Growth ROI Layers

Growth ROI layers are the episode’s practical accounting frame for judging whether acquisition spending creates real value. In 发券、裂变、极速版，如何用红包设计增长？丨字节跳动 第8集, 徐鸿亮 / Tom labels the layers ROI1, ROI2, and ROI3: ROI1 compares LTV against CAC, ROI2 adds variable costs such as content or compute, and ROI3 adds fixed costs.

The frame extends LTV-Based Growth Budgeting by making cost structure explicit. In ad-supported feeds, user time can often be monetized with low marginal cost once the content and ad machine are in place. In AI products, every new user may consume tokens, GPU capacity, and electricity, so acquisition looks worse if model usage is heavy and willingness to pay is weak.

Key Claims

• ROI1 can make a campaign look attractive if LTV exceeds CAC, but it may hide content, compute, delivery, or reward costs.
• ROI2 is critical for products where marginal use is expensive, including AI assistants and some content businesses.
• ROI3 approaches a net-profit view by adding fixed costs, not only acquisition and variable cost.
• Red-packet, Lite-app, coupon, and AI growth campaigns should be judged against different ROI layers because their cost structures differ.
• AI consumer products need retention and usage, but more usage can destroy value if inference cost and monetization do not match.

Connections

• LTV-Based Growth Budgeting, AI Consumer Growth Metrics, AI Inference Cost Structure, and AI Commercialization Pressure — adjacent economics pages.
• Red Packet Growth, Lite App Growth, Coupon-Led Transaction Growth, and Spring Festival Growth Campaign — campaign types requiring ROI discipline.
• Doubao, DeepSeek, and Product Led Willingness To Pay — AI-era cases and product-value constraint.