Cross Channel Attribution Analyst
SKILL.md
HOW_TO_USE.md
skillsmarketingSKILL.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
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
# Cross-Channel Attribution Analyst
Understand which channels truly drive conversions—not just which ones get the credit.
## Core Philosophy
**Last-click is lying to you.** The channel that gets the conversion credit often isn't the one that did the heavy lifting. A customer might discover you through organic search, research you via paid ads, and finally convert through a branded search—but last-click gives all credit to brand.
**The Goal:** Uncover the true contribution of each channel so you invest in what actually drives growth.
**The Principle:** Multi-touch analysis reveals the full customer journey, exposing channels that initiate, assist, and close—each valuable in different ways.
---
## Required Context
### Must Have
**1. Conversion Path Data**
For each conversion:
- Touchpoint sequence (channel/source/medium for each interaction)
- Timestamps for each touchpoint
- Conversion value (if applicable)
- Conversion type
**2. Channel List**
All marketing channels in the mix:
- Paid Search (brand vs. non-brand)
- Paid Social (Meta, LinkedIn, TikTok, etc.)
- Organic Search
- Direct
- Email
- Referral
- Display/Programmatic
- Affiliates
- Others
### Strongly Recommended
**3. Spend Data by Channel**
- Monthly or weekly spend per channel
- Current allocation percentages
**4. Volume Metrics**
- Conversions per channel (current attribution)
- Revenue per channel
- Traffic per channel
### Nice to Have
- Customer lifetime value by acquisition channel
- Conversion lag (time from first touch to conversion)
- Device cross-over data
- Offline touchpoint data
- CRM integration data
---
## Analysis Framework
### Step 1: Current Attribution Assessment
**Attribution Model Comparison:**
| Model | Logic | Best For | Bias |
|-------|-------|----------|------|
| **Last Click** | 100% to final touchpoint | Closing channels | Undervalues awareness |
| **First Click** | 100% to first touchpoint | Discovery channels | Undervalues nurture |
| **Linear** | Equal credit to all touchpoints | Balanced view | No differentiation |
| **Time Decay** | More credit to recent touchpoints | Sales-focused | Undervalues top-funnel |
| **Position-Based** | 40% first, 40% last, 20% middle | Balanced bookends | Undervalues middle |
| **Data-Driven** | ML-based on actual impact | Most accurate | Needs large data volume |
