UGC for ecommerce ads: the complete guide

Be specific about what you mean before you commission anything, because the word covers four different things with four different costs and uses.

The first is content customers made on their own, without being asked: a photo in a review, a video posted about your product, a message sent to your support inbox. This is the most believable and the least controllable. You did not commission it and you cannot direct it, but you can ask permission to use it.

The second is content you asked existing customers to make. You reach out to people who already bought, give them a light brief, and offer something in return. It keeps most of the credibility because the person genuinely uses the product, and you get some influence over what gets filmed.

The third is creator content, made by people who film product videos as work. They may not have an audience worth speaking of, and that is fine, because you are buying footage rather than reach. It is the fastest way to get volume and the most directable, and it is also the easiest to spot as an advert if the brief is poor.

The fourth is content from someone with an audience, where part of the value is that their followers see it. That is a different arrangement with different pricing and different disclosure obligations, and it should be judged on reach as well as on footage.

Decide which of the four you need from the job the creative has to do. If you need volume for testing, creators are the practical route. If you need credibility for a product that people are sceptical about, customers are worth the extra effort. If you are trying to reach a specific community, someone with an audience in that community is what you are buying. Most stores end up using more than one, and the mistake is paying creator prices when a customer photo would have worked better.

Both kinds of content have a job, and the useful decision is which one goes where rather than which one is better. Choosing by task rather than by preference is what stops a store from having a folder of unusable video.

Customer and creator footage earns its place where the buyer's problem is belief. Does it fit in a normal kitchen. Is it as heavy as it looks. Does the colour match the photo. Is it hard to assemble. Does it work on hair like mine, in a room like mine, on a bike like mine. A person showing that in an ordinary setting settles the doubt in a way a lit product shot cannot, because the setting itself is part of the evidence.

Studio and controlled photography earns its place where the buyer's problem is detail. Texture, stitching, finish, the exact colour, what is included in the box, how the parts fit together. It is also what your product feed and your product page hero images need, because those have to be clean, consistent and accurate.

In practice most ads use both. A common structure opens with a person in a real setting and cuts to a close, controlled shot when the ad needs to show what the product actually looks like. That combination handles belief and detail in the same fifteen seconds.

There is also a middle option many stores forget: your own footage, filmed by you, in a real place. It costs nothing, it is fully directable, and it is honest as long as you do not pretend it is a customer. Packing an order, opening a delivery, using the product on your own bench: these are cheap to shoot and they often test well.

Decide per angle rather than per campaign. Some reasons to buy need a face and a room, some need a close up and a caption, and briefing them the same way wastes the strength of both.

Start sourcing inside your own customer list, because those people are cheaper to reach, more credible on camera and already know the product. Only go outside when you need volume or a specific look you cannot find among your buyers.

The simplest route is your post purchase email. A short message a few weeks after delivery asking whether they would share a photo or a short clip of the product in use, with something in return, produces a steady trickle at almost no cost. Make it easy: one link, clear instructions, no account to create.

Your reviews are the second source. People who left a detailed positive review, especially one with a photo, are the obvious people to ask for more. Reach out individually and reference what they said, because a personal message gets a far better response than a mass send.

Your inbox and social mentions are the third. People who already tagged you have shown they are willing, and asking permission to use something they already posted is a small request. Keep a list of who has said yes so you are not asking the same people repeatedly.

When you go outside, be clear that you are hiring for footage. Look at whether the person's existing videos show products clearly, whether they speak in a way that suits your buyers, and whether the setting matches where your product is used. Follower count is irrelevant if you are only running the footage as an ad.

Build a small roster rather than a long list of one time collaborations. Working with the same handful of people repeatedly means less briefing, more consistent quality, and a group who genuinely know the product. It also lets you plan a monthly batch instead of scrambling every time a test needs new assets.

Aim for variety across the roster. Different ages, settings, skin tones, body types and use cases produce different results, and the only way to find out which resonates is to have the range available.

Give the person a structure and let them use their own words. A tight script produces something that sounds like an advert read aloud, which removes the only advantage this kind of content has.

Write the brief as beats. A workable set is: open by showing the product or naming the problem in your own words, say what you were using or doing before, show the thing being used, point at the detail that surprised you, and finish with who you would recommend it to. Five beats give you enough structure to edit with and enough freedom for the person to sound like themselves.

