Street interview marketing: 10 playbooks for D2C brands

Short answer

Street interview marketing uses vox pop, unscripted answers from people on the street, as brand content and ads. Kreators.Army's ten playbooks make it a system for D2C brands: build familiar trust, write for forwarding, keep the pauses, own one question, lead with doubt, move fast, log real words, find hidden beliefs, design for gossip and price each usable moment.

  • Street interviews borrow the trust people give to someone they know. Ask about the viewer's real habits, in their words, until the brand feels familiar.
  • Before posting, name the one person someone would forward the clip to. Then write the caption as that message.
  • Keep at least two honest pauses or stumbles in every cut. Dead air goes. Thinking stays.
  • Lead with doubt. A sceptic saying "ye sab marketing hai" on camera opens a video more honestly than a fan does.
  • Log each answer's exact words, emotion, speaker type and question. The log becomes your copy and your next questions.
  • Judge shoots by your own cost per usable moment. Nobody can honestly promise you a return in advance.
In this guide17 sections
CircleStore: Trust BlindlyWatch it in the video library

Your ad account wants new creative every few weeks. Your last influencer post looked like every other one in the feed. And somewhere in your comments, a customer has written one plain sentence that sells your product better than anything your team scripted.

Street interviews go out and find more sentences like that. A host takes one question to a footpath, people answer in their own words, and the best answers become posts and ads.

Most guides we've read sell the format without teaching it, and few are written for India. This one has ten playbooks you can run in-house or with a studio, each with a method, a street example and a tool. Where a number matters, it's yours to fill in.

Why are D2C brands turning to street interviews?

People trust people more than they trust ads, and a street interview puts an ordinary person's words where an ad would usually sit. It also shows a brand how its buyers really talk.

In Nielsen's 2021 global survey, 88% of people said they trust recommendations from people they know more than any other source1. An ad can't be someone you know.

A stranger on a footpath, answering honestly, is about as close as advertising gets.

One street shoot gives a brand four things:

  • Faces that look like your buyers, where a polished ad would usually sit.
  • Openings you can test, because every shoot gives you several first lines.
  • Your buyers' own language for captions, product pages and briefs.
  • A reason to post often without writing a new script each time.

The format also fits where Indian buyers already spend their time. About 588 million people in India watched short-form video in 2025, and slightly more of them lived in rural India than in cities2. Street interviews are made for those feeds: vertical, quick and carried by a face and a voice.

Paid reach gets more expensive whenever more brands bid for the same audience. So it helps to own content that earns attention before you pay to spread it.

The research on why ordinary voices land is less obvious than "people love real content". In a 2022 experiment with TV news items, viewers rated vox pops as less trustworthy than an expert quoting a poll. Yet the vox pops moved their personal opinions more3.

Our reading: viewers don't trust the format by default. They're moved when they recognise themselves in it, and the ten playbooks below earn that recognition on purpose.

What is a street interview ad, and what isn't one?

A street interview ad is a paid or boosted post made from an unscripted street interview. The person on screen isn't an actor, isn't reading lines and wasn't told what to say.

Vox pop video
A vox pop video is a short street interview. A host takes one question to the street, real people answer on camera without a script, and the best reactions are edited into a video for Reels, Shorts or other short-form feeds. Brands use vox pop to show how real people talk about a product or idea.

Our guide to what vox pop content is covers the format from scratch. Street interview marketing is the wider practice: using those answers in organic posts, in ads, on product pages and in your own research.

Kreators.Army is a Bengaluru studio that makes vox pop street-interview videos for D2C brands. We plan the question, the setup and who to approach. We don't script what anyone says.

The quickest comparison is who's talking, and what the viewer assumes about them.

Street interview ads and three other formats, as we see them
FormatWho speaksWhat viewers tend to assumeBest used for
Street interview adA stranger, unscripted"That could be me"Habits, objections and comparisons
User-generated content (UGC) adA creator or customer, often briefed"Probably paid, maybe honest"Demos and routines
Influencer postA creator with a following"Paid, but I trust them"Reaching that following
Scripted adActors or the brand"This is an ad"Clear offers and launches

None of these wins every time, and our vox pop vs UGC comparison covers when each one fits. What a street interview ad must never be is a fake one:

  • Staged and passed off as real. Actors playing strangers, or answers fed in advance, make an ad that lies about what it is.
  • A testimonial in disguise. A happy customer reading out praise isn't a street interview, and the praise isn't proof the product works.
  • A pitch holding a microphone. If the host spends the clip selling, the viewer can tell within seconds.

Here are the ten playbooks, one card each. Swipe through, then read the ones that fit your next shoot.

Slides

The ten playbooks at a glance

  1. 1. The KIRANA principle

    Be the shop on the corner. Ask about real habits, in the viewer's words.

  2. 2. The FORWARD test

    Name who would forward this clip, and to whom. Caption it as that message.

