电商广告高阶优化

平台内广告竞争激烈，出价只是入场券。真正决定 ACOS 的是商品页质量、关键词结构和自动化投放策略的组合拳。

电商广告优化框架

关键词竞价管理器

from dataclasses import dataclass
from enum import Enum
class MatchType(Enum):
EXACT = "exact"
PHRASE = "phrase"
BROAD = "broad"
@dataclass
class EcomKeyword:
"""电商广告关键词"""
keyword: str
match_type: MatchType
current_bid: float
acos: float         # 广告销售成本比 (%)
impressions: int
clicks: int
orders: int
@property
def ctr(self) -> float:
return self.clicks / self.impressions if self.impressions > 0 else 0
@property
def cvr(self) -> float:
return self.orders / self.clicks if self.clicks > 0 else 0
class KeywordOptimizer:
"""关键词出价优化器"""
def __init__(self, target_acos: float = 25.0):
self.target_acos = target_acos
def suggest_action(self, kw: EcomKeyword) -> dict:
"""根据表现推荐操作"""
if kw.impressions < 100:
return {"action": "观察", "reason": "数据不足，继续积累"}
if kw.acos <= self.target_acos * 0.7:
# 表现优秀——提高出价抢更多流量
new_bid = round(kw.current_bid * 1.2, 2)
return {
"action": "提价",
"new_bid": new_bid,
"reason": f"ACOS {kw.acos:.1f}% 远低于目标，可扩量",
}
elif kw.acos <= self.target_acos:
return {"action": "保持", "reason": "ACOS 在目标范围内"}
elif kw.acos <= self.target_acos * 1.5:
# 略微超标——降价
new_bid = round(kw.current_bid * 0.85, 2)
return {
"action": "降价",
"new_bid": new_bid,
"reason": f"ACOS {kw.acos:.1f}% 略超标，适度降低",
}
else:
# 严重超标——暂停或加否定
if kw.orders == 0:
return {"action": "暂停", "reason": f"ACOS {kw.acos:.1f}%，零转化"}
new_bid = round(kw.current_bid * 0.6, 2)
return {
"action": "大幅降价",
"new_bid": new_bid,
"reason": f"ACOS {kw.acos:.1f}% 严重超标",
}
def batch_optimize(
self, keywords: list[EcomKeyword]
) -> list[dict]:
"""批量优化"""
results = []
for kw in keywords:
suggestion = self.suggest_action(kw)
suggestion["keyword"] = kw.keyword
suggestion["match_type"] = kw.match_type.value
results.append(suggestion)
return results
# 使用示例
optimizer = KeywordOptimizer(target_acos=25.0)
keywords = [
EcomKeyword("蓝牙耳机", MatchType.EXACT, 2.5, 18.0, 5000, 150, 12),
EcomKeyword("无线耳机 降噪", MatchType.PHRASE, 1.8, 45.0, 3000, 80, 3),
EcomKeyword("耳机", MatchType.BROAD, 1.0, 60.0, 10000, 200, 2),
]
for result in optimizer.batch_optimize(keywords):
print(f"{result['keyword']} ({result['match_type']}): {result['action']} — {result['reason']}")

分时段预算分配

# 不同时段的转化效率差异巨大
HOURLY_MULTIPLIER = {
# 时段: (出价系数, 预算占比)
"00-06": (0.5, 0.05),   # 凌晨低谷
"06-09": (0.8, 0.10),   # 早通勤
"09-12": (1.2, 0.20),   # 上午高峰
"12-14": (1.0, 0.15),   # 午休
"14-18": (1.1, 0.20),   # 下午
"18-21": (1.3, 0.20),   # 晚间黄金
"21-24": (0.9, 0.10),   # 夜间
}
def calculate_hourly_budget(daily_budget: float) -> dict:
"""按时段分配预算"""
result = {}
for period, (multiplier, ratio) in HOURLY_MULTIPLIER.items():
result[period] = {
"budget": round(daily_budget * ratio, 2),
"bid_multiplier": multiplier,
}
return result

电商广告平台策略对比

本章小结

• ACOS 驱动决策——低于目标 70% 提价扩量，高于 150% 暂停止损
• 三级匹配——精确词守转化、词组词拓流量、广泛词挖新词
• 分时段投放——晚间 18-21 点是黄金时段，凌晨降到 50%
• 否定词——广泛匹配必须配合否定词，避免垃圾点击
• 数据驱动——至少 100 次展示以上再做优化决策

下一章：KOL 合作与软广