Best CTA Formulas for UGC Video Ads (Ranked by CTR)

The highest-CTR close in UGC is the one that pretends it is not a close. The creator finishes the proof, glances off-camera, and mentions the link as a favour rather than a command. It works because it matches the rest of the video: a person who was sceptical, then convinced, then willing to tell a friend. Adding a shouted 'SHOP NOW' after that performance throws away the trust the first 20 seconds bought.

Write it as a throwaway, then protect the throwaway in the edit. The line should be one breath, no smile-for-the-logo, no sudden music swell. Variants that consistently hold CTR: 'it's in the link if you want to look', 'I wouldn't wait on this one but you do you', 'I already ordered a second, link's there'. Pair it with a quiet on-screen URL or 'link in bio / shop' so muted viewers still get a next step. The formula fails when the offer is complicated — a bundle with three conditions cannot be an afterthought, and forcing it here just sounds evasive.

Do not confuse this with having no CTA. Accounts that skip the close entirely leak clicks from people who were ready. The afterthought is a CTA with the volume turned down, not an invitation to fade to black. If your media buyer insists on a button, keep the spoken line soft and put the hard language on the platform CTA (Shop now, Learn more) where it is expected. That split — human voice soft, button hard — is the version that usually wins the CTR read. If a buyer still wants a spoken command, test it as a separate cell rather than bolting it onto a working afterthought — mixed closes are how you lose the read on which formula did the work.

Social proof at the close, from the only person on screen. Saying you already re-ordered does three jobs in one line: it implies the product lasts, it implies you would spend your own money twice, and it creates a small scarcity ('I wasn't sure they'd stay in stock'). CTR lifts because the next step is modelled, not requested. The viewer is copying a behaviour, not obeying a command. Keep the tone offhand; a proud announcement of the second order sounds like a script and the lift disappears.

The line has to be visually true. Hold the second unit, or the empty first, on camera. A spoken 'I bought extras' with no prop is just a claim, and claims at the close are where scepticism returns. Specifics beat adjectives: 'the 2-pack, not the travel size' outperforms 'I stocked up'. If you cannot show a second unit, show the order confirmation — ugly, screenshotted, personal. That is more persuasive than a lifestyle pack shot. If the second unit is a different colour, say so — specificity is the proof, not the pile.

Compliance is the limit. You cannot imply a stock-out that is not real, and you cannot invent a repurchase. If the product is a one-time durable, this formula is the wrong tool; use a guarantee or a next-step close instead. On subscriptions, 'I didn't cancel after month one' is the honest version and often the stronger one. Keep the rest of the CTA light — this line is the CTA. On digital goods, the honest analogue is 'I still have it installed' or 'I used it again this morning', not a fake box. Stretching the formula onto a product you cannot physically restock makes the creator look like a liar.

A CTA that names the SKU, the size or the bundle outperforms a generic 'shop now' because it removes a decision. UGC traffic that clicks and then bounces on a PDP with six variants is a CTA problem, not a landing-page problem. The creator, who just used the product on camera, is the right person to say which version to buy. CTR is good; more important, add-to-cart on those clicks is usually better.

Be bossy about the variant, not about the person. 'Get the 2-pack' is useful; 'you need to buy this right now' is the hard sell sitting at rank 8. Name the thing in the words the PDP uses, or the click arrives on a search that returns nothing. If the Shop or site supports it, match the spoken SKU to the product tag or collection so the auction is not sending people to a homepage. If the auction lands on a collection page, the spoken SKU must still be findable in two seconds or the click was a bounce you paid for.

This formula is how you sell the offer without sounding like a late-night ad. A bundle, a 'free liner with the mascara', a size that actually fits — the close is merchandising, delivered in the same voice as the demo. Failure mode: naming a variant that is out of stock, or a bundle that only exists on one landing page the ad does not hit. Check inventory on the destination before you lock the line. When the catalogue rotates weekly, put the named SKU in a caption you can swap without a reshoot. Spoken-only next steps go stale the day the 2-pack disappears.

Callout hooks work at the open; they also work at the close. An identity-filter CTA tells the wrong people to leave and the right people to click, which is why raw CTR can look average while CPA looks excellent. It is the close version of 'if you have oily skin, stop scrolling' — same mechanic, applied when the viewer already has the proof. Write it as one breath at the end, not as a second hook fighting the first.

Sharpen the identity to a behaviour, not a demographic. 'If you cook on weeknights' beats 'if you care about your kitchen'. The behaviour should be the one the demo just showed, so the close is a logical end not a new pitch. This is also how you protect return rate in categories like apparel: 'if you're between S and M, size up — then use the link' is a CTA and a fit instruction in one breath. If the identity needs a sentence of explanation, it is not sharp enough to close with. A behaviour plus a pain in one breath is the whole craft.

The risk is shrinking the audience the algorithm will find. On prospecting that can be a feature. On retargeting, where the person already knows you, a tight identity filter can leave clicks on the table — they are in the audience because they already qualified. Use this close on cold UGC; use a softer afterthought on warm traffic. Do not stack it with a callout hook unless the two identities match, or the ad argues with itself. On broad Advantage+ this filter is doing targeting the audience stack no longer does. That is a feature on cold, and a leak on people who already added to cart — they do not need to be told they are the kind of person who buys.

Risk reversal belongs at the end, after proof, not at the start where it sounds like you expect a refund. A short, plain guarantee close lifts CTR on higher-AOV and first-time-category products because it answers the last objection while the product is still on screen. It is also one of the few CTAs that legal teams like, provided the words match the actual policy. Say the window once, plainly, then stop talking; extra reassurance sounds like you expect a return.

Quote the policy you really have. 'Send it back' on a 30-day money-back is true; the same line on a store credit-only policy is a complaint waiting in the ads-manager comments. Name the window if it is a selling point ('30 days, I checked') and skip it if the window is stingy. Show the returns URL or the unboxing of a prepaid label only if that is actually how it works — invented operational detail is worse than no guarantee. If finance hates refunds, do not run this close until the policy is one you can afford to advertise.

Do not lead with it. Opening on a refund is an admission. Closing on a refund is confidence. The rest of the video still has to do the selling; a guarantee on a weak demo just advertises that you expect disappointment. Pair with a soft next step ('link's there, try it') rather than a countdown. This formula also travels well into AI UGC, because it is a sentence, not a stunt. A useful test: same body, guarantee close versus afterthought close, on the same warm audience. If CTR rises and refund rate does not, you found a line; if refunds follow the clicks, you advertised a safety net people intended to use.

When the number is the objection, say the number, then put it next to something the viewer already buys without thinking. A reframe close outperforms a percent-off overlay because it keeps the creator's voice and does not train the auction on discounters only. CTR is strongest when the comparison is boring and local, not when it is a theatrical 'would you believe'. Keep the comparison in the same week as the purchase, not in a yearly budget the viewer does not feel.

Use a comparison from the same weekly budget: coffee, a delivery order, a refill they already buy. Category-inappropriate frames ('less than a tank of petrol' on a lip balm) feel written. Say the actual price at least once; a reframe with no number looks like you are hiding it. If you run a sale, put the sale in the spoken line as an afterthought to the reframe, not as a sticker that replaces it. On localised cuts, rebuild the comparison; a US coffee frame in a market that does not buy coffee that way is just a translated cliché.

This is a poor close when the product is premium-on-purpose. Reframing a £90 serum as 'less than two facials' can work; reframing it as 'less than a pizza' cheapens the brand you just spent 20 seconds elevating. Check the destination: if the ad says £29 and the PDP has jumped to £35, the click is a refund in slow motion. Keep currency and