  3. 3. The PEAK-VALLEY map

    Mark highs and pauses before you edit. Keep at least two honest pauses.

  4. 4. The FIRST QUESTION advantage

    Own one unasked category question for a whole quarter.

  5. 5. The SKEPTIC flip

    Open with the viewer's own doubt. Keep the sceptic's words, whatever they say.

  6. 6. The VELOCITY gap

    A street cycle has fewer steps. Spend the speed on testing questions.

  7. 7. The OWNED intelligence asset

    Log exact words, emotion, speaker type and question. It becomes your copy.

  8. 8. The INVISIBLE category frame

    Ask one step away from the product to find the belief that decides the purchase.

  9. 9. The GOSSIP architecture

    Build for sharing: a real tension and a question with no clean answer.

  10. 10. TRUST arbitrage

    Price formats by your own cost per usable moment.

Playbook 1: the KIRANA principle for familiar trust

Ask about the viewer's real life, in their words, again and again, until the brand feels as familiar as the shop on the corner. That's the KIRANA principle: trust built the way a neighbourhood shop builds it.

Think of the kirana (neighbourhood grocery) store near your home. The owner doesn't run ads. They know your name, which atta your family buys and that someone will come in for curd on Sunday morning. Nobody calls that marketing, yet it's the reason you keep going back.

Most D2C marketing aims the other way. It talks wide, about "premium quality" or "a healthy lifestyle", and sounds like every rival in the category. A street interview can do what the shop owner does, at a scale no shop could: show the viewer someone who talks like their cousin, about a moment from their own day.

The difference starts with the question.

A wide question or a narrow one

Wide

  • "What do you think about skincare?"
  • Gets a polite opinion
  • Sounds like a survey

Narrow

  • "What does your skincare routine actually look like at 11 pm?"
  • Gets a scene from someone's night
  • Viewers see their own evening

Ask about a habit and nobody has to judge a product. People relax, and the details come out.

The narrow version invites detail: the face wash half rinsed off, the phone in one hand, the step everyone admits to skipping. Details like those make a viewer stop and think "that's me".

Framework

KIRANA

Six letters for familiar trust.

  1. K

    Know the person

    One buyer and their evening, before any audience.

  2. I

    Inhabit their world

    Ask about their day. The product can wait.

  3. R

    Reflect their words

    Their phrases in captions, Hinglish included.

  4. A

    Anchor to the familiar

    Aim for "that's me".

  5. N

    Narrow the target

    Specific enough to feel personal.

  6. A

    Accumulate

    Stack small moments of recognition across every post.

The last A is the one brands skip. One video is a moment of recognition. A run of them, all in your buyers' own words, is how a brand starts to feel known.

Think of it this way: one shoot gives you answers, and five give you patterns.

When this doesn't work: products people buy to feel different from everyone around them. Aspiration sells on distance, so lead with taste and comparison questions instead.

Playbook 2: the FORWARD test before you post

Before you publish, ask whether someone would send this clip to one particular person. If nobody comes to mind, the clip isn't ready yet, however good it looks. That one check is the FORWARD test.

A paid ad reaches whoever you target. A forwarded clip reaches whoever your viewer trusts, with a personal note attached. In India that note often travels on WhatsApp, which has more than 500 million users in the country4.

Platforms reward the same behaviour. Instagram's head, Adam Mosseri, has said the three signals that matter most for ranking are watch time, likes and sends5. A street answer people want to send to a friend fits that well.

In the edit, look for the moment that makes you think of a particular friend. That's the forward moment. Then ask what sharing it says about the sharer. "I found something funny" is close. "I know something you don't" passes. Cut whatever delays that moment: the long setup, the brand intro, the explanation.

Before and after

One caption, rewritten as a message

To a crowd

"Check out our new video on skincare habits."

  • The brand is talking
  • Nobody comes to mind

To one person

"Asked people about their 11 pm skincare routine. Sending this to whoever takes forty minutes in the bathroom."

  • A person is talking
  • You already know who gets it

Same clip. Only the caption changed.

Checklist

The FORWARD test, on your draft caption

Keep your draft clip and caption open, and tick each check it passes. The seven checks spell FORWARD. Your ticks stay on this device.

It passes. Post it, then watch the sends as closely as the views.

Where it breaks: products people buy privately. Few people forward a clip about something they'd rather not be seen buying. Let those clips work in the feed, and lean on the next two playbooks.

Playbook 3: the PEAK-VALLEY map for the edit

Mark the high points and the pauses in your footage before you edit, then keep at least two pauses in the final cut. That's the PEAK-VALLEY map. Peaks are the laugh, the strong opinion and the surprise. Valleys are the pause, the confusion, the half-finished sentence.

Most brand edits keep only the peaks. Every pause goes, then every stumble and every look away. What's left is tight and watchable, and it looks like every testimonial on the internet.

