Social Proof Audit

Social Proof & Trust Audit

> Judge whether a site's proof actually builds trust — and call out the fake or > manipulative tactics.

What it does

Inventories every trust and social-proof element on a site, evaluates each for credibility, specificity, and placement relative to the decision point, and flags manipulative patterns (fake scarcity, unverifiable claims, dark patterns). Returns a trust scorecard and prioritized fixes — or, for a competitor, a read on how they manufacture trust.

When to use this skill

• "Audit the social proof / trust signals on this page."
• "Are these testimonials credible? Why won't people trust this?"
• "Review the trust and urgency tactics on ."
• Pre-launch credibility check on a sales or landing page.

Inputs needed

• Required: the page or site URL (or screenshots).
• Helpful: the conversion goal and the audience's main fears/objections — proof

is only effective when it answers a real anxiety.

• If reviewing your own page, ask what objection most often stops buyers.

Method

1. Inventory trust elements:

• Testimonials (named? photo? role/company? specific result?)
• Case studies (with real numbers and a clear before/after?)
• Third-party reviews/ratings (G2, Trustpilot, app store, Google)
• Customer/partner logos and counts ("10,000+ users")
• Guarantees / refund terms / risk reversal
• Authority signals (press, awards, certifications, expert endorsements)
• Security/compliance badges (SSL, payment, GDPR, SOC2)
• Urgency/scarcity (timers, limited seats, cart close)

1. Grade each on credibility & specificity — named + specific + verifiable

beats anonymous + vague. Specific numbers and real identities score highest.

1. Check placement — does proof appear at the decision points (near CTAs,

pricing, forms) and match the objection at that moment?

1. Match proof to objections — list the buyer's likely fears and whether a

proof element answers each. Find unaddressed objections.

1. Flag manipulation — fake/perma countdown timers, "results not typical"

income claims, invented review counts, fake activity popups, pressure tactics. Note these honestly; recommend ethical alternatives.

1. Score & recommend — a trust score plus prioritized fixes (add, strengthen,

reposition, or remove).

Output format

# Social Proof & Trust Audit — 

## TL;DR
- Trust score __/5. The biggest credibility gap, and any manipulative tactics found.

## Trust inventory
| Element | Present? | Credibility (specific/named/verifiable) | Placement |
|---------|----------|------------------------------------------|-----------|
| Testimonials | | | |
| Case studies | | | |
| Reviews / ratings | | | |
| Logos / counts | | | |
| Guarantee | | | |
| Authority signals | | | |
| Security badges | | | |
| Urgency / scarcity | | | |

## Proof ↔ objection match
| Likely objection | Addressed by | Gap? |
|------------------|--------------|------|

## Manipulative / risky tactics
- (List, with an ethical alternative for each.)

## Prioritized fixes
| # | Fix | Element | Why | Impact | Effort |
|---|-----|---------|-----|--------|--------|

Quality bar

• Rate proof by specificity and verifiability, not quantity.
• Call out fake/manipulative trust tactics plainly; never recommend them.
• Don't assume a claim is true or false you can't verify — mark it "unverifiable".
• Tie missing proof to the specific objection it would resolve.

Common mistakes to avoid

• Counting testimonials instead of judging their credibility.
• Ignoring placement — strong proof in the footer does little at the CTA.
• Praising urgency tactics that are obviously fake.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: fjilvvi
• Source: fjilvvi/growthlens
• License: MIT

Install and usage instructions live in the source repository linked above.