Style 09 of 10
Street interview
Also called “man on the street”, “vox pop” or “blind taste test”.
Someone with a microphone stops a passer-by and asks one question: which one tastes better, what do you think this costs, what's the one thing you'd tell a new business owner. The stranger hesitates, answers, and the reveal lands. It's one of the most native-looking formats in the feed, and one of the easiest to fake badly.
- Consumer brands
- Lenders & financial services
- Cold audiences
Opens OpenCraft Studio with this style’s brief ready to edit. Fill in your product; nothing is made until you approve the storyboard.
What it is
A street interview puts the product in front of strangers who don't know it's coming. The format borrows its credibility from the stranger: they weren't briefed, so their reaction reads as real. The shape is question, hesitation, answer, reveal.
The best ones are games, not endorsements. A blind taste test ("which one would you pick?"), a price guess, or a quiz with a factual answer ("this one has zero sugar") gives the ad a payoff the viewer wants to see, and the reveal is a fact rather than a stranger's opinion.
That distinction matters when the strangers aren't real. Real passers-by giving honest first reactions are fine. But the FTC's 2024 rule bans testimonials from people who don't exist or never used the product, and a cast or generated "passer-by" raving about the taste, presented as a real stranger, is that kind of testimonial. A quiz whose answer is a fact, with any actors labelled, stays on the right side of it.
It's also the youngest style here. Every long-running street interview we found started in July or August 2026, and none had passed 63 days when we checked, so treat it as promising rather than proven. The ones working now are all consumer products you can try in five seconds.
Anatomy
- 0–2sThe questionThe host's question, on a mic in frame and as big text: "Which one's the zero-sugar one?" Outdoor daylight, city or beach texture.
- 2–6sThe tryThe stranger tastes, holds or looks. One close-up of the product in their hand.
- 6–10sThe hesitationA real beat of thinking before they answer. It's the moment that makes the format believable; don't cut it.
- MiddleThe revealThe answer and the twist: they picked yours, the cheap one was the expensive one, the sweet one has no sugar. The brand is shown or said here.
- Last 3sCloseThe host hands over the product, or a quick end card with where to buy.
- Length
- 15–30 seconds. One stranger, or two or three answering the same question in quick succession.
- Shot length
- Handheld, 1.5–3 seconds. Reaction close-ups punctuate.
- Talking head
- High but split between host and stranger, both on camera.
- Captions
- Big word-by-word captions, and the question as a title for the first seconds. Brand names pop up as graphics on the reveal.
- Sound
- Street sound under the voices, with a light bed. Practitioners warn against slick music; keep it low.
- End card
- Short, with where to buy (a store, a site).
What it’s good for, and bad for
Good for
- Taste and price. Drinks, snacks and anything where a blind test or a price guess has a surprising answer.
- Beating an objection. "Zero sugar can't taste good" is an objection; a stranger picking it blind answers it.
- Cold audiences. It looks like content, not an ad, and the question hooks people who've never heard of you.
Bad for
- Serious or private topics. Nobody wants to be stopped on the street about an injury, a debt or a medical treatment.
- Products that need explaining. If the product can't be tried in five seconds, the stranger has nothing to react to.
- Invented reactions. Generated or scripted strangers praising the product are fake endorsements. Make it a quiz with a factual answer, and label actors.
By business
- Consumer brands
- The natural home: food and drink taste tests, price guesses for fashion and home goods, "smell this" for personal care.
- Lenders & financial services
- Ask business owners, not shoppers: "What would you do with $50,000 tomorrow?" or "What's the one thing you wish you'd known when you started?" Keep it a question, not an endorsement.
- Local & professional services
- Weak. Home services and law firms have nothing for a passer-by to try. A local business can ask neighbours a trivia question about the area, but it's a stretch.
Real examples
Real brands’ ads, cited for reference. We didn’t make them. Unless marked “ran”, each was still running in the Meta Ad Library on Sep 29, 2026. Days live is the only performance signal the Library shows, and an ad that keeps running is usually paying for itself.
- Better ThanZero-sugar soda
A host with a mic stops two young men outside a beachfront hotel for a blind taste test: their strawberry lemonade against another brand's, with graphics of both cans popping up. The guys sip from two cups, argue it out, and pick one before the reveal. Big white word-by-word captions carry the whole exchange. The post text names the result and where to buy it: "Find us at Sprouts."
Take from it: A blind test gives the ad a payoff the viewer waits for, and the stranger's back-and-forth is the hook.
View in the Meta Ad Library · ID 1539494144586380 - PelviniWatches
Guess the price: on a busy city sidewalk a host in a wide-brimmed hat hands a young man in a Yankees cap a watch box and asks what he thinks it costs, and his reactions play out as he opens it. Then it cuts to a short product film of the green-dial watch on its box and on a wrist, ending on "Buy 1 get 1 free". The post text: "You don't need a $5,000 watch to look powerful."
Take from it: A price guess works for any product whose value is hidden, and the gap between the guess and the price is the pitch.
View in the Meta Ad Library · ID 1667909584324475 - Saje WellnessPeppermint roll-on
On a park path at golden hour, a host with a handheld mic stops young women one at a time ("Asking people if they know Saje Peppermint Halo") and has them roll it onto their temples. Close-ups catch the cooling moment, eyes closing, then she asks for one word and to be honest. The host signs off to camera: "Shop online now."
Take from it: If the product has a felt moment (a cooling tingle, a first sip), the stranger's face does the selling.
View in the Meta Ad Library · ID 1365062095116555 - Fantom WalletCard-fanning wallet
Two borrowed proofs in one ad. The first 15 seconds is a wallet reviewer's desk demo, framed as a clip. Then a label switches to "Then, they tried the Fantom Wallet" and it cuts to strangers on the street working out the lever and fanning their cards ("Whoa."). The post text says it straight: a reviewer ranked it top tier, then they handed it to strangers.
Take from it: An expert's verdict followed by strangers' first try covers both kinds of doubt.
View in the Meta Ad Library · ID 2562182907628729
Common mistakes
- A question with no tension. "Do you like soda?" has no answer worth waiting for. Ask something with a surprising reveal.
- Cutting the hesitation. The pause before the answer is what makes it feel unstaged. Trimming every breath makes it look scripted.
- Too polished. A lit, stabilized, graded street interview looks like a commercial. Handheld, daylight, a visible mic.
- Staged raves. Strangers (or actors, or generated people) gushing about the product are fake testimonials. A quiz with a factual reveal is honest and more interesting.
- Burying the brand. If the reveal doesn't show the product clearly, the viewer remembers the stranger and not what they drank.
The prompt
The brief we’d give OpenCraft to make one. Swap the brackets for your own details. The director turns it into a script and a storyboard for you to approve before anything is filmed.
Make a 20-second street-interview ad for [product] from [brand]. A host with a handheld mic stops [who, e.g. two friends outside a beach café] and asks one question with a surprising answer: [e.g. a blind taste test, a price guess]. Show the try, keep a real beat of hesitation, then the reveal: [the fact, e.g. the one they picked has zero sugar]. Make it a quiz with a factual answer, not a stranger praising the product. Handheld, outdoor daylight, big word-by-word captions, street sound with a light music bed under it. End with the host handing over the product and [where to buy]. The people on screen are actors, so label them.