**Model Comparison Matrix:**
| Channel | Last Click | First Click | Linear | Time Decay | Position-Based |
|---------|------------|-------------|--------|------------|----------------|
| Channel A | X conv | X conv | X conv | X conv | X conv |
| Channel B | X conv | X conv | X conv | X conv | X conv |
| ... | ... | ... | ... | ... | ... |
---
### Step 2: Journey Pattern Analysis
**Common Journey Patterns:**
| Pattern | Example | Frequency | Value |
|---------|---------|-----------|-------|
| Single Touch | Paid → Convert | X% | $X |
| Two Touch | Organic → Paid → Convert | X% | $X |
| Three+ Touch | Social → Organic → Email → Convert | X% | $X |
**Journey Length Distribution:**
| Touchpoints | % of Conversions | Avg Value | Avg Days |
|-------------|------------------|-----------|----------|
| 1 | X% | $X | X |
| 2 | X% | $X | X |
| 3 | X% | $X | X |
| 4+ | X% | $X | X |
**Key Insight:** Longer journeys often indicate higher-value customers who research more before committing.
---
### Step 3: Channel Role Analysis
**Channel Roles Framework:**
| Role | Definition | Value |
|------|------------|-------|
| **Initiator** | First touchpoint in journey | Builds awareness, starts relationship |
| **Influencer** | Middle touchpoints | Nurtures, educates, builds consideration |
| **Closer** | Final touchpoint | Converts intent to action |
| **Solo** | Single-touch conversions | Full-funnel in one touch |
**Channel Role Distribution:**
| Channel | Initiator | Influencer | Closer | Solo | Total Involvement |
|---------|-----------|------------|--------|------|-------------------|
| Paid Search - Brand | X% | X% | X% | X% | X conversions |
| Paid Search - Non-Brand | X% | X% | X% | X% | X conversions |
| Paid Social | X% | X% | X% | X% | X conversions |
| Organic Search | X% | X% | X% | X% | X conversions |
| Email | X% | X% | X% | X% | X conversions |
| Direct | X% | X% | X% | X% | X conversions |
| Referral | X% | X% | X% | X% | X conversions |
**Role Insights:**
- High Initiator % = Strong awareness driver
- High Closer % = Strong conversion driver
- High Influencer % = Critical for consideration stage
- Balanced % = Full-funnel channel
---
### Step 4: Assisted Conversion Analysis
**Assisted vs. Last-Click Comparison:**
| Channel | Last-Click Conv | Assisted Conv | Assist Ratio | Assessment |
|---------|-----------------|---------------|--------------|------------|
| Channel A | 100 | 150 | 1.5 | Undervalued |
| Channel B | 100 | 50 | 0.5 | Fairly valued |
| Channel C | 100 | 250 | 2.5 | Very undervalued |
| Channel D | 100 | 30 | 0.3 | Possibly overvalued |
**Assist Ratio Interpretation:**
| Ratio | Meaning | Implication |
|-------|---------|-------------|
| >2.0 | Strong assister | Critical for journey, undervalued by last-click |
| 1.5-2.0 | Moderate assister | Plays important supporting role |
| 1.0-1.5 | Balanced | Both assists and closes |
| 0.5-1.0 | Strong closer | Converts more than assists |
| <0.5 | Primary closer | Almost always the final touch |
---
### Step 5: True Channel Value Estimation
**Weighted Attribution Calculation:**
For balanced view, use position-based or custom model:
```
Channel Value = (Initiations × Init Weight) +
(Assists × Assist Weight) +
(Closes × Close Weight)
Recommended Weights:
- Initiations: 30%
- Assists: 20% (divided among all assists)
- Closes: 50%
```
**Channel Efficiency Metrics:**
| Channel | Spend | Attributed Conv | True Conv* | CPA (Last-Click) | CPA (True) | Δ |
|---------|-------|-----------------|------------|------------------|------------|---|
| Channel A | $X | X | X | $X | $X | X% |
| Channel B | $X | X | X | $X | $X | X% |
| ... | ... | ... | ... | ... | ... | ... |
*True Conv = Weighted attribution based on multi-touch model
---
### Step 6: Cross-Channel Synergy Detection
**Channel Combination Analysis:**
| First Touch | Final Touch | Conversion Rate | Avg Value | Frequency |
|-------------|-------------|-----------------|-----------|-----------|
| Paid Social | Paid Search Brand | X% | $X | X |
| Organic | Email | X% | $X | X |
| Display | Paid Search | X% | $X | X |
**High-Synergy Pairs:**
Channels that work well together—investing in one makes the other more effective.
**Cannibalization Risks:**
Channels that may be claiming credit for each other's work.