We think the valleys are where belief comes from. A smile can be coached. A three-second pause after a hard question is much harder to fake, and it shows the viewer the question landed. "Actually, I've never thought about it that way" means more after a silence than straight after the question.

This sits beside a rule we already edit by: no unnecessary pauses. The word doing the work is "unnecessary". Dead air while someone hunts for their keys goes. A pause in which someone changes their mind stays.

Framework

PEAK-VALLEY

Ten letters for an edit that keeps the thinking.

  1. P

    Plot the arc

    Mark peaks and valleys before the first cut.

  2. E

    Evaluate every valley

    Watch it twice before cutting.

  3. A

    Anchor on one valley

    Build around the pause that reveals most.

  4. K

    Keep two

    Two valleys survive the final cut.

  5. V

    Value the silence

    The beat after a question counts too.

  6. A

    Allow confusion

    A puzzled face shows the question was real.

  7. L

    Leave the stumble

    A restarted sentence is the most human frame.

  8. L

    Layer for contrast

    A valley right before your strongest peak.

  9. E

    Edit the turn

    Open on the shift from valley to peak.

  10. Y

    Your edge

    It's in the footage others throw away.

That turn from valley to peak often makes the strongest opening. YouTube's Shorts analytics measure how often viewers stayed to watch past the first seconds6, so the first beat has to earn the next ones. Our guide to street interview hooks has more openings built that way.

Try the map on a short transcript. We invented it, so it isn't from a real shoot. Decide how you'd mark each line, then tap to compare.

ScriptIllustrative transcript: "Would you buy a second-hand phone?"

Invented for this guide. One host, one participant, a busy footpath.

  1. HostWould you ever buy a second-hand phone?
  2. ParticipantSecond-hand? No. Never.
  3. A pause. They glance down at the phone in their hand.
  4. ParticipantActually... this one's second-hand. It was my cousin's.
  5. HostSo why "never"?
  6. ParticipantI don't know. I think I just don't like saying it out loud.
  7. ParticipantIt's been fine for a year.

Tap to reveal

Mark the peaks and valleys

  1. Line 2: "Second-hand? No. Never."

    The line

    A quick, confident answer.

    Our mark and the edit

    Peak, but a thin one. Keep it as the setup.

  2. Line 3: the pause and the glance

    The line

    Nothing is said.

    Our mark and the edit

    Valley, and the anchor. The answer changes before a word is spoken. Keep all of it.

  3. Line 4: "Actually... this one's second-hand."

    The line

    The admission.

    Our mark and the edit

    Peak. Straight after the pause, it's the turn. Open on lines 3 and 4.

  4. Line 6: "I just don't like saying it out loud."

    The line

    A half-explanation, a little awkward.

    Our mark and the edit

    Valley. Keep it. It names the real belief: embarrassment, with no doubt about the phone itself.

  5. Line 7: "It's been fine for a year."

    The line

    A calm close.

    Our mark and the edit

    Peak, a quiet one, and a clean ending. Drop line 5 if the cut runs long.

So the cut opens on the pause, lands the admission, keeps the awkward reason and ends on the calm line. Two valleys and two peaks, and nobody was told what to say.

When this doesn't work: rapid-fire formats, where pace is the point. Save one valley for the very end instead.

Playbook 4: the FIRST QUESTION advantage in your category

Pick one question nobody in your category has asked on camera, and keep asking it for a quarter. We think the first brand to ask a category's question in public gets linked to that conversation. That's the FIRST QUESTION advantage.

Ownership here isn't legal. Anyone can ask your question after you. But the first brand to ask sets the frame, and whoever comes next looks like they're replying. A domain name works the same way: the first to register it keeps it, and everyone after needs a variation.

A strong first question usually has four traits:

  • It's about the buyer's life, and your product isn't in it.
  • People already think it, but nobody has said it on camera.
  • It touches identity: money, family, status or a habit people defend.
  • It still gets new answers when you ask it again next month.

Here's the shape across a few categories, as starting points to adapt:

  • A kitchen-appliance brand: "What's the appliance you bought and never used?"
  • An innerwear brand: "What's the most uncomfortable thing you wear every day?"
  • A snack brand: "What do you eat when nobody's watching?"
  • A second-hand electronics store: "What would you never buy second-hand, and why?"

Each one invites a small confession: the air fryer under a dust cover, the 4 pm namkeen nobody admits to. Answers like that are hard to fake and easy to recognise, so the question keeps working all quarter.

Framework

FIRST QUESTION

Thirteen letters, read two at a time.

  1. F, I

    Find the gap, identify the belief

    An unasked question, and the belief behind it.

  2. R, S

    Run it first, seed it everywhere

    Shoot before rivals, then post on every platform.

  3. T, Q

    Track what spreads, question as an address

    Shared answers become your story.

  4. U, E

    Unclaimed ground, each shoot a flag

    Every post marks a little more territory.

