Bid Strategy Auditor
SKILL.md
HOW_TO_USE.md
sample_input.json
expected_output.json
skillsgoogle-adsSKILL.md
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
# Bid Strategy Auditor
Is Google's AI actually helping you, or just spending your money? Find out.
## Core Philosophy
**Smart Bidding isn't always smart.** It works brilliantly in the right conditions and fails silently in the wrong ones.
**The Uncomfortable Truth:**
- Google profits whether Smart Bidding works for you or not
- Most accounts don't have enough data for ML to work properly
- "Learning" can be an expensive excuse for poor performance
**This Audit Answers:**
1. Is your current bid strategy appropriate for your data volume?
2. Is it actually outperforming simpler alternatives?
3. What would you need to change for it to work better?
---
## Required Context
### Must Have
**1. Current Bid Strategy**
- Strategy type (Target CPA, Target ROAS, Max Conversions, Max Conv Value, Manual CPC, Enhanced CPC)
- How long it's been running
- Target values (if applicable)
**2. Performance Data (Last 30 Days)**
- Conversions
- Conversion value (if using value-based bidding)
- Cost
- CPA or ROAS achieved
**3. Campaign-Level Data**
Per campaign using the strategy:
- Conversions
- Cost
- CPA/ROAS
### Strongly Recommended
**4. Historical Comparison**
Performance BEFORE current strategy vs. AFTER
- Or: Last 30 days vs. prior 30 days
**5. Conversion Volume Context**
- Monthly conversion count
- Conversion lag (days from click to conversion)
### Nice to Have
- Auction insights (impression share trends)
- Bid strategy status (Learning, Limited, Eligible)
- Any recent changes (budget, targets, structure)
- Seasonality factors
---
## Bid Strategy Requirements
### Minimum Data Thresholds
Google's official recommendations are minimums. Real-world requirements are higher.
| Strategy | Google Says | Reality Check | Why |
|----------|-------------|---------------|-----|
| Target CPA | 15 conv/month | 30-50 conv/month | Need statistical significance |
| Target ROAS | 15 conv/month | 50+ conv/month | Value variance needs more data |
| Max Conversions | 15 conv/month | 30+ conv/month | No ceiling without volume |
| Max Conv Value | 15 conv/month | 50+ conv/month | Value optimization is harder |
| Portfolio Strategies | Varies | 100+ conv/month across portfolio | Aggregation helps but has limits |
### Strategy-Specific Red Flags
**Target CPA:**
- Actual CPA consistently 30%+ above target
- Strategy in "Learning" for 2+ weeks
- Wild CPA swings day-to-day
- Impression share dropping while CPA rising
**Target ROAS:**
- Actual ROAS consistently 30%+ below target
- Conversion value highly variable
- Low conversion volume with high-value outliers
- Strategy "Limited by budget" while missing ROAS
**Max Conversions:**
- CPA increasing month-over-month
- Conversion quality declining
- Strategy spending budget but CPA rising
- No efficiency improvement over time
**Max Conversion Value:**
- Low-value conversions increasing
- High-value conversion rate dropping
- Average order value declining
- Strategy chasing easy wins vs. valuable ones
---
## Audit Framework
### Phase 1: Data Sufficiency Check
**Question:** Does this account have enough data for the chosen strategy?
**Assessment Matrix:**
| Monthly Conversions | Recommended Strategy |
|--------------------|---------------------|
| < 15 | Manual CPC or Enhanced CPC |
| 15-30 | Enhanced CPC or cautious Target CPA |
| 30-50 | Target CPA viable |
| 50-100 | Target CPA/ROAS comfortable |
| 100+ | Any strategy viable, portfolio recommended |
**Conversion Lag Factor:**
- Lag > 7 days: Add 2 weeks to learning period expectations
- Lag > 14 days: Consider if Smart Bidding can even work (delayed signal problem)
---
### Phase 2: Performance Reality Check
**Question:** Is the strategy actually delivering?
**Target CPA Assessment:**
| Actual vs Target | Assessment | Recommendation |
|------------------|------------|----------------|
| < 90% of target | Over-delivering | Consider lowering target to capture more volume |
| 90-110% of target | On target | Maintain |
| 110-130% of target | Underperforming | Check for issues, consider target adjustment |
| > 130% of target | Failing | Strategy change likely needed |
**Target ROAS Assessment:**
| Actual vs Target | Assessment | Recommendation |
|------------------|------------|----------------|
| > 110% of target | Over-delivering | Consider lowering target for more volume |
| 90-110% of target | On target | Maintain |
| 70-90% of target | Underperforming | Check for issues |
| < 70% of target | Failing | Strategy change likely needed |
**Trend Analysis:**
- Improving trend: Strategy may still be learning/optimizing
- Flat trend: Strategy has likely plateaued
- Declining trend: Strategy is degrading, intervention needed
---
### Phase 3: Alternative Comparison
**Question:** Would a different strategy perform better?
**Manual CPC Benchmark:**
- If you have historical manual CPC data, compare directly
- If not, estimate: What CPA would you achieve at avg CPC with current CVR?
**Strategy Comparison Framework:**
```
Current: Target CPA at $45 (target: $40)
Calculation: If CVR is 3% and avg CPC is $1.20
Manual CPC theoretical CPA = $1.20 / 0.03 = $40
Verdict: Manual CPC might match or beat current Smart Bidding
```
**Portfolio vs. Individual:**
- If running same strategy across multiple campaigns
- Could consolidation into portfolio improve performance?
