网店动态商品广告投放规划
围绕网店商品目录制定投放方案,涵盖数据检查、商品分组、受众预算、创意模板与监测规则。
完整提示词
<context>
The user runs an online shop and wants catalogue (dynamic product) ads that show each person products from the feed, either products they viewed or products the platform predicts they will want. These ads fail quietly for technical reasons more than creative ones. What experts know: the product IDs in pixel and server events must match the catalogue IDs exactly, or retargeting shows the wrong items or nothing; price and stock must sync often enough that ads never promote sold-out items or old prices; product sets should be built around business logic (best-sellers, high margin, in stock, price bands, new arrivals) rather than every item; showing the product someone viewed works for retargeting, but catalogue ads to broad audiences can also prospect well when the catalogue is large; plain white-background product images look like ads, so frames, overlays with price or offer and lifestyle images help; and recent purchasers must be excluded or shown complementary items instead.
</context>
<task>
Store: {{store}}
Platform: {{platform}}
{{#catalog_status}}Catalogue status: {{catalog_status}}{{/catalog_status}}
Goal and budget: {{goal_and_budget}}
1. Readiness checks: list what must be true before launch (catalogue connected and syncing, error-free core fields, product ID match between events and catalogue, view content, add to cart and purchase events with IDs, price and availability updates) and how to verify each in the platform's diagnostics. If the status is unknown, put these first.
2. Product sets: propose 3 to 6 sets with the filter rule for each (category, margin label, price band, in stock, new, best-seller) and which campaign uses it.
3. Campaigns and audiences: a retargeting campaign split by intent stage if volume allows (viewed, added to cart, started checkout) with lookback windows, a broad prospecting catalogue campaign if the catalogue and budget support it, and a cross-sell set for past buyers. Give budget split and optimisation event.
4. Creative templates: frame or overlay ideas, which fields to show (price, discount, rating, free shipping), mixing catalogue and lifestyle or video, and primary text variants by stage.
5. Exclusions and frequency: purchasers by window, out-of-stock items, low-margin or returned-heavy items, and frequency to watch.
6. What to watch: metrics per set and campaign, signs of feed or ID mismatch (low match rate, irrelevant products shown), and decision rules.
</task>
<constraints>
- Do not state catalogue field requirements, image specs or match-rate thresholds as fact; tell the user to confirm in the platform's current commerce or catalogue documentation.
- Do not invent ROAS or conversion rates; use the user's targets.
- Do not recommend retargeting messages that reveal sensitive browsing (health, intimate or financial products) in a way the viewer would find intrusive; flag such categories.
</constraints>
<output_format>
## Readiness checks
Checklist with how to verify each item.
## Product sets
A table: Set | Filter rule | Size estimate | Used in.
## Campaigns and audiences
A table: Campaign | Audience and window | Product set | Optimisation event | Budget share.
## Creative templates
Bullets per campaign, with example primary text.
## Exclusions and frequency
Checklist.
## What to watch
Metrics, warning signs and decision rules.
</output_format>可替换内容
- {{store}}
- {{platform}}
- {{#catalog_status}}
- {{catalog_status}}
- {{/catalog_status}}
- {{goal_and_budget}}
分类与标签
电商运营、商品目录广告、再营销、商品分组、投放监测