  5. S, T

    Surprise once, take the lead

    Ask first, and rivals end up replying.

  6. I, O

    Identity first, one a quarter

    Questions about who people are, one at a time.

  7. N

    Nail distribution

    Post it until people link it to you.

Ask the same question of three different groups, so one question gives you three angles. One way to plan it: office crowds at lunch, students after college and families at a weekend market. Decide who you're asking before you decide where.

Then seed each version where it fits. The sharpest answer becomes a Reel or a Short, a longer cut goes to YouTube, and the most surprising clip goes out as something people will forward. Our vox pop strategy for Indian brands shows how one question sits inside a full content plan.

Where it breaks: a question so broad that any brand could ask it. "What does success mean to you?" belongs to nobody. Tie it to the moment your product lives in.

Playbook 5: the SKEPTIC flip, or leading with doubt

Open with the doubt your viewer already has. A sceptic says it on camera, you keep the pause after it, and then they react to something they can check for themselves. That's the SKEPTIC flip.

So the doubt is already in the room. Many viewers scrolling past your ad are thinking some version of "ye sab marketing hai": it's all marketing. Most brands cut that line when someone says it on camera.

The SKEPTIC flip opens with it.

We think hearing your own doubt said out loud by somebody else lowers your guard a little. Nobody expects an ad to show the objection first.

ScriptIllustrative SKEPTIC flip: the blind pick

Invented to show the shape. Nobody makes a product claim, and the participant checks the product for themselves.

  1. HostWhen an ad says "India's favourite" anything, what goes through your head?
  2. ParticipantYe sab marketing hai. Everyone says that.
  3. Hold the pause. The host does not argue.
  4. HostFair. Want to try these two blind and pick one to take home?
  5. Two unlabelled samples. The participant tries both.
  6. ParticipantThis one. Okay, that's not what I expected.
  7. HostStill all marketing?
  8. ParticipantSome of it. Not this bit.

Notice what the script avoids. The participant never says the product works, heals or protects anything. India's advertising code expects objective claims to be substantiated9, and a stranger's reaction isn't substantiation. Keep street questions about habits, tastes and feelings, and let viewers draw their own conclusions.

Framework

SKEPTIC FLIP

Eleven letters for leading with doubt.

  1. S

    Surface the doubt

    Say it before the viewer thinks it.

  2. K

    Keep the sceptic

    The doubting voice stays in the cut.

  3. E

    Embrace the doubter

    They're the main character.

  4. P

    Put proof in reach

    A label, a price or a blind test.

  5. T

    Track the shift

    Mark where their doubt moves, if it does.

  6. I

    Introduce late

    The product comes in after the doubt.

  7. C

    Close in their words

    Whatever they ended up saying.

  8. F

    Film before and after

    Doubt first, reaction second.

  9. L

    Let them write copy

    Their phrasing becomes your caption.

  10. I

    India's own doubt

    Open on lines like "ye sab marketing hai".

  11. P

    Pause after

    The silence makes the doubt real.

Our library has two CircleStore street videos that take both sides of one doubt: buying second-hand. Watch them as a pair.

CircleStore: Bad Second HandWatch it in the video library
CircleStore, "Bad Second Hand": the doubt side of buying second-hand.
CircleStore: Second Hand IntelligentWatch it in the video library
CircleStore, "Second Hand Intelligent": the other side. Is buying second-hand the intelligent choice?

Running both sides is the point. A 2018 study of Flemish TV news, drawing on about 2,000 vox pop interviews, found that most vox pop items showed only one point of view10. Show the doubt and the defence, and let the viewer pick a side.

When this doesn't work: categories where the doubt is about safety. A street reaction can't settle whether something is safe to use, and shouldn't try. Answer that with substantiated facts elsewhere.

Playbook 6: the VELOCITY gap between street and studio

A street shoot needs no studio booking or script approval, so it can move faster than a scripted production. The VELOCITY gap is that difference in speed, and the best use of it is testing more questions.

A scripted campaign usually runs through a brief, an agency concept, a script, approvals, a shoot, an edit and more approvals. A street cycle has four steps: a question, the street, the edit and the post. Fewer steps means you can answer a moment while people are still talking about it.

Picture the first heavy monsoon week in Mumbai. Trains run late, office shoes are soaked and every group chat is full of it. A footwear brand could be on the footpath that week asking "What did the rain ruin for you?" A scripted ad about the same moment would arrive after the rain had stopped.

How fast one cycle runs depends on the brief, the spot and the weather. Our own batches usually take 14 to 21 days, with the exact dates in the pilot agreement. The point is relative: whatever your scripted cycle takes, a street cycle has fewer places to wait.

Step by step

One fast street cycle

A possible shape for one cycle, run in-house or with a studio. The order matters more than the days.

  1. Pick the question

    Read this week's comments and messages for the question buyers keep circling.

