UGC Ads That Convert: How to Brief a Creator.

The video came back better than you hoped. Good light, a likeable person, your product in their hands. You put a few hundred dollars of spend behind it and waited for the orders. They didn't come. Click-through flat, cost per purchase climbing, and a quiet suspicion that you'd been sold on UGC for nothing.

Here's what almost nobody tells you first: the video probably wasn't the problem. The brief was.

A UGC ad that pays for itself and one that quietly drains budget usually start from the same place. Same kind of creator, same product, sometimes the same talent. What separates them is what the brand asked for before anyone hit record. That's good news, because the brief is the one part of this you fully control.

You're not here to be convinced UGC works. You already suspect it does, and the research backs you: in Bazaarvoice's 2024 study of more than 8,000 consumers, 65% said they lean on user-generated content like photos, videos, and reviews when deciding what to buy.¹ So the question isn't whether creator content sells. It's why some creator ads convert and others burn money. This guide is the answer, in the order you'll actually face it: what to commission, how to brief it, how to script it, and how to read whether it worked.

Why a video that crushed on a feed can still die as an ad

Start with the trap, because it catches almost everyone once. You get back something genuinely good. It would do well on the creator's own feed. You promote it, and it flops.

That's not UGC failing. It's a video made for a feed being asked to do the job of an ad, and those are two different jobs.

Organic content earns its attention. Someone follows the creator, so a post can wander for ten seconds before it gets going and a fan will still stick around. An ad has no such patience. It interrupts a stranger who didn't ask to see you, and it has about three seconds to stop the scroll, land a point, and earn a click. Same creator, same product, completely different edit.

The payoff when you get it right is real. Emplifi's Q3 2025 benchmarks found that social posts featuring UGC drove 10.38 times higher conversion rates than posts without it.² (That figure covers social posts broadly, not paid ad creative specifically, so don't read it as "UGC ads convert ten times better." Read it as the direction every ad account eventually confirms: when content looks like a person made it, people act on it.) The lift is there. Whether your ad captures it comes down to the decisions below.

Decide what to commission (and order a slate, not a video)

Before you write a word of a brief, decide what you're buying. Two decisions save the most money here.

Pick the format that fits where the viewer is. The type of content that converts shifts across the funnel:

• Cold traffic, people who don't know you yet. Hook-led testimonials, unboxings, and in-use lifestyle shots. The job is to stop the scroll and spark curiosity, so the opening carries everything.
• Warm traffic, people who've seen you but haven't bought. Testimonials, before-and-afters, and honest comparisons. Specificity sells here: "it cleared my breakouts in two weeks" beats "I love this."
• Retargeting, people who almost bought. Demos and quick social-proof cuts. They need a reason to finish, not an introduction.

If you're only commissioning one thing, make it a hook-led testimonial. It's the most reusable asset you can own: it works as cold creative, recuts for retargeting, and gives you raw material for a dozen variations later. For a wider menu, these are the video formats brands commission and how to brief each one.

Then order a slate, not a single video. This is the part first-timers most often get wrong. You're not commissioning one perfect ad. You're commissioning a small batch of distinct concepts, maybe three to five different hooks or angles, knowing most will lose and one might carry the whole account. Trying to perfect a single video is the exact behavior that flops, because you can't predict which opening will land. Test several cheaply, then put your budget behind the winner.

That reframes the money question too. A single creator-made video commonly runs from a few dozen to a couple of hundred dollars depending on the creator and the length, and if you run it as a paid ad rather than just posting it, expect a separate usage fee on top, often billed for as long as the ad stays live. Those two costs are different things, and first-timers routinely confuse them. So budget for a small test batch, not one expensive hero video. (The full breakdown lives in our UGC pricing guide.)

The brief that's actually built for ads

Most "UGC brief" advice hands you a blank template with fields like "tone" and "key messages." That's the foundation, and we cover the six things every brief needs in how to write a brief that gets great content. An ad brief is that foundation plus a layer the organic version skips, and that layer decides whether the footage is usable as paid creative at all.

Here's the ad-specific layer to add on top of a normal brief:

• Hook directions, not one line to read. Give the creator three to six opening angles to try, not a single scripted sentence (more on the patterns below). Nothing else in the brief moves performance as much.
• Multiple hooks filmed in one session. Ask for the same core video with several different openings shot back to back. Now you can test openers without commissioning a whole new video each time.
• A clean master with no baked-in captions or music. Request the footage without burned-on text or a soundtrack, so you control captions and audio per placement and per test.
• At least one CTA-free cut. A version without a hard call to action can run as top-of-funnel or mid-funnel creative, where a sales-y close would scare people off. One shoot, more places to use it.
• Specs for where it's running. 9:16 for Reels, TikTok, and Stories; 4:5 for Feed. Keep most ads in the 15 to 30 second range, and ask for a captioned cut, because most feed video plays muted. Tell the creator the placement so they frame for it and keep anything important out of the corners where the platform's buttons sit. Our platform-by-platform breakdown weighs which formats and specs each channel rewards.
• Two or three key messages, maximum. An ad that says five things says nothing. Pick the points that matter and cut the rest.
• The exact CTA wording, and where it lands. Spoken at the end, a text overlay, or both. Don't leave it to chance.
• What not to over-polish. This is the one most brands get backwards. The instinct, once you're paying for an ad, is to make it look professional: a logo intro, stock music, slick color grading. That polish is exactly what tips a viewer off that it's an ad, and the authenticity you paid for evaporates. Say plainly what you don't want.

