- **Day 7 behavior signals real intent, not passive browsing.** A completed post-video survey is a deliberate time investment and self-disclosure; it reveals training needs, budget context and decision role — stronger than views or video watches. 完成视频后问卷是主动投入时间并披露需求/角色/预算信号，比单纯观看更能说明真实采购意向。 - **Score rose 44→48, but stays below the 60 C→B threshold.** The AI weights behavior type and recency; one high-value action adds ~4 points, not a pool jump. Staying in C is correct — premature upgrade inflates pipeline with unqualified deals. 分数从44升到48，仍低于60的升级线；单次高价值行为只加约4分，留在C池可避免把未成熟线索提前塞进销售管道。 - **Trigger precise, low-cost nurturing — not a proposal yet.** Actions: tag survey answers into the lead profile, send a tailored content follow-up (e.g., 100-seat program case study), set a 3-day re-score task, and alert sales only as a "watchlist" note. 触发低成本精准培育：把问卷答案写入画像、推送定制内容（如100人项目案例）、设3天复评任务，销售仅作“观察名单”提示，暂不发正式方案。 - **Gate the B2B proposal and sales follow-up on the B-pool crossing.** Once score ≥60 (e.g., repeat visit + content download), auto-generate the 100-unit distribution proposal and create a sales follow-up task with SLA. 方案与销售跟进应由B池门槛触发：分数≥60（如再次访问+资料下载）时自动生成100套分销方案并建销售跟进任务。 - **Business outcome: protects conversion rate and sales efficiency while keeping the 100-unit inquiry warm.** Correct pool discipline means sales time goes to qualified leads, and this account keeps accumulating intent signals toward a larger, better-informed deal. 业务结果：守住转化率与销售效率，同时让100套询盘持续升温，为更大、更成熟的订单积累意向证据。