  2. Write the follow-up

    Decide the second question before you leave. Its answer is usually the one you post.

  3. Shoot at two spots

    Two places with different crowds give you two kinds of answer.

  4. Cut two lengths

    A tight cut for Reels and Shorts, and a longer one for the whole conversation.

  5. Test the opening

    Instagram's Trial Reels show a cut to non-followers first11, a cheap way to compare two openings.

  6. Log it and repeat

    Add the answers to your log (Playbook 7) and let them pick next week's question.

Framework

VELOCITY GAP

Eleven letters for spending the speed well.

  1. V

    Volume

    Several pieces from each cycle.

  2. E

    Experiment

    Three questions where you'd test one concept.

  3. L

    Learn

    A shorter loop from asking to knowing.

  4. O

    Optimise

    See what holds attention before a launch.

  5. C

    Catch the moment

    Answer a topic while it's live.

  6. I

    Inform the next shoot

    Write down what each cycle taught you.

  7. T

    Timing

    Join your category's conversation early.

  8. Y

    Your next question

    This week's answers choose it.

  9. G

    Gather signals

    Comments and messages are next week's brief.

  10. A

    Act early

    Post the honest version this week.

  11. P

    Publish regularly

    More posts, more chances to learn.

Where it breaks: launches with claims that need legal or regulatory sign-off. Speed can't skip an approval that exists for a reason.

Playbook 7: the OWNED intelligence asset, a log of real words

Keep a running log of what people say on your shoots: exact words, emotion, speaker type and the question behind each answer. Over months, that OWNED intelligence asset becomes research a rival can't easily copy.

Every brand has a marketing budget. Very few keep a record of how their buyers talk. A street shoot gives you that record, yet it's common to keep the clips and throw the language away.

What you learn from hundreds of real answers is hard for a competitor to copy. They didn't ask your questions, in your places, framed your way. It's the same loop we work in: hypothesis, interaction, production, edit, learn, improve. The log is the learn step, written down.

It needs no special software. A shared spreadsheet works. After every shoot, someone watches the raw footage and logs each useful answer in four columns. They hold the exact words, the emotion, an archetype (a short label for the kind of person, never a name) and the question that brought it out.

An intelligence log, with made-up example rows
Exact wordsEmotionArchetypeQuestion
"My mother would say it's a waste."Guilty, half jokingFirst-job earner, lives with familyLast thing you bought just for yourself?
"I check the price on three apps first."ProudStudent on a budgetHow do you know a deal is real?
"I just don't like saying it out loud."EmbarrassedYoung professionalWould you buy a second-hand phone?

Question builder

Log one answer as a CSV row

Type one answer from your footage, then copy the CSV row into your sheet.

CSV row

  1. "My mother would say it is a waste","guilty, half joking","first-job earner, lives with family","Last thing you bought just for yourself?"

Caption draft

  1. "My mother would say it is a waste" Who else has heard this at home?

Follow-up to test

  1. Next shoot, ask: when did you last hear "My mother would say it is a waste", and what did you do next?Personal experiences

Opening line

  1. We asked: Last thing you bought just for yourself? One answer: "My mother would say it is a waste"Unexpected questions

Header row, pasted once at the top: Exact words,Emotion,Archetype,Question. Never log names or phone numbers.

Then read the log once a month. Sort by question and see which ones keep producing stories. Sort by emotion and look for the feeling that repeats. Search for a word like "waste" or "mother" and count how often it comes up. The lines that keep returning are your next captions.

Our vox pop market research guide turns the same log into a full research method.

Two rules keep the log safe. Store words, not identities: no names, numbers or faces in the sheet. And don't paste raw footage of identifiable people into third-party tools without their consent. A video of an identifiable person is personal data, and India's data protection rules are coming into force in phases through 202712.

Framework

OWNED INTELLIGENCE

Seventeen letters, read two at a time.

  1. O, W

    Output and words

    Lines no copywriter would write.

  2. N, E

    Narratives and emotions

    Stories matched to who felt what.

  3. D, I

    Data and intelligence

    Which questions bring out which answers.

  4. N, T

    Nobody else has it, transcripts beat guesses

    Your log comes from real conversations.

  5. E, L

    Every shoot adds, language becomes copy

    Hooks come straight from the sheet.

  6. L, I

    Longer is deeper, insights take time

    This can't be rushed or bought.

  7. G, E

    Gaps show, each insight feeds the next

    Findings shape the next question.

  8. N, C

    Net result, competitive edge

    Know your buyer better on the same budget.

  9. E

    Equity

    It stays with the brand, whoever runs the account.

Playbook 8: the INVISIBLE category frame behind a purchase

Every category has an unspoken belief that decides how people buy, and you won't find it by asking about your product. Ask one step away from it, and listen for the reason behind the reason. That belief is the INVISIBLE category frame.