One more line belongs in the brief, and it bites hardest if you skip it: rights and disclosure. If you're going to run the content as a paid ad, agree paid-usage rights in writing before the shoot, not after, and make sure any paid partnership is disclosed per FTC guidelines. (If you later want to run the ad from the creator's own handle for extra trust, that's whitelisting, and it has its own setup, covered in UGC whitelisting and Spark Ads.) Worth burning in: a claim the creator makes becomes your liability as the advertiser, so you can't run "this cured my eczema" even if they said it unprompted.

A 30-second script, beat by beat

Give a creator a structure to work from rather than a script to read, and you'll get something that sounds like a person. Here's a skeleton that works for most direct-response goals. Say you sell a magnesium drink for sleep:

Each beat earns the next. The hook buys the problem, the problem makes the turn feel relevant, the proof makes the CTA believable. Lose the first three seconds and none of the rest gets watched.

For the hook specifically, give the creator named directions to choose from:

• Problem-first: open on the pain before the product appears. "My white sneakers were gray by week two."
• Surprising result: lead with an outcome that sounds slightly too good. "This $14 thing replaced the $80 one on my shelf."
• The question they're already asking: say out loud what they'd type into a search bar. "Does anyone actually know how to get gel polish off without wrecking your nails?"
• "I was skeptical until...": honest doubt is disarming, and a creator voicing the buyer's own objection reads as a peer, not a pitch.
• Pattern interrupt: a visual that breaks the scroll before a word is spoken, useful when the product is visual and the feed is noisy.

The one opening to ban: "Hi guys! So today I want to talk about this amazing product I've been loving…" Three seconds gone, no problem, no payoff, and it announces itself as an ad before anything happens. For more on what makes an opening land, video hooks goes deeper. And remember length is a function of placement, not a rule: cut a tighter ten or fifteen second version for cold traffic and let the longer one run in retargeting.

The five-minute check before you spend

When the files come back, don't promote anything yet. Run a quick check, because this is the cheapest moment to catch a problem:

• Does the hook land in the first three seconds with the sound off?
• Is there a captioned cut, since most feed video is watched muted?
• Did you get the clean master and the raw clips you asked for, so you can make variations later?
• Would it pass the platform's ad review?

Five minutes here saves a week of spend behind a video that was never going to work. If something's off, give feedback like you're talking to a person, not correcting a machine. Be specific about the hook or the CTA ("the open is too slow, can we start on the problem?") and leave the personality alone. Asking a creator to sharpen a weak hook gets you a better ad. Asking them to sand off everything human gets you the polished nobody-look you were trying to escape.

Reading the results without a measurement degree

You don't need an attribution stack to run UGC ads. You need to read a few numbers as instructions for what to fix. (For the full measurement setup, that's its own guide: how to measure UGC ROI.)

Watch the hook rate first, the share of people who keep watching past the first few seconds, then your cost per purchase against your product's margin. Between them they tell you most of what you need:

• Hook rate is low. It's the opening, not the product. Reshoot the first three seconds with a different hook direction. This is exactly why you asked for several hooks up front.
• Click-through is fine but cost per purchase is high. The ad did its job; the problem is after the click. Look at the landing page, the price, or the offer, not the creative.
• It's working. That's your signal to put more budget behind it, and to make variations of the winner before it tires.

Two rules worth burning in. First, don't put your UGC video and a static image in the same ad set to "see which wins." The auction picks an early front-runner on first delivery signals, not on which creative is actually better, and starves the other before it gets a fair read. Test one variable at a time, in its own set. Second, expect UGC to tire faster than polished brand creative, sometimes within weeks once spend ramps up, because a specific face and story get familiar quickly. When it fades, the fix usually isn't a brand-new shoot. It's a fresh hook on footage you already have, which is the whole point of turning one video into many ad variations.

That's the loop: the winner gets iterated, the losers tell you what to put in the next brief. Each round you learn what your audience actually responds to, and your briefs get sharper.

When AI UGC makes sense (and when it doesn't)

Search "UGC ads" today and you'll hit a wall of AI tools promising a finished ad in sixty seconds, no creator required. It's worth knowing honestly where they fit, because they do have a place.

AI generators are fast and cheap, which makes them handy for one job in particular: spinning up a high volume of hook and angle variations to test when you don't yet know what works. If you need twenty openings to throw at cold traffic this week, that's a strength.

