Core Concept · Trust Layer

Machine Trust Signals.

Not what looks trustworthy. What AI can actually verify. These are the specific crawlable signals AI checks before it will recommend a store.

Category Core ConceptChecks Crawlable evidenceRelated AI Trust Layer
Opening store...
SSL Certificate
Organization
Contact
Return Policy
Shipping
Reviews Schema
AggregateRating
FAQ Schema
Machine Trust Signals
73%
AI doesn't trust what looks trustworthy. It trusts what it can verify.
Definition
Machine Trust Signals are the structured, crawlable signals AI systems verify before including a store in recommendation flows. They are distinct from visual trust cues that humans see. Machine Trust Signals include AggregateRating schema, return policy text in HTML, SSL certificate presence, legal identity in the footer, shipping information in crawlable text, and contact details. They must be machine-readable to count.
Two different kinds of proof

Human trust vs machine trust

A human forms trust from design and copy. AI cannot read either one. It only counts what it can parse.

Human · ★★★★★
Trust Badge?
Beautiful Reviews?
Guarantee?
Secure Checkout?
AI: ? ? ? ?
Machine
AggregateRating
Organization
Policy HTML
Shipping HTML
SSL
AI: VERIFIED
Looks trustworthy ≠ Machine Trust.
Nobody else builds this

AI forensic table

Picture a lab bench. AI lays every signal out and scans them one at a time, the same way a forensic investigator processes evidence.

SSLPending
OrganizationPending
ReviewsPending
PoliciesPending
ShippingPending
ContactPending
Verified
This is what AI actually sees

X-ray mode

Everything a human notices, from photography to layout, disappears. Only the machine-readable signals remain.

Price
Review Schema
Organization
Contact
Policy
Shipping
Diagnostics, not a checklist

Signal detector

Every signal reads at a different strength, the same way a diagnostic panel reads a car.

Organization
Detected
AggregateRating
Weak
Policy
Strong
Shipping
Strong
Illustrative example, not a specific measured result.
The exact difference

Machine vs visual

Same policy. Same words. Only the format decides whether AI can use it.

As an image
"30 Day Returns"
AI: ✗ Image
As HTML
<p>30 Day Returns</p>
AI: ✓ HTML
Every part has a place

Trust signal blueprint

Seen from above, each signal is a component. All of them feed the same assembly.

OrganizationPolicyReviewsShippingContactSSLMachine Trust
What the scan actually logs

AI validation log

This is not a summary. It is the log itself, one check at a time.

Reading merchant data...
✓ Checking Organization...
✓ Checking Reviews...
✓ Checking Contact...
✓ Checking SSL...
✗ Checking Policies...
✓ Checking Shipping...
↓ Building Trust... 82%
One signal moves the whole number

Missing signal

Four out of five signals confirmed. The one that is missing is enough to change the outcome.

OrganizationConfirmed
PolicyConfirmed
ShippingConfirmed
ContactConfirmed
ReviewsMissing
Recommendation Confidence
91%
62%
Illustrative example, not a specific measured result.
Every signal, on file

Signal library

Click a card to see what AI is actually looking for.

Organization
Identity
Legal business name and entity in machine-readable Organization schema.
AggregateRating
Reputation
Review counts and scores exposed as structured data, not just a widget.
Contact
Verification
Real address, phone, and email in crawlable text, not an image.
Policies
Commerce Safety
Return and refund terms written as readable HTML, not a graphic.
Shipping
Operational Evidence
Shipping terms and timelines present in crawlable page text.
AI verdict
Merchant
Verified ✓
Eligible for Recommendation

See which Machine Trust Signals AI can actually verify.

Find out exactly which crawlable signals your store already has, and which ones are still invisible to AI.

Check My Store's Evidence →