Buyers rarely think about a category the way its brands do. Brands talk about glow, comfort or performance. Buyers carry a private model built from habit, disappointment and family. It seldom shows up in a survey or a dashboard, because nobody asks the question that would bring it out.

Ask about the product, or one step away

Direct

  • "Why don't you buy skincare?"
  • Invites a defence
  • Gets "too expensive" or "no time"

One step removed

  • "What's the last thing you bought just for yourself that cost more than five hundred rupees?"
  • Invites a memory
  • Shows how people feel about spending on themselves

The second question answers the first.

Imagine the answers to that second question come back as "I can't remember", "my mother would say it's a waste" and "I feel I should spend it on the house". None of them mentions skincare. All of them describe the real frame: guilt about spending on yourself.

A brand still talking about glow is answering a question nobody asked.

Name that belief on camera, in the participants' words, before you challenge it. After that, the product can enter as the thing that eases the guilt.

Framework

INVISIBLE

Nine letters for finding the belief behind the purchase.

  1. I

    Identify

    The unspoken belief about the whole category.

  2. N

    Name it

    In their words first.

  3. V

    Validate

    Show it on camera before you push back.

  4. I

    Introduce late

    The product comes in once the belief is seen.

  5. S

    Surface the gap

    Between what they believe and what's possible.

  6. I

    Invert the story

    Tell the category from the buyer's side.

  7. B

    Build on it

    Base your messaging on the belief.

  8. L

    Layer

    Every later piece sits on it.

  9. E

    Everyone else

    Rivals talk product. You talk about the buyer's world.

Some titles in our library point the same way. CircleStore's "Acceptable Vs Taboo" and "Branded Vs Unbranded" are, by their titles, about norms and labels more than any one product. That is where hidden beliefs tend to show.

When this doesn't work: purchases that really are about features, like replacing a broken appliance. Ask comparison questions there instead.

Playbook 9: the GOSSIP architecture for sharing

Build clips that people pass to each other, instead of messages the brand pushes out. Broadcast reach follows spend, while gossip reach follows relevance. The GOSSIP architecture tests one thing: does the clip give someone something to say?

Broadcast or gossip

Broadcast

  • The brand pushes a message out
  • Reach rises with spend
  • "Our new bedsheets are the softest in India. Shop now."

Gossip

  • One person passes it to another
  • Reach rises with relevance
  • "Settle this for us: do you make your bed every morning?"

Few people forward the first. The second is a conversation waiting to start.

People share for their own reasons: to look informed, to start an argument, to tell someone something before they hear it elsewhere. A clip needs no gimmick for that, only a real tension that two reasonable people could disagree about.

Five caption and opening templates, with blanks for you to fill:

  1. "We asked people at [place] whether they've ever [small, shared habit]. Watch who hesitates."
  2. "Settle this: is [option A] or [option B] the normal way to [everyday task]?"
  3. "Nobody at [place] could agree on [question]. Where do you stand?"
  4. "Send this to the one friend who [does the habit]."
  5. "We asked [kind of person] what they'd never buy [new, second-hand or online]. Here's what they said."

Keep the brand as the one who asked, and let viewers reach the conclusion. We think implication travels further than a statement, and it keeps you clear of claims you can't support.

One rule on numbers. If a caption uses a count, make it a real count from your own footage: "Of the people we asked on [day], [number] said yes." Never round up for drama, and never invent the number.

Framework

GOSSIP ARCHITECTURE

Eighteen letters, read two at a time.

  1. G, O

    Ground it, open on a person

    A real tension, and someone the viewer thinks of.

  2. S, S

    Something inside, seed a question

    News that feels early, with no clean answer.

  3. I, P

    Introduce friction, position the brand

    Mild disagreement. The brand only asked.

  4. A, R

    Anchor early, remove brand voice

    The shareable moment first, no marketing lines.

  5. C, H

    Create a reason to tag, hide the ask

    Skip "tag a friend". The moment is the ask.

  6. I, T

    Imply, then test

    Does it give someone something to say?

  7. E, C

    Easier to share, count sends

    Track shares as closely as views.

  8. T, U

    Their reasons, understood

    People share for social reasons.

  9. R, E

    Relevance, exit broadcast

    Stop writing for a crowd.

When this doesn't work: topics where disagreement hurts someone. Mild friction about bedsheets is gossip. Friction about religion, caste or bodies is a pile-on. Don't mock anyone, and drop any answer that would embarrass the person who gave it.

Playbook 10: TRUST arbitrage, priced per usable moment

Compare formats by what each costs per moment you can actually use. Work out your cost per usable moment from your own shoot, then set it beside what your current creative costs per finished piece. That comparison is TRUST arbitrage.

The name points to a possible gap: trust bought through polish may cost more than trust found on a footpath. If that gap exists for your brand, it's worth acting on.

We won't tell you it exists. We'll show you how to measure it.

