electric skateboard" → Agent retrieves structured product data → recommends Backfire G2 with image, price, rating & reasoning. ↑ 用户在 ChatGPT 问 "推荐电动滑板" → Agent 检索结构化商品 → 返回 Backfire G2 带图片、价格、评分、推荐理由。

Join dtc.lab 加入 dtc.lab
Limited access · Co-construction application 限量开放 · 共建申请