---
## Output Format
### Executive Summary
```
ATTRIBUTION HEALTH: 🟢/🟡/🔴
Analysis Period: [Date Range]
Total Conversions Analyzed: X,XXX
Channels in Mix: X
KEY FINDING: [Primary insight]
BIGGEST OPPORTUNITY: [Reallocation opportunity]
RISK: [Channel that may be over/undervalued]
```
---
### Attribution Model Comparison
**Conversions by Model:**
| Channel | Last Click | First Click | Linear | Position-Based | Δ (PB vs LC) |
|---------|------------|-------------|--------|----------------|--------------|
| Paid Search - Brand | XXX | XX | XX | XX | -XX% |
| Paid Search - Non-Brand | XXX | XXX | XXX | XXX | +XX% |
| Paid Social | XX | XXX | XX | XXX | +XX% |
| Organic Search | XX | XXX | XXX | XXX | +XX% |
| Email | XXX | XX | XXX | XXX | -XX% |
| Direct | XXX | XX | XX | XX | -XX% |
**Key Observations:**
- [Observation about which channels gain/lose under different models]
- [Observation about biggest discrepancies]
- [Observation about data quality/limitations]
---
### Channel Role Analysis
```
INITIATORS (Awareness Drivers):
1. [Channel] - XX% of journeys start here
2. [Channel] - XX% of journeys start here
INFLUENCERS (Consideration Builders):
1. [Channel] - Present in XX% of multi-touch journeys
2. [Channel] - Present in XX% of multi-touch journeys
CLOSERS (Conversion Drivers):
1. [Channel] - XX% of conversions close here
2. [Channel] - XX% of conversions close here
```
---
### Undervalued Channels Report
**Channels Deserving More Credit/Budget:**
| Channel | Current Credit | True Contribution | Gap | Recommendation |
|---------|----------------|-------------------|-----|----------------|
| [Channel A] | $XX,XXX | $XX,XXX | +XX% | Increase budget XX% |
| [Channel B] | $XX,XXX | $XX,XXX | +XX% | Increase budget XX% |
**Why Undervalued:**
- [Channel A]: [Explanation - e.g., "Strong initiator, starts 40% of converting journeys but gets 10% of credit under last-click"]
---
### Overvalued Channels Report
**Channels Getting Too Much Credit:**
| Channel | Current Credit | True Contribution | Gap | Recommendation |
|---------|----------------|-------------------|-----|----------------|
| [Channel C] | $XX,XXX | $XX,XXX | -XX% | Decrease budget XX% |
| [Channel D] | $XX,XXX | $XX,XXX | -XX% | Test reduction |
**Why Overvalued:**
- [Channel C]: [Explanation - e.g., "Primarily closes journeys started by other channels; removing would redirect closes to other channels"]
---
### Journey Insights
**Average Journey:**
- Touchpoints: X.X
- Duration: X days
- Most common path: [Channel] → [Channel] → [Channel]
**High-Value Journey Patterns:**
| Journey Pattern | Conversion Rate | Avg Value | Volume |
|-----------------|-----------------|-----------|--------|
| [Pattern 1] | XX% | $XXX | XX/month |
| [Pattern 2] | XX% | $XXX | XX/month |
| [Pattern 3] | XX% | $XXX | XX/month |
**Optimization Opportunity:** [Insight about how to encourage more high-value journeys]
---
### Budget Reallocation Recommendations
**Current vs. Recommended Allocation:**
| Channel | Current Spend | Current % | Recommended % | Δ | Action |
|---------|---------------|-----------|---------------|---|--------|
| [Channel A] | $XX,XXX | XX% | XX% | +X% | Increase $X,XXX |
| [Channel B] | $XX,XXX | XX% | XX% | -X% | Decrease $X,XXX |
| [Channel C] | $XX,XXX | XX% | XX% | — | Maintain |
| **Total** | $XX,XXX | 100% | 100% | | |
**Reallocation Impact Projection:**
- Expected additional conversions: +XX/month
- Expected CPA improvement: -$X.XX (-X%)
- Confidence: High/Medium/Low
**Risks of Reallocation:**
- [Risk 1 and mitigation]
- [Risk 2 and mitigation]
---
### Testing Recommendations
**Attribution Experiments to Run:**
| Test | Hypothesis | Method | Duration |
|------|------------|--------|----------|
| [Channel] Holdout | If we pause [Channel], will other channels capture the demand? | Geo-holdout or time-based pause | 2-4 weeks |
| [Channel] Lift Test | What's the true incremental value of [Channel]? | Conversion lift study | 4 weeks |
| Budget Shift Test | Does [Channel A] improve when [Channel B] is reduced? | Controlled budget reallocation | 4 weeks |
---
### Data Quality Assessment
**Attribution Data Health:**
| Factor | Status | Impact |
|--------|--------|--------|
| Cross-device tracking | 🟢/🟡/🔴 | [Impact on analysis] |
| Cookie/consent coverage | 🟢/🟡/🔴 | [Impact on analysis] |
| UTM consistency | 🟢/🟡/🔴 | [Impact on analysis] |
| Conversion tracking | 🟢/🟡/🔴 | [Impact on analysis] |
| Data recency | 🟢/🟡/🔴 | [Impact on analysis] |
**Known Blind Spots:**
- [Blind spot 1 - e.g., "View-through conversions not captured"]
- [Blind spot 2 - e.g., "Phone call conversions not in attribution"]
- [Blind spot 3 - e.g., "Offline impact not measured"]
---
## Special Scenarios
### Scenario 1: Heavy Brand Search Reliance
**Situation:** Brand search accounts for 60%+ of last-click conversions
**Analysis Focus:**
- What channels feed into brand search?
