The Strategist Center

Attribution & Cross-Channel Correlation

The Attribution Problem

Google Ads uses Last-Click attribution by default. This means:

If someone sees your YouTube ad on Monday, searches your brand on Wednesday, and converts via a Shopping ad on Friday—the Shopping campaign gets 100% credit. The Search campaign gets nothing. The YouTube ad is invisible.

This causes chronic underinvestment in top-of-funnel channels.

Your CFO sees: "YouTube campaign: $5K spend, $2K revenue, 0.4x ROAS. Kill it."

But the reality is: YouTube drove the brand search that drove the Shopping conversion. Without YouTube, that $15K Shopping revenue doesn't exist.

The Strategist Center makes this invisible value visible.

1. Time-Lagged Correlation Analysis

How it works:

We pull daily spend from your Display/Video campaigns and daily conversions from your Search campaigns. Then we calculate Pearson's r between these two time series with various lags (0, 1, 2... up to 14 days).

Interpreting the results:

+0.5 to +1.0 (strong positive): When Display spend goes up, Search conversions follow

Near 0: No relationship between channels

-0.3 to -1.0 (negative): Campaigns may be cannibalizing each other

The Optimal Lag:

The system highlights the time delay where correlation peaks.

If Display → Search correlation peaks at lag=5 days, it means Display ads take approximately 5 days to drive Search activity. This is your awareness-to-intent pipeline.

Example output:

Lag 0 days: r = 0.12 (no same-day effect)

Lag 3 days: r = 0.38 (moderate)

Lag 5 days: r = 0.67 (strong) ← Optimal lag

Lag 7 days: r = 0.41 (fading)

Lag 14 days: r = 0.09 (no effect)

Translation: Your Display ads create intent that materializes as Search conversions approximately 5 days later.

Data requirements: Minimum 30 days, ideally 90 days. At least $1K/day total spend for statistical reliability.

2. Ghost Value (The Assisted Revenue Problem)

What it is:

A Display campaign shows $5K direct revenue in Google Ads. But it assisted $80K in downstream conversions. That $80K is ghost value—real revenue that wouldn't exist without the Display campaign, but isn't credited to it under Last-Click.

How we surface it:

The Strategist Center analyzes:

View-Through Conversions: Users who saw your Display/Video ad, didn't click, but later converted via another channel

Cross-Campaign Paths: GA4 conversion paths showing which campaigns appeared before the final click

Ghost Value Calculation:

For each campaign, we calculate:

Ghost Value = Assisted Conversions × Average Order Value

The Ghost Value Card:

For each upper-funnel campaign:

Direct Revenue (Last-Click): $5K

Ghost Value (Assisted): $80K

Total True Value: $85K

True ROAS: $85K / $10K spend = 8.5x

Last-Click ROAS: $5K / $10K = 0.5x

The insight: Your "failing" Display campaign is actually your most valuable asset—it's just invisible under Last-Click.

Use this to justify awareness budgets to your CFO. Export the Ghost Value card as evidence.

3. Halo Effect Detection

The question: Does non-brand advertising drive branded search volume?

How we detect it:

The system tracks:

Daily non-brand campaign spend (Display, Video, Non-Brand Search)

Daily branded search impressions (Brand Search campaigns)

When non-brand spend increases and branded search volume follows (with a lag), that's the Halo Effect.

Visual: Dual-axis line chart showing Non-Brand Spend (left axis) and Branded Search Volume (right axis) over 90 days.

What to look for:

Pattern 1: Clear Halo

Non-brand spend increases 30% → Branded search volume increases 15-20% within 7 days

Translation: Your awareness campaigns are working. People see your ads, remember your name, search for you later.

Pattern 2: No Halo

Non-brand spend increases 50% → Branded search volume stays flat

Translation: Your awareness campaigns aren't creating brand recall. The creative might be generic, or the targeting is too broad.

Pattern 3: Inverse Halo

Non-brand spend decreases → Branded search volume decreases proportionally

Translation: Your brand is dependent on continuous paid awareness. If you cut Display/Video, brand search will decline. Budget accordingly.

Important note: This analysis uses correlation, not causation. External factors (press coverage, word of mouth, seasonality) can also drive branded search. Use this as directional evidence, not proof.

