Alibaba Cloud Media and Entertainment China Translated from Chinese

Focus Media (分众传媒)

Marketing Personalization

Focus Media needed a more scalable and data-driven way to design offline brand-marketing plans for large advertisers while also lowering the creative and strategic barrier for smaller advertisers that lacked dedicated marketing resources. Traditional service delivery depended on heavy human interpretation, making it harder to translate the company's media data and industry experience into fast, customized, repeatable AI-assisted workflows.

covers 400 million urban mainstream consumers
operates more than 3 million elevator terminals

Reported impact

Business outcome

Focus Media turned its offline media expertise into a reusable AI marketing engine. The vertical model broadened its service capacity for large accounts, lowered the barrier for smaller advertisers to access strategy and creative support, and created a new AI-driven workflow for brand insight, ad-copy generation, and delivery optimization across a network that already reaches hundreds of millions of urban consumers.

  • Customer Experience
  • Productivity Improvement
  • Revenue Growth

Solution

Focus Media worked with Alibaba Cloud to build a marketing vertical large model on Bailian and Bailian-Quanmiao. The foundation combines a domain media model, industry data, multi-agent workflows, Focus Media's historical ad data, and its own slogan-scoring and evaluation models. Through Bailian's training and optimization toolchain, the company continuously fine-tunes the stack so it can generate customer insight, advertising creativity, targeting strategy, and delivery-optimization recommendations in one workflow. On top of the model layer, Focus Media launched the Zhongzhi AI product and an AI chat assistant that use Tongyi for semantic understanding while calling the vertical model for specialized outputs. That gives sales and service teams a guided interface for brand diagnosis and plan generation, and gives smaller advertisers lower-cost access to copy generation, AI design, and offline media planning capabilities that previously required more manual strategic work.

Data flow

Historical advertising data, industry knowledge, and media-specific evaluation logic are ingested into the Bailian training stack, where the vertical marketing model is optimized and then exposed through Zhongzhi AI and chat-style interfaces. Brand requirements are interpreted by Tongyi, routed into the marketing model, and returned as audience insight, creative text, strategy guidance, and delivery recommendations.

Solution architecture

4 components · 3 layers
  1. Compute
    • Bailian-Quanmiao Supplies the domain media model, industry data, and multi-agent workflow foundation for the marketing solution.
    • Tongyi large model Handles semantic understanding and intent interpretation for user-facing marketing interactions.
  2. Serving
    • Zhongzhi AI Delivers customer insight, creative generation, strategy creation, and delivery optimization to advertisers and service teams.
  3. Orchestration
    • Bailian Provides the training, optimization, and deployment platform for Focus Media's marketing large-model stack.

Architecture clues

  • Focus Media contributes labeled historical ad data and internal scoring models as part of the optimization loop.
  • Multi-agent workflows are embedded directly into the model foundation rather than added only at the application layer.
  • The model stack mixes Tongyi for general understanding with a separate vertical marketing model for domain outputs.
  • The user-facing products expose both chat interaction and generated output for creative and strategy tasks.

Evidence from the source

Based on the large model, Focus Media launched Zhongzhi AI to help millions of small and midsize advertisers solve marketing problems and lower offline ad-delivery barriers. (original: 分众已基于大模型推出“众智 AI”产品,赋能千万中小广告主解决营销难题,降低广告投放门槛。)
Focus Media and Alibaba Cloud deployed a dedicated marketing large model by combining Bailian, Bailian-Quanmiao, media data, and multi-agent workflows. (original: 分众传媒携手阿里云百炼、阿里云百炼-全妙... 以阿里云百炼-全妙里沉淀的大传媒模型、行业数据及 Multi-agent 工作流为基础。)
The AI chat assistant uses Tongyi for semantic understanding and then calls the vertical model to provide brand insight and marketing-strategy recommendations. (original: AI 小智助手基于通义千亿大模型进行准确的语义理解和意图分类,并根据用户的具体提问场景调用“众智 AI 大模型”回答问题并多轮交互。)
The company stacked its own labeled advertising data and slogan-scoring model on top of Alibaba Cloud training tools to build a Focus Media-specific vertical model. (original: 叠加上分众对历史广告数据的人工精标和深度挖掘、分众独有的广告语打分和评估模型,借助阿里云百炼的各种训练工具持续迭代和优化,最终打造出分众自己的营销垂类大模型。)
Transparency Evidence & methodology 71/100 reference value · 93% confidence · 1 source

Review how this reference was scored and trace each normalized claim back to its documented source.

01

Scoring methodology

How scores are calculated

Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.

Customer reference value71/100
  • Problem clarity20/20
    Measures whether the customer's operational problem is specific enough to compare.
  • Implementation depth25/25
    Rewards named components, data flow, integrations, and architecture detail.
  • Outcome specificity16/20
    Measures whether the reported benefit is concrete rather than a generic claim.
  • Quantified impact0/25
    0 metrics describe measurable impact.
  • Deployment scale10/10
    Rewards evidence of a live or scaled deployment and meaningful integration scope.
Evidence confidence93%
  • Source authority20/25
    The main evidence is an official vendor-published customer story.
  • Claim traceability25/25
    12 of 12 key fields link to a named source.
  • Direct evidence20/20
    4 direct source quotes retained.
  • Date provenance10/10
    The publication date comes from the source page or publisher.
  • Source transparency10/10
    1 source reference document publisher, URL, authority, retrieval, and supported fields.
  • Implementation specificity8/10
    Rewards source-backed components, data flow, quantified impact, and a specific outcome.
02

Traceability

Sources & provenance

Field coverage shows which facts each source supports. Dates are shown only when documented by the source.

  1. Alibaba Cloud

    分众传媒携手阿里通义大模型,开拓品牌广告AI营销新模式_传媒与音视频_阿里云客户案例

    First-party
    https://www.aliyun.com/customer-stories/media-and-entertainment-2024-focusmedia
    Source type
    Vendor-published customer story
    Supports
    • Customer
    • Region
    • Company size
    • Business problem
    • Solution
    • Products
    • Solution components
    • Data flow
    • Integrations
    • technical area
    • use case category
    • Business outcome
    • outcome category
    • Metrics
    • Architecture
    • evidence quotes

Date evidence

  • Published on source page Aug 30, 2024 · Alibaba Cloud · First-party
    Open date source
Not separately sourced Industry. Treat these normalized attributes as lower-confidence until a dedicated source is added.