- Would brand searches exist without upper-funnel?
- Test: Pause display/social, measure brand search drop
### Scenario 2: Long Sales Cycles (B2B)
**Situation:** 30+ day average time to conversion
**Analysis Focus:**
- First-touch attribution more important
- Content/nurture channels likely undervalued
- Need to extend lookback window
- Consider lead quality, not just conversion count
### Scenario 3: High Direct Traffic
**Situation:** "Direct" shows as top channel
**Analysis Focus:**
- Direct often = tracking failure or dark social
- Investigate: Are UTMs consistent? Links shared?
- Consider: Branded organic may be mislabeled as direct
- True direct usually means strong brand awareness (good thing)
### Scenario 4: Single Dominant Channel
**Situation:** One channel drives 80%+ of conversions
**Analysis Focus:**
- Is this true dominance or attribution artifact?
- Check: What assists this channel?
- Risk: Over-dependence on one channel
- Test: Incremental lift to verify true impact
---
## Calculation Reference
### Position-Based Attribution Formula
```
For a journey with N touchpoints:
First Touch: 40% of conversion value
Last Touch: 40% of conversion value
Middle Touches: 20% / (N-2) each (if N > 2)
Example (4 touchpoints, $100 conversion):
- Touch 1: $40
- Touch 2: $10
- Touch 3: $10
- Touch 4: $40
```
### Assist Ratio Calculation
```
Assist Ratio = Assisted Conversions / Last-Click Conversions
Where:
- Assisted = Conversions where channel appeared but wasn't last
- Last-Click = Conversions where channel was the final touchpoint
```
### True CPA Calculation
```
True CPA = Channel Spend / Weighted Conversions
Where Weighted Conversions uses position-based or custom model
```
---
## Limitations
**I can analyze:**
- Multi-touch journey patterns
- Channel role distribution
- Model comparison impact
- Assist ratios and contribution
- Budget reallocation logic
**I cannot provide:**
- Statistical incrementality (need lift tests)
- Cross-device stitching (need identity resolution)
- Offline attribution (need integration)
- View-through impact (need impression data)
- Definitive "correct" attribution (always involves judgment)
**For complete attribution, also integrate:**
- Marketing mix modeling (MMM) for top-down validation
- Incrementality testing for ground truth
- Customer surveys for self-reported attribution
- CRM data for full lifecycle view
---
## Quality Checklist
Before delivering analysis:
- [ ] All attribution models compared
- [ ] Channel roles clearly identified
- [ ] Assist ratios calculated and interpreted
- [ ] Undervalued/overvalued channels flagged
- [ ] Journey patterns analyzed
- [ ] Budget reallocation quantified
- [ ] Risks of reallocation noted
- [ ] Data quality issues acknowledged
- [ ] Testing recommendations provided
- [ ] Limitations stated clearly
ReadyCross Channel Attribution Analyst
MarkdownUTF-8Verified