Where human creators still win is the believable part. A person's hands actually using the product, a testimonial a viewer trusts, an unboxing reaction that isn't staged, the small unscripted moments that make a stranger stop and think "that could be me." That trust is the whole reason UGC converts, and it's what audiences say they want: in EnTribe's 2023 survey of over a thousand US consumers, 90% said they'd prefer to see brands share content from actual customers, and 86% said they're more likely to trust a brand that publishes UGC than one that leans on influencer posts.³ There's also a rising expectation, and in some places a legal one, that synthetic "people" in ads get disclosed.

So treat it as a mix, not a fight. Plenty of brands use AI to test angle volume and human creators to make the hero creative that carries the conversion. We go deeper on the trade-offs in UGC vs AI-generated content.

The mistakes that quietly drain budget

Most wasted UGC ad spend traces back to a short list:

• Over-polishing. A logo intro, stock music, heavy grading. The moment it looks like an ad, it gets scrolled like one. The rough edges are the feature.
• The wrong aspect ratio for the placement. A cropped or letterboxed video reads as recycled and out of place.
• No hook in the first three seconds. Opening on a beauty shot or a slow pan kills the ad before it starts.
• Running organic content as-is. A clip that did well on a creator's feed usually needs a new hook, a tighter cut, and a clear CTA before it works as paid.
• Over-scripting the creator. Word-for-word reads sound like ads. Brief the angle, not the lines.
• Betting on one video. Commission a slate, expect most to lose, scale the winner.
• Testing UGC against a static image in one ad set. It hands you a false winner. Give each its own set.

Frequently asked questions

What are UGC ads? Creator-made photos and videos, a testimonial, an unboxing, a demo, that a brand owns and runs as paid advertising. The creator makes the content; you deploy it as your own ad creative to your own target audience. The look is authentic and feed-native; the use is paid media.

Are UGC ads legal? Yes, with the basics handled. Agree usage rights in writing before the shoot, disclose paid partnerships per FTC guidelines, and use licensed or platform-provided music rather than a track lifted from somewhere. Remember that a claim the creator makes becomes your liability as the advertiser, so you can't run an unsupported health or earnings claim even if they said it unprompted.

How is a UGC ad different from working with an influencer? An influencer post buys their audience's attention, so you're paying for reach. A UGC ad buys the content itself, so you own it and show it to your own audience through your ad account. You can do both; they solve different problems.

Do I have to ship my product, or can the creator use something they already own? Usually you provide the product so they can show your actual packaging and features. For some categories a creator can work with something they own or a service they can access, but if the ad needs your specific product on screen, plan to get it into their hands.

How much do UGC ads cost? A single video commonly runs from a few dozen to a couple of hundred dollars, with a usage fee on top if you're running it as paid. Budget for a small batch of concepts rather than one expensive video, since you're testing to find a winner. It's well below studio production, and you can start with a few.

What about the AI UGC tools I keep seeing? They're good for generating lots of hook variations to test cheaply. For the creative that has to earn a stranger's trust, a person using your actual product still does it better. Most brands end up using both.

Start with the brief

The brands getting UGC ads that convert aren't spending more. They're briefing better. A clear hook direction, the right format for the funnel stage, the correct specs for the placement, a slate instead of a single bet, and honest feedback to the creator will out-perform a bigger budget behind a vague brief almost every time.

Once you know what to ask for, you need someone to make it. That's where a marketplace like Modliflex fits: you browse and filter creators, send your brief, and get the photos and videos back, with payment held in escrow until you approve the work. A person using your product, made for your ads, not an avatar reading a script.

Footnotes

1. Bazaarvoice, Shopper Experience Index Vol. 18 (November 2024), Savanta survey of 8,000+ consumers across seven countries. The report found that 65% of consumers rely on UGC, such as ratings, reviews, photos, and videos, in their buying decisions. https://www.bazaarvoice.com/press/bazaarvoice-shopper-experience-index-vol-18-88-of-shoppers-want-an-omnichannel-experience-a-third-of-shoppers-say-that-includes-social/ ↩

2. Emplifi, Q3 2025 Social Media Benchmarks Report (October 2025): social media posts featuring user-generated content "drove 10.38X higher conversion rates compared to non-UGC posts." Figure reflects social posts generally, not paid ad creative. https://emplifi.io/press/ugc-delivers-10x-higher-conversion-rates/ ↩

3. EnTribe survey of 1,000+ US consumers (April 2023): "90 percent stated they would prefer to see brands share content from actual customers," and "86 percent of respondents mentioned they are more likely to trust a brand that publishes user-generated content as opposed to influencers." https://www.globenewswire.com/news-release/2023/06/06/2682877/0/en/New-EnTribe-Survey-Reveals-User-Generated-Content-Impacts-Consumer-Purchases-More-Than-Social-Media-Influencers.html ↩