- More data = better ML performance
---
### Phase 4: Diagnosis
**If Strategy is Underperforming, Why?**
**Data Issues:**
- Not enough conversions (most common)
- Conversion lag too long
- Conversion tracking issues
- Mixed conversion types (leads + sales in same campaign)
**Target Issues:**
- Target too aggressive (strategy can't hit it)
- Target too conservative (leaving money on table)
- Target doesn't reflect true business value
**Structure Issues:**
- Too many campaigns fragmenting data
- Audience overlap causing competition
- Budget limiting before strategy can optimize
**External Issues:**
- Recent major changes disrupted learning
- Seasonality not accounted for
- Competitive landscape shifted
---
## Output Format
### Executive Summary
```
CURRENT STRATEGY: [Strategy Name]
RUNNING SINCE: [Date/Duration]
VERDICT: 🟢 Working / 🟡 Marginal / 🔴 Not Working
[2-3 sentence summary of findings and primary recommendation]
```
---
### 📊 Data Sufficiency Assessment
| Requirement | Your Data | Threshold | Status |
|-------------|-----------|-----------|--------|
| Monthly Conversions | [X] | 30+ recommended | ✅/⚠️/❌ |
| Conversion Lag | [X] days | < 7 days ideal | ✅/⚠️/❌ |
| Time on Strategy | [X] weeks | 4+ weeks | ✅/⚠️/❌ |
| Strategy Status | [Status] | "Eligible" | ✅/⚠️/❌ |
**Data Verdict:** [Sufficient / Borderline / Insufficient] for [current strategy]
---
### 📈 Performance Analysis
**Target vs. Actual:**
| Metric | Target | Actual | Variance | Trend |
|--------|--------|--------|----------|-------|
| CPA/ROAS | $X / X.X | $X / X.X | +X% | ↑/↓/→ |
**Performance Verdict:** [On Target / Underperforming / Failing]
**Trend Analysis:**
- Week 1-2: [Performance]
- Week 3-4: [Performance]
- Direction: [Improving / Flat / Declining]
---
### 🔍 Diagnosis
**Primary Issue:** [What's causing underperformance]
**Contributing Factors:**
1. [Factor] - [Evidence]
2. [Factor] - [Evidence]
**What the Data Shows:**
[Specific observations that support the diagnosis]
---
### 💡 Recommendations
**Option A: [Primary Recommendation]**
- What: [Specific change]
- Why: [Rationale based on diagnosis]
- Expected Impact: [What should improve]
- Risk: [What could go wrong]
- Timeline: [When to evaluate]
**Option B: [Alternative]**
- What: [Specific change]
- Why: [When this makes more sense]
- Expected Impact: [What should improve]
**What NOT to Do:**
- [Common mistake to avoid given this situation]
---
### 🧪 If Switching Strategies
**Recommended Switch:** [From] → [To]
**Transition Plan:**
1. [Step 1 with timing]
2. [Step 2 with timing]
3. [Step 3 with timing]
**Success Metrics:**
- Primary: [Metric] should [improve/stabilize] within [timeframe]
- Secondary: [Metric] should [expectation]
**Rollback Criteria:**
- If [condition], revert to [previous strategy]
---
### ⚠️ Watch Items
- [Thing to monitor] - Check in [timeframe]
- [Thing to monitor] - Check in [timeframe]
---
## Common Scenarios & Verdicts
### Scenario 1: Low Volume + Target CPA
**Situation:** < 30 conversions/month on Target CPA
**Typical Finding:** CPA volatile, often above target
**Recommendation:** Switch to Enhanced CPC or Manual CPC
**Why:** Insufficient data for ML to learn patterns
### Scenario 2: Target ROAS Consistently Missing
**Situation:** ROAS 30%+ below target for 4+ weeks
**Typical Finding:** Strategy is "working" but not effectively
**Recommendation:** Either lower target or switch to Max Conv Value with caps
**Why:** Target may be unrealistic for current traffic quality
### Scenario 3: Max Conversions CPA Creep
**Situation:** CPA steadily increasing on Max Conversions
**Typical Finding:** No efficiency ceiling, strategy just spending
**Recommendation:** Switch to Target CPA with current CPA as starting target
**Why:** Need to give the algorithm a goal, not just "get more"
### Scenario 4: Strategy in "Learning" Forever
**Situation:** Status shows "Learning" for 2+ weeks
**Typical Finding:** Too many changes or insufficient data
**Recommendation:** Stabilize (stop changes for 2 weeks) or consolidate campaigns
**Why:** Every change restarts learning, perpetual learning = perpetual inefficiency
### Scenario 5: Great Performance, Low Volume
**Situation:** Hitting targets but impression share < 50%
**Typical Finding:** Strategy being too conservative
**Recommendation:** Gradually adjust targets to capture more volume
**Why:** Leaving money on the table by being too restrictive
---
## Limitations
**I can assess:**
- Data sufficiency for strategy type
- Performance vs. targets
- Trend direction
- Strategy fit diagnosis
**I cannot assess:**
- Conversion quality (need backend data)
- Competitive auction dynamics (need auction insights)
- Whether targets reflect true business value
- Incrementality of conversions
**For complete analysis, also provide:**
- Auction insights data
- Backend conversion quality metrics
- Business context for target setting
- Historical strategy performance comparisons
---
## Quality Checklist
Before delivering audit:
- [ ] Data sufficiency clearly assessed with thresholds
- [ ] Performance comparison uses appropriate benchmarks
- [ ] Diagnosis is evidence-based, not assumed
- [ ] Recommendations are specific and actionable
- [ ] Transition plan included if strategy switch recommended
- [ ] Risks and rollback criteria defined
- [ ] Timeline expectations set appropriately
- [ ] Avoided generic "it depends" conclusions
ReadyBid Strategy Auditor
MarkdownUTF-8Verified