4. Multi-Touch Path Analysis

What it shows:

The most common conversion paths from GA4, showing every touchpoint before the final conversion.

Example paths:

Path 1 (23% of conversions):

Display Ad → Organic Search → Brand Search → Purchase

Path 2 (18% of conversions):

YouTube Ad → Direct Visit → Shopping Ad → Purchase

Path 3 (12% of conversions):

Non-Brand Search → Brand Search → Purchase

Why this matters:

Under Last-Click, Path 1 gives 100% credit to Brand Search. Display gets nothing.

But Display initiated the journey. Without it, the user never would have searched your brand.

The Touchpoint Frequency Table:

TouchpointFirst Touch %Middle Touch %Last Touch %
Display34%12%5%
YouTube22%8%3%
Non-Brand Search28%35%22%
Brand Search8%30%45%
Shopping5%15%25%

The insight: Display and YouTube are journey starters (high First Touch %). Brand Search and Shopping are journey closers (high Last Touch %). Both are essential.

Recommended credit model: Consider switching from Last-Click to Position-Based (40% first touch, 40% last touch, 20% middle) for more balanced attribution.

When to Use This Dashboard (vs. Other Tools)

Use Strategist Center when you want to:

Justify awareness/upper-funnel budget to stakeholders

Understand cross-channel relationships

Detect the Halo Effect of non-brand campaigns

See the full customer journey, not just Last-Click

Don't use Strategist Center when you want to:

Optimize individual campaign performance (use PCC or Creative Lab)

Reallocate budgets within Search (use Budget Balancer)

Find wasted search terms (use Search Hygiene)

Who Should Use This:

✅ Strategists and CMOs (cross-channel budget decisions)

✅ Analysts (attribution modeling)

✅ Agency leads (justifying channel mix to clients)

❌ Daily practitioners (too strategic for daily optimization)

How Often: Monthly. Quarterly for formal reporting to stakeholders.

Technical FAQ

Q: How much historical data do I need for correlation analysis?

Minimum 30 days, ideally 90 days. We also need at least $1K/day total spend for statistical power—low-budget accounts will see noisy results.

Q: What if I see negative correlation between my campaigns?

Negative correlation means campaigns might be cannibalizing each other. For example, if non-brand and brand Search both target similar queries, increasing non-brand spend might decrease brand conversions by capturing the same users earlier. This isn't necessarily bad—just means they're substitutes, not complements.

Q: Can I use this to justify brand awareness budget to my CFO?

Absolutely. Export the correlation heatmap showing how Display spend predicts Search revenue at a specific lag. Combine it with the Ghost Value calculation showing total assisted revenue. Data-backed story: 'Every $1 in Display generates $0.20 direct ROAS plus $1.50 in assisted downstream revenue within 7 days.'

Q: Does this replace Google's Data-Driven Attribution model?

No, it complements it. Google's DDA is a black box—you can't see the logic. Our correlation analysis is transparent and gives you the 'why' behind cross-campaign relationships. Use both: DDA for in-platform optimization, Strategist Center for strategic budget allocation.

Q: What's the difference between Ghost Value and Assisted Conversions in Google Ads?

Google Ads shows Assisted Conversions as a count. Ghost Value converts that into a dollar amount (Assisted Conversions × AOV) so you can compare it directly to campaign spend and calculate True ROAS. It's the same data, presented in a more actionable format.

Q: How reliable is the Halo Effect detection?

It's directional evidence, not definitive proof. External factors (press, word of mouth, seasonality) can also drive branded search. Use the Halo Effect chart alongside other signals. If correlation is strong AND consistent over 90+ days, confidence is high.

Q: Can I share the attribution analysis with clients?

Yes. Click 'Export Report' to generate a PDF with the correlation matrix, Ghost Value cards, Halo Effect chart, and top conversion paths. This makes an excellent appendix for monthly client reports.

Q: What if my business only runs Search campaigns? Is this tool useful?

Less so. The Strategist Center is most powerful when you run multi-channel campaigns (Search + Display + Video). If you only run Search, the correlation analysis won't have Driver/Harvester pairs to compare. Focus on the Path Analysis tab, which still shows useful cross-campaign journey data within Search.

The Strategist Center | ClickCatalyst