Analyzing paid search impact on organic traffic
29.6% of paid clicks would have come organically
30% of paid clicks would likely have occurred organically without ads. This means 70% of your ad spend is driving genuinely new traffic.
Industry average for brand campaigns is 50-80% cannibalization. Your 30% is well below average — your brand ads are more incremental than typical.
Current strategy appears effective. Maintain brand defense while exploring growth opportunities elsewhere.
When you run Google Ads, you pay for clicks. But what if people would have clicked your free organic listing anyway? That's cannibalization — paying for traffic you'd get for free.
🍕 Pizza Shop Analogy: Imagine you own a pizza shop on Main Street. You're already the only pizza place in town (great organic ranking). Now you pay for a billboard right next to your shop. People see your billboard and come in — but they were already walking to your shop anyway. You paid for customers you would have gotten for free.
We use a conservative, evidence-based approach that combines two data sources:
1. Empirical Evidence (70% weight): We compare organic clicks during two periods:
If organic clicks increased when ads were off, that's evidence of cannibalization. If organic stayed the same or decreased when ads were off, there's no evidence that ads were stealing organic clicks.
2. Position-Based Priors (30% weight): Research from Google and Bing incrementality studies suggests expected cannibalization rates based on organic ranking:
🔬 Scientific Method: Think of position-based rates as a "hypothesis" and organic traffic changes as "experimental results." If the experiment contradicts the hypothesis (organic didn't increase when ads turned off), we trust the data over the theory.
If you rank #1 organically for a search term, most people will click your organic result — the ad is largely redundant. But if you rank #15? An ad might be your only chance to be seen on page 1.
However, position alone isn't enough. We need to see organic traffic actually recover when ads are off to confirm cannibalization is happening.
Branded terms include your company name. You almost always rank #1 organically — so ads here often have higher cannibalization.
Non-branded terms are generic searches where competition is higher and organic rankings vary — ads here are typically more incremental.
🏠 Home Address Analogy: Branded search is like someone typing your home address into GPS — they're coming to YOU specifically. Non-branded is like searching "coffee shop near me" — you're competing with everyone, and an ad helps you stand out.
Not all estimates are equally reliable. We flag segments with low confidence when:
Small samples can produce misleading averages. When you see a "Low Confidence" warning, treat that segment's cannibalization rate with caution.
This is the estimated amount you paid for clicks that would have come through organic search anyway. It's not truly "wasted" in all cases — ads provide other benefits like:
Think of "wasted spend" as opportunity cost — money that could be redirected to acquire genuinely new customers.
Just because organic clicks changed doesn't mean it's meaningful — it could be random noise. We run statistical tests to determine if the difference is real:
Welch's t-test: Compares average daily clicks between periods. The p-value tells you the probability the difference happened by chance. If p < 0.05, we're 95% confident the difference is real (not noise).
Mann-Whitney U test: A backup test that doesn't assume normal distribution. Useful when data has outliers or is skewed.
Cohen's d (effect size): Measures how big the difference is:
This analysis has inherent limitations you should be aware of:
| Position | Queries | Avg Rank (Ads Off) | Avg Rank (Ads On) | Paid Clicks | Cannibalization | Spend | Wasted |
|---|---|---|---|---|---|---|---|
| 1-3 (Top) | 110 | 1.7 | 2.1 | 1,938 | 39% | $1,539 | $455 |
| 4-5 | 63 | 3.9 | 4.5 | 1,498 | 40% | $907 | $399 |
| 6-10 | 61 | 5.7 | 7.4 | 1,237 | 38% | $3,087 | $437 |
| 11-20 | 5 | 14.9 | 14.7 | 6 | 46% | $19 | $12 |
Are the differences between ads-on and ads-off periods statistically meaningful, or just noise?
Daily data not available — only aggregate comparison possible. Without daily data, we cannot calculate p-values or confirm statistical significance. For proper statistical tests, re-export GSC data with the Date dimension included.
Average daily organic clicks from Google Search Console during each period.
Organic clicks were 9.3% higher when ads were off. This suggests ads may be capturing clicks that would have gone to organic listings.
All ad data is from a brand campaign. Every search term that triggered an ad is considered a branded query. Brand campaigns typically have higher cannibalization because users searching for your brand name would likely find you organically anyway.
When someone searches for your brand name, they're already looking for you — not your competitors. You likely rank #1 organically for these terms, so the ad is often redundant. However, brand ads still provide value: they defend against competitor ads, control messaging, and dominate SERP real estate. The "wasted spend" is better thought of as defensive/branding spend rather than pure waste.