How do I make a Street Interview Ad?

Fake vox-pop: an interviewer stops a stranger, the answer pivots into the product. The gold standard UGC opener right now.

model⚡ Veo 3
agency would charge$3,500
you'll spend~$6
runs onTikTok · Reels

THE REFERENCE — BILLY ON THE STREET — THE FORMAT'S PEAK

THE PROMPT — COPY, SWAP THE BRACKETS, RUN

Vertical 9:16 video, 8 seconds, handheld street-interview style, realistic lip sync and natural location audio. SUBJECT: male interviewer, late 20s, denim jacket over a hoodie, flag-topped handheld mic; the stranger is a woman, mid-20s, tote bag, pulling one AirPod out. LOCATION: busy city sidewalk — storefront glass behind them, pedestrians passing, a parked delivery bike. LIGHT: overcast daylight, soft and even. CAMERA: chest-high phone gimbal, slight walk-and-stop wobble, medium two-shot.

BEATS —
0.0–1.5s: Camera already rolling mid-approach; interviewer thrusts the mic in and asks fast, "What's one app you'd actually pay for?" The stranger flinches, then laughs — the flinch is the hook.
1.5–3.0s: Stranger, thinking out loud: "Honestly? [PRODUCT — e.g. "Ledgr, a $9/mo budgeting app"]." Said with a small shrug, zero salesmanship.
3.0–6.5s: She warms up: "[ONE-SENTENCE BENEFIT — e.g. "It found $140 of subscriptions I forgot I had"]" — said with rising disbelief, small laugh at the end.
6.5–8.0s: Interviewer turns to camera, deadpan: "Okay. That's the whole video." Quick whip-pan toward the street.

Street ambience throughout: traffic, distant chatter. No captions baked in, no brand logos or signage, no watermark, hands with five fingers, mic never covers either mouth.

HOW TO RUN IT

  1. Run in Veo 3 (audio + lip sync is the whole trick — Seedance needs a separate VO pass)
  2. Generate 3 takes, keep the most natural read
  3. Add captions + end-card CTA in CapCut
browse all 101 ad styles →