Kreators.Army never promises returns, conversions or a lower cost per customer, because those depend on your platform, audience, offer and spend. What you can measure from the first shoot is the input: how much each usable street moment costs you.

Usable moment
A clip from a street shoot that you would publish: a complete answer, a clear reaction or a valley-to-peak turn, with clean audio and a signed release.

Count strictly. A great answer with an auto horn blaring through it doesn't count. Neither does a funny line from someone who walked off before signing a release. If you'd have to explain the clip before posting it, leave it out.

Calculator

Cost per usable moment

Put in your own figures from one shoot. Nothing here is a benchmark.

What the shoot and edit cost you, in rupees (₹)
Answers you filmed (answers)
Usable moments among them (moments)
Finished posts made from those moments (posts)
Cost of one finished piece in your current format, in rupees (₹)
Cost per usable moment
₹3,000
Cost per finished post
₹5,000
Share of filmed answers you could use
25 %
Current cost per piece, divided by street cost per post
3.0 ×

Worked out with the starting values above. Change them on the page to use your own.

Your own numbers, for comparison only. The starting figures are placeholders, not benchmarks or our prices, and a usable moment is an input, not a result.

Read the last figure with care. Above 1 means a street post costs you less than a piece of your current creative. Below 1 means it costs more. Neither tells you which one will sell, so run both against the same goal and let your own analytics decide. For what a done-for-you shoot costs, see how much a vox pop video costs in India.

Framework

TRUST ARBITRAGE

Fourteen letters, read two at a time.

  1. T, R

    Track today, run one shoot

    Today's cost per piece, then one street shoot.

  2. U, S

    Understand the gap, scale what holds

    Back what holds up in your own numbers.

  3. T, A

    Test the yield, add it up

    No gap exists until you've done the sum.

  4. R, B

    Reuse and build

    One shoot, several posts, a growing question bank.

  5. I, T

    Invest once, think in runs

    Set-up carries over. Judge a run of shoots.

  6. R, A

    Review and avoid vanity

    Never compare on views alone.

  7. G, E

    Gauge honestly, exit by evidence

    Count only moments you would publish.

How do the ten playbooks fit together?

They run as a loop. Playbooks 1, 4, 5 and 8 shape the question. Playbooks 2, 3 and 9 shape the edit and how it spreads. Playbooks 6, 7 and 10 make the whole thing repeatable and measurable.

Use the table to pick the two or three that fit where your brand is now.

The ten playbooks: when each applies and what to measure
PlaybookWhen it appliesWhat to measure
1. KIRANA principleWriting questions, every shootComments that quote a line back
2. FORWARD testBefore every postSends and shares, as well as views
3. PEAK-VALLEY mapEvery editHow many viewers stay past the opening
4. FIRST QUESTION advantageOnce a quarterWhether people link the question to you
5. SKEPTIC flipCategories with doubt or past disappointmentWhere viewers drop off after the doubt
6. VELOCITY gapWeekly, or around launchesQuestions tested in a month
7. OWNED intelligenceAfter every shootRows logged, and lines reused in copy
8. INVISIBLE category frameRepositioning, or a stalled categoryNew beliefs found per shoot
9. GOSSIP architectureCaptions and openingsSends, saves and tags
10. TRUST arbitrageBudget reviewsCost per usable moment over time

Starting from zero, we'd add them in this order:

  1. Start the log (7) on your first shoot. Everything else feeds it.
  2. Edit with the PEAK-VALLEY map (3), and run the FORWARD test (2) on every post.
  3. Once the log holds a few shoots, find the invisible frame (8) and your first question (4).
  4. Add the SKEPTIC flip (5) and gossip openings (9) when you know the objections well.
  5. Work out your cost per usable moment (10) after three shoots.

Which mistakes undo the playbooks?

Six mistakes undo most of this work, because each one breaks the trust the playbooks depend on.

  • Staged answers. If anyone on camera was paid, cast or works for you, India's endorsement guidelines expect clear labels such as "advertisement" or "paid promotion"13. A staged clip presented as spontaneous misleads.
  • Over-cleaned edits. Peaks alone read as performance. Cut the dead air and keep the thinking.
  • The brand too early. A logo or pitch in the first seconds says "ad" before anyone has said anything worth hearing.
  • No consent. Get a signed release from everyone you feature, say where the video will appear, and never film anyone under 18 without a parent's written consent.
  • Testimonials as proof. A happy stranger can't prove a product works. Keep health and efficacy claims out of street answers entirely.
  • Judging on views alone. Views say a clip was seen. Your log, sends and cost per usable moment say whether the system works.

Consent deserves more than a bullet. Data protection duties in India are arriving in phases, so have a lawyer check your release form before you rely on it. This is general information, not legal advice, and our guide to filming consent in India goes further.

There's also a limit worth being honest about. We control the question, the setup, who we approach, the follow-up and the edit. We can't control exactly how a stranger will respond in the moment.