Include the angle you are testing, in plain language. If the ad is about how quickly it sets up, say that the video needs to show the setup happening, in real time, with no cuts. If it is about durability, ask for the specific thing that proves it. Vague briefs produce vague footage.

Be explicit about the technical requirements, because these are where footage most often becomes unusable. Film vertically unless told otherwise. Film in landscape orientation as well if you need it for other placements. Keep the phone steady or use a support. Film in daylight or good light. Record where it is quiet. Do not add music, captions, filters or effects, because you will add those in the edit. Send the original files rather than a version that has been through a social app, which compresses them.

Say what is not allowed, too. Health, safety and performance claims your product cannot support. Comparisons to named alternatives. Anything about results the product cannot produce. If you sell in a regulated category, list the specific phrases that cannot be used, because a creator will not know them.

Finally, ask for more than you need. A person filming five short takes of the same beat costs the same as one and gives the editor options. Requesting variety at the shoot is far cheaper than asking for a reshoot.

Sort the paperwork before you spend money promoting someone else's face, voice or footage. This is the part stores skip and the part that causes real problems later.

Get permission in writing, and be specific about what it covers. A comment saying yes on a social post is not a licence to run a paid advert. Your agreement should say where the content can be used, including paid advertising, your website, email and any marketplace listing. It should say how long for, whether you can edit it, and whether you can use the person's name or face. If you want to keep running a strong asset in a year, an open ended or long term licence is worth negotiating at the start, when it costs very little.

Make sure the person owns what they are giving you. If they filmed it themselves with their own phone in their own home, that is straightforward. If a photographer took it, if there is recognisable music playing, or if other people appear in it, the rights are not theirs alone to give.

Disclosure is a separate obligation from permission. Where someone has been paid, given free product, or has any other material connection to your business, that relationship has to be disclosed clearly to the people seeing the content. This applies to advertising authorities in most markets, and the platforms have their own tools and labels for paid partnerships that should be used where they apply. Do not rely on a hashtag buried in a caption.

Keep the records somewhere findable: who, what content, what the licence covers, when it expires, what they were paid or given. Attach it to the asset in your library. When someone asks whether you can run a particular video in a new market, you want to answer from a record rather than from memory.

Finally, respect withdrawal. If a customer asks you to stop using their content, stop, including in campaigns that are performing. It is the right thing to do and the alternative is a public complaint about your brand.

Give the person a shot list so one filming session produces everything the edit needs. Going back for a missing shot costs more time than asking for it in the first place.

The moments that get used most often are predictable. The unboxing or first look, which works as an opening because it carries curiosity. The product in the setting it belongs in, shot wide enough that you can see the space. Hands using it, close, so the viewer understands the mechanics. The detail that surprised them, held up to the camera. A talking segment where they say what changed. And a closing line about who it suits.

Add product specific shots from your own angle list. For clothing that means movement, fit on a real body, and the fabric close up. For anything assembled, the assembly happening at real speed. For consumables, the routine it fits into. For anything with a size question, the product next to something everyone recognises.

Ask for both orientations of the key moments if you plan to run the footage in more than one placement, or ask for it filmed with headroom so it can be cropped. Cropping a tightly framed vertical video into a square usually cuts off the product or the face.

Request a few seconds of quiet at the start and end of each take, and ask people not to speak over the moment they are demonstrating. Editors need clean handles to cut on and clean audio to work with.

When the footage arrives, review it against the shot list before you pay or before the person moves on to another job. It is much easier to ask for one more take that week than to arrange a reshoot a month later. Keep everything usable, including the takes you do not cut into this ad, because the same footage will feed the next round of tests.

Cut the footage for where it will run, and put the strongest evidence at the front. Raw creator footage is rarely an ad; the edit is what makes it one.

Start with the hook. The strongest opening is usually the moment that shows the product doing something, or the person saying the problem in the same words your customers use. Resist opening with a logo, a slow reveal, or a greeting. If the best moment is thirty seconds in, move it to the front and rebuild around it.

Then assemble the middle in the order a sceptic needs. Show the thing working. Show the detail that answers the main objection. Cut in a clean product shot where the viewer needs to see exactly what they would receive. Keep it moving, because a static talking clip loses people quickly in feeds.

Add captions, and write them properly. Many people watch without sound, and captions are read even when audio is on. Keep them short, keep them out of the areas where platform interface elements sit, and use them to reinforce the claim rather than to transcribe every word.

