TikTok · Trust & Safety

Account verification flow

UX copy for TikTok's verification badge application, a flow where every screen has to earn enough trust to ask for a government ID.

Role

Content Designer

Content Strategist

Scope

End-to-end flow

Error messages

Edge cases

Surface

TikTok mobile app

Audience

Reputable creators (celebrities, artists, athletes, influencers)

Credible businesses and institutions

Phase 1: Getting them in

The entry screen leads with what verification does for the user. No mention of IDs or review times; that friction comes later, once they're invested.

Goals

Reduce confusion and friction

Increase verified accounts

Basic requirements checklist showing five account requirements

Back

The flow

Prevention

One ✓ and five ✗ examples: visual do/don't beats prose rules. The most common failure ("no screenshots or photocopies") is bolded on its own line.

Confirmation

"Remove your ID?" states the consequence, not just the question. "Keep" comes first; the destructive action never gets the default.
Impact

10M+ in annual savings by reducing ID verification requests

Higher completion rates

Each verification type is defined by who it's for, with concrete examples. The biggest misfiling risk, political figures, is handled inside the option text itself.

"Incorrect selections may delay or affect your application."

The costliest mistake in the flow, warned before the choice, not after the failure

The application

Phase 2: Make the requirements impossible to misread

Requirements read as personal progress ("You meet {x} of the {total} requirements"), not a wall of rules. The identity form says the one thing that matters: enter details exactly as they appear on your ID.

Recovery

Errors name the problem and the fix together: "Unsupported ID type. Upload a valid, supported document." Privacy reassurance sits directly under the upload, right where the anxiety happens.

Confirm your identity form with legal name, date of birth, and ID upload fields

The follow-through

Beyond the happy path

Errors, warnings, and second thoughts

What this flow taught me

Notability screen with content category and trusted public sources requirements

✓ "Independent, non-sponsored media coverage"

✗ "Blogs, social posts, or videos"

Notability, the top rejection reason, turned into a concrete accept/reject list

Phase 3: Eliminate the “now what?”

Submission sets the expectation (channel and timeframe: email, within 30 days) so nobody contacts support tomorrow.

Approval resists over-celebrating: one exclamation point, then the useful part, where the badge now lives.

Takeaway

Trust-sensitive flows are written backwards: start from every way the application can fail, then write the happy path so those failures never happen. The best line here isn't on the approval screen. It's the warning on screen two.

Impact: this work contributed to $10M+ in annual savings by reducing reliance on third-party ID verification requests. Clearer copy meant more users completing verification in-flow.

Next project