Every playbook here is built around that limit.

Where should you start this week?

Write one question about your buyer's day. Leave your product out of it. Ask it to three people you don't work with. If any of them answers with a story, you have a street interview question and the first row of your log.

Before you write a brief, watch a few real street interviews in our video library and note which ones made you think of a particular person.

If you'd rather hand over the footpath, Kreators.Army plans, shoots and edits street interviews for D2C brands. See the format on your own brand first: scope, the number of trial videos and any fee are confirmed in writing. Request a trial video.

Frequently asked questions

What are street interview ads?

Street interview ads are paid or boosted posts built from unscripted street interviews. A host asks people in public one question, and the best real answers are cut into a short video for Reels, Shorts or other feeds. Street interview ads differ from scripted ads because nobody on camera reads lines, and from influencer posts because the speaker is a stranger rather than a known creator.

Do street interview ads work?

Street interview ads can work, but no format works every time. In a 2022 experiment, vox pops in TV news shifted viewers' personal opinions more than an expert quoting a poll did3. Results still depend on the platform, audience, offer and spend, so test street interview ads against your current creative. Kreators.Army does not guarantee views, reach or sales.

Are street interviews staged?

Good street interviews are not staged. The brand plans the question, the place and who to approach, but not what people say. Some videos online do use actors or scripted answers. If anyone on camera was paid, cast or works for the brand, India's endorsement guidelines expect that to be disclosed clearly. Kreators.Army does not script what strangers say.

Do street interviews suit D2C brands?

Usually, yes. Street interviews suit D2C brands with physical products best, because people can hold, taste or try the product and react on camera. They also suit brands whose buyers have a strong habit or doubt worth discussing. They suit a category less well when the main question is safety or medical efficacy, which a stranger's reaction can't answer.

How often should a brand run street interviews?

Often enough to keep learning, which in our view means a steady rhythm rather than one big shoot a year. One approach is to anchor each quarter on one owned question, then shoot around launches and moments in your category. Logging every shoot matters more than the exact frequency, because the log is what sharpens each new street interview.

How do you measure street interview content?

Measure street interview content at two levels. For inputs, count usable moments per shoot, cost per usable moment and rows added to your log. For attention, check how many viewers stay past the opening, which YouTube reports as "stayed to watch", plus sends, saves and comments that quote a line back. Judge sales with your own analytics, against your current creative.

Can street interviews be used in paid ads?

Yes, if each participant's signed release covers advertising and the usage rights are yours. Kreators.Army's videos can be used on your own channels and in ads on the terms of the pilot agreement, and usage rights transfer on full payment. Trial videos are different: they are for evaluation only and can't be published or used in ads.

What do street interview ads need legally?

Street interview ads need a signed release from everyone featured, saying where the video will appear, including ads. Disclose any paid, cast or staff participants, and never feature anyone under 18 without a parent's written consent. India's data protection rules are coming into force in phases through 2027, so have a lawyer check your forms. This is general information, not legal advice.

Sources

  1. Beyond martech: Building trust with consumers and engaging where sentiment is high (opens in a new tab), Nielsen, 2021
  2. India's internet user base tops 950 million as rural growth, AI drive adoption (opens in a new tab), Business Today, on the IAMAI and Kantar Internet in India 2025 report, 29 January 2026
  3. Power of the people or the expert? The influence of vox pop and expert statements on news-item evaluation, perceived public opinion, and personal opinion (opens in a new tab), Communications 47(1), De Gruyter Brill, 2022
  4. India lifts WhatsApp payment curbs (opens in a new tab), TechCrunch, 31 December 2024
  5. Instagram shares algorithm insights for 2025 (opens in a new tab), Social Media Today, January 2025
  6. Understand your YouTube content performance (opens in a new tab), YouTube Help
  7. India slips to third place in Edelman Trust Barometer 2025 rankings (opens in a new tab), Adda247 Current Affairs, on the Edelman Trust Barometer 2025, 2025
  8. Voice of the Consumer Survey 2024 (India), press release (opens in a new tab), PwC India, July 2024
  9. The ASCI Code (opens in a new tab), Advertising Standards Council of India
  10. Opinion balance in vox pop television news (opens in a new tab), Journalism Studies, via Drexel University Research Discovery, 2018
  11. Instagram expands access to Trial Reels (opens in a new tab), Social Media Today, 2025
  12. Govt notifies DPDP Rules; implementation planned in phases spread over 12-18 months (opens in a new tab), PTI, via The Week, 14 November 2025
  13. Endorsement Know-hows! (opens in a new tab), Department of Consumer Affairs, Government of India, 20 January 2023

Want to see a vox pop made with your product?

See the format on your own brand first. Scope, the number of trial videos and any fee are confirmed in writing.

Request a trial video
Browse video libraryFinished vox pops, by brand.