End with something specific. What to do, and why now if there is an honest reason. A vague sign off wastes the attention the rest of the video earned.

Produce more than one cut from the same source. A long version, a short version, and a version with a different opening give you three tests from one shoot. Changing only the first three seconds and running the two against each other is one of the cheapest useful tests available.

Keep the finish honest. Heavy colour grading, dramatic music and slick transitions make the video look like the studio content it was meant to contrast with. Light editing usually performs better, and it is faster to produce.

Run this kind of content through the same testing process as everything else, rather than assuming it will win or treating it as a special category. Some products need a face and a room; some sell better on a clean close up. Only a fair test tells you which yours is.

Set up the comparison properly. Same audience settings, same landing page, same offer, same budget, running at the same time. Then vary only the creative. If you run customer footage this month and studio work last month, you are comparing two different weeks of market conditions rather than two approaches.

Decide what you are testing. Person versus no person is one test. One creator versus another is a different test. The same script filmed by two people is a third. Each answers a specific question, and running all three at once answers none of them.

Judge on the same stages you use for any creative: whether people watched, whether they clicked, whether they added to cart, whether they bought. Content of this type often produces high watch time and click rates with weaker purchases, which usually means the video entertained without making a clear case for the product. That is a fixable editing problem rather than a reason to abandon the format.

Expect a shorter working life. Because this content depends on feeling current and personal, it tends to tire faster than a clean product image, particularly with smaller audiences. Plan the roster and the shooting cadence so replacements are ready.

When something wins, work out why. Was it the person, the setting, the opening line, the demonstration or the claim. Write that down as a lesson and put it into the next brief, so the roster gets better rather than just bigger.

An asset that works as an ad usually works everywhere else, so plan for each piece of content to appear in several places. Most stores license footage for a campaign and then let it expire without ever putting it in front of the people closest to buying.

The product page is the first stop. A short clip of the product in use, placed near the buying decision, answers the same doubts the ad answered, for a visitor who is already interested. Photos from customers work well further down the page, alongside reviews, where they act as evidence rather than decoration.

Email is the second. Cart and browse abandonment messages are the obvious place, because the person hesitated and this content addresses hesitation. Post purchase sequences are a quieter opportunity: a clip showing how to get the most out of the product reduces support tickets and returns, and it costs nothing extra since you already own the footage.

Your marketplace and social storefront listings usually accept video and lifestyle imagery too, and those placements are often ignored. Check the image rules for each surface first, because product feeds in particular have strict requirements about promotional text and backgrounds that lifestyle images can fail.

Organic social is the third. Content that performs as a paid ad often performs organically, and reposting with permission gives the creator a reason to work with you again.

Before you do any of this, check the licence. Permission for paid social does not automatically cover your website, your email or a marketplace listing. This is why the rights conversation is worth having properly at the start: the same money buys far more use if you ask for it up front, and it is awkward to go back later and ask for permission for a video that is already working.

Pick a way of compensating people that you can afford to repeat, because this kind of content only works when it is a steady supply rather than a one off project. The model matters more than the amount.

There are a few workable approaches. Product only works with genuine customers and enthusiasts for small requests, particularly photos and short clips, and it is the cheapest way to build a bank of real content. Product plus a fee is standard for creators producing filmed content to a brief, and the fee should reflect the shoot time, the number of deliverables and the licence length. Store credit or a discount code suits customer requests and keeps the relationship inside your own economics. A retainer with a small roster suits stores that need a predictable batch every month and want people who genuinely know the product.

Price the licence, not just the film. Content you can only run for thirty days is worth much less to you than content you can run for a year across every channel. Agreeing broad, long usage at the point of hiring is almost always cheaper than extending later, and it protects you from having to retire an asset that is performing.

Make the process easy on your side. A short agreement, a clear brief, a deadline, a single place to send files, and prompt payment. People who are treated well come back, and a roster of returning contributors is worth more than a constant search for new ones.

Track cost per usable asset rather than cost per collaboration. Someone who charges more but delivers six usable cuts from one session can be better value than someone cheaper who delivers one. Over a few months that number tells you who to keep working with.

Finally, budget for it as part of creative production, not as an extra. If creative is the main variable in your paid performance, the supply of content is a running cost, and treating it as a project that finishes is how stores end up with tired ads and no replacements.