Loamly vs Plausible: Which Analytics Tool Actually Detects AI Traffic?

I get asked this a lot: "Why did you build Loamly when Plausible exists?"

Fair question. I use Plausible myself. It's excellent.

But Plausible and Loamly solve different problems. Let me explain why I built Loamly even though privacy-first analytics tools already exist.

The Short Answer

Plausible is a privacy-first general analytics tool that recently added AI traffic detection as a feature.

Loamly was built specifically to solve the AI traffic blind spot, using a different technical approach.

If you need full website analytics with some AI visibility, use Plausible.

If AI traffic detection accuracy is critical to your business, or you need brand visibility monitoring (what AI says about you, not just clicks), consider Loamly.

That's the honest summary. Now let me get into the details.

What Plausible Does Well

Plausible is genuinely excellent at what it does:

• Privacy-first by design: No cookies, no personal data, GDPR compliant without banners
• Lightweight script: 75x smaller than Google Analytics (~1KB vs ~45KB)
• Simple dashboard: One page with essential metrics, no complexity bloat
• Open source: You can self-host or use their cloud
• Real-time updates: Data refreshes every 30 seconds
• 16,000+ paying customers: Proven at scale

I respect what Marko and Uku built. Plausible proved there's a market for GA alternatives.

The AI Traffic Challenge

Here's the problem both tools face: AI platforms don't always tell you where their traffic came from.

When someone clicks a link from ChatGPT, Perplexity, or Claude, what happens?

Platform	Referrer Behavior	Detectable in Analytics?
Perplexity	Passes referrer header	Yes - shows as referral
ChatGPT	Sometimes strips referrer	Partial - depends on context
Claude	Inconsistent	Partial - varies by link type
Gemini	Often strips referrer	Limited

This inconsistency is the core problem. Traditional analytics tools (including Plausible) rely on the HTTP referrer header to identify traffic sources. When AI platforms strip that header, the traffic appears as "Direct" instead of being attributed to the AI source.

How Plausible Handles AI Traffic

Plausible added AI traffic detection in 2024. Their approach:

1. Referrer pattern matching: Filter by known AI referrer strings (perplexity.ai, chatgpt.com, etc.)
2. User agent analysis: Identify AI crawler signatures
3. Custom channel groups: Create an "AI" channel to segment this traffic

This works for platforms that pass referrer data. Perplexity, Microsoft Copilot, and sometimes ChatGPT will show up correctly.

The limitation: when AI platforms strip the referrer (which ChatGPT and Claude often do), Plausible can't see it. The traffic just appears as Direct. There's no signal to detect.

According to Plausible's own analysis, they saw 2,200% growth in trackable AI traffic in 2024. But that's only the traffic they can detect through referrers.

How Loamly Handles AI Traffic

I built Loamly to specifically solve this detection gap. Our approach is different:

1. Cryptographic Signature Verification (RFC 9421)

We implemented the HTTP Message Signatures standard (RFC 9421) that OpenAI, Anthropic, and Google are adopting for their AI agents.

When an AI platform sends traffic with a cryptographic signature, we can verify the origin with certainty - even without a referrer header.

This is more accurate than referrer-based detection because:

• Signatures can't be spoofed
• They work regardless of referrer stripping
• They provide cryptographic proof of origin

The tradeoff: not all AI platforms use signatures yet. It's an emerging standard.

2. Brand Visibility Monitoring

Here's what no analytics tool can show you: AI recommendations that never generate clicks.

When ChatGPT answers "What's the best analytics tool for startups?" and lists five options, those five brands got visibility. But only the one the user clicked on gets any traffic attribution.

Loamly monitors what AI actually says about your brand by running real queries against ChatGPT, Claude, Perplexity, and Gemini. We track:

• How often you're mentioned
• What topics trigger your mentions
• How you're positioned (leader, alternative, also-ran?)
• What competitors appear alongside you

This is separate from traffic analytics. It's about understanding your presence in AI responses.

Honest Feature Comparison

Feature	Plausible	Loamly
General website analytics	Full dashboard	Basic (use alongside GA4/Plausible)
Privacy-first	Yes	Yes
Self-hostable	Yes	Yes (open source)
AI referrer detection	Yes (referrer-based)	Yes (signatures + referrer)
Detection of stripped referrers	No	Yes (when signatures exist)
Brand visibility in AI responses	No	Yes
Revenue attribution	Yes	Via integrations
Funnels	Yes (Business plan)	No
eCommerce tracking	Yes	No
Price	From $9/mo	Free tier, then $89/mo

When to Use Each Tool

Use Plausible if:

• You need comprehensive website analytics
• AI traffic is a nice-to-have, not critical
• You want the simplest possible dashboard
• Budget is tight (starts at $9/mo)
• You're replacing Google Analytics entirely

Use Loamly if:

• AI traffic accuracy is business-critical
• You need to know what AI says about your brand (not just clicks)
• You're in a market where AI recommendations drive decisions
• You're building a GEO (Generative Engine Optimization) strategy
• You want to track AI visibility over time

Use Both if:

This is what I actually recommend for most teams.

Plausible for general analytics (traffic, conversions, pageviews). Loamly for AI-specific intelligence (visibility, brand monitoring, accurate AI traffic attribution).

They solve different problems.

The Data on AI Traffic Quality

Here's why AI traffic detection matters:

Research from Microsoft Clarity's study of 1,200 publisher websites found that AI-referred traffic converts at significantly higher rates:

• AI traffic → 1.66% signup rate
• Search traffic → 0.15% signup rate
• Direct traffic → 0.13% signup rate

Ahrefs' June 2025 analysis shows AI traffic represents 0.5% of their traffic volume but generates 12.1% of signups—a 23x conversion rate advantage.

If you can't measure it, you can't optimize for it. And you're making budget decisions on incomplete data.

What I Got Wrong Building Loamly

Let me be honest about mistakes I made:

Mistake 1: Initially I tried to build a full analytics replacement. Wrong approach. The market has good analytics tools. What was missing was AI-specific intelligence. I pivoted to focus there.

Mistake 2: I underestimated how inconsistent AI platforms are. ChatGPT's referrer behavior changed multiple times during my development. You have to build for chaos.

Mistake 3: I assumed everyone cared about AI traffic. They don't - yet. But the teams that get it now will have data advantages later.

The Future of AI Traffic Detection

Both Plausible and Loamly will improve AI detection over time. Some predictions:

1. AI platforms will standardize: RFC 9421 adoption is growing. OpenAI, Anthropic, and Google are all moving toward signed requests for their agents.

2. Referrer stripping will decrease: Google already fixed an "accidental" AI Mode bug that was stripping referrers. Pressure is mounting.

3. Brand visibility becomes a metric: Knowing what AI says about you will be as important as knowing your search rankings.

4. Integration is the answer: No single tool will do everything. The best stacks will combine general analytics + AI intelligence.

Try the Free Check

Not sure if AI traffic matters for your site? Run a free visibility check at loamly.ai/check.

It takes 3 minutes and shows you:

• Your current AI visibility score
• Which AI platforms mention your brand
• Where you appear vs competitors
• Specific GEO improvement recommendations

No signup required for the basic check.

---

Comparison data current as of December 2025. Plausible feature list from their official pricing page. AI traffic statistics from Microsoft Clarity and Ahrefs research studies.