Does AI traffic really convert better than organic traffic? Decoding a study that challenges conventional wisdom.

Does AI traffic really convert better than Google? This burning question has divided the SEO world for months. Between studies that claim… conversion rates 4 times higher and those that temper the enthusiasm, it’s difficult to find your way around.

I’m going to break it down for you. Amsive’s most recent study who analyzed 54 websites over 6 months. Spoiler alert: the results will surprise you and probably Change your approach to artificial intelligence traffic.

The Amsive Study That Calls Everything into Question

Amsive Digital has just published a study that has caused a sensation in the SEO community. For six months, they have analyzed conversion behavior across 54 different websites.

The numbers speak for themselves:

Traffic Source	Conversion Rate	Statistical Significance
LLM Traffic	4.87%	p = 0.794
Organic Traffic	4.60%

Do you see this p = 0.794? In statistics, this means that there is no no significant difference between the two traffic sources. The 0.27% difference could very well be due to chance.

“Contrary to sensationalist claims, our data does not show AI traffic superiority in terms of conversion,” explains the Amsive team.

This conclusion challenges all strategies based on the assumption that AI converts better than Google.

The War of Numbers: Why Do Studies Contradict Each Other?

But wait… Other studies claim the opposite! How can we explain such discrepancies between the research ?

• Writesonic: Claims that ChatGPT converts 2.08 times better than Google
• Seer Interactive: Announcement of conversions 9 times higher for AI traffic
• Amsive: Does not find no significant difference

The answer can be summed up in three words: methodological biases.

The Pitfalls of Attribution

Let me tell you the story of Marc, owner of an online computer hardware store in Lyon. He was convinced that his ChatGPT traffic converted at 15% compared to only 3% for Google.

Problem: its tracking was completely messed up.

Visitors arrived via ChatGPT, but Google Analytics classified them as “direct traffic” due to a missing referrer. The result? All of its best conversions were attributed to the wrong channel.

Once the tracking was corrected, the reality was quite different: 4.2% for AI, 4.1% for organicVirtually identical.

The Variables That Disrupt Everything

The conflicting studies can be explained by several factors:

1. Sample size: 54 sites (Amsive) vs 12 sites (Writesonic)
2. Business sector: B2B vs B2C vs e-commerce
3. Definition of conversions: Sales vs. leads vs. engagement
4. Analysis period: 6 months vs 3 months vs one-off

The method completely changes the results.

What French Data Really Reveals

In France, the situation is even more nuanced. A study by Webmarketing-com reveals that 51% of French users prefer AI for sensitive issues such as health or finances.

But this preference doesn’t automatically translate into better conversions. Here’s why:

Postman	Impact on Conversion	Explanation
Intent Quality	Neutral	Same level of purchase intention
User Behavior	Similar	Time spent and pages viewed are equivalent.
Trust Level	Variable	Depends on the sector and age

French users are cautious with AI. They use it to get information, but convert at the same rate only via Google.

Volume Remains the Achilles’ Heel of AI Traffic

Even if AI traffic converted better (which, according to Amsive, is not the case), there is a size problem major:

AI traffic represents less than 1% of total traffic, compared to 31.9% for organic traffic.

Let’s do the calculations for a site receiving 100,000 monthly visitors:

• Google traffic: 31,900 visitors × 4.60% = 1,467 conversions
• AI Traffic: 800 visitors × 4.87% = 39 conversions

Even with a “slightly” higher rate, AI generates 37 times fewer conversions than Google in absolute volume.

Impressive Growth… But

Certainly, AI traffic has exploded +970% this yearBut going from 0.02% to 0.8% is mathematically impressive without being economically revolutionary.

To match Google, growth of 4000% moreWe’re not there yet.

How to Effectively Measure and Track AI Traffic

Enough about contradictory figures. Let’s move on to concrete How to properly track AI traffic on your site?

Google Analytics 4 setup

Here is my step-by-step method to avoid attribution errors:

1. Create custom channels for each AI source
2. Define the reference points specific (chatgpt.com, perplexity.ai, etc.)
3. Configure the UTM parameters when possible
4. Segmenting audiences to analyze behavior

Tracker’s Referents

AI platform	Reference to Configure	Expected Volume
ChatGPT	chatgpt.com	60% of AI traffic
Perplexity	perplexity.ai	25% of AI traffic
Gemini	gemini.google.com	10% of AI traffic
Others	Variables	5% of AI traffic

Attention : 30% of AI traffic is still classified as “direct” due to missing referrers. Monitor for unusual spikes in direct traffic.

Optimization Strategies for Both Channels

Now that we know that AI doesn’t convert any better than GoogleWhat strategy should be adopted for 2025?

The Intelligent Hybrid Approach

My recommendation? Don’t put all your eggs in one basket with AI, but don’t ignore it either.

• 80% of your efforts they remain focused on traditional SEO (volume + proven conversions)
• 20% experimentation on LLM optimization (future potential)

Optimize for Google AND AI

Good news: techniques that work for Google often work for AI:

Technical	Google SEO	AI/LLM SEO
Quality content	✅ Essential	✅ Essential
Clear structure	✅ Important	✅ Crucial
Direct answers	✅ Featured snippets	✅ AI Quotes
Domain Authority	✅ Ranking factor	🔶 In development

Concrete Actions for 2025

Here is your action plan based on real data:

1. Maintain your SEO excellence (31.9% of traffic, proven conversions)
2. Optimize your direct responses (beneficial for Google AND AI)
3. Track correctly AI traffic to measure its true performance
4. Experiment in moderation LLM optimization in a few pages
5. Monitor the developments volumes and behaviors

FAQ: Your Questions About AI Traffic vs. Google

In the end, AI traffic does not convert better than organic traffic. according to the most reliable data we have today. And you know what? Maybe it’s better this way.

This prevents us from going off in all directions chasing an Eldorado that doesn’t (yet?) exist. Let’s focus on what already works, and keep a watchful eye on future developments.

After all, in our line of work, the revolutions that are announced are often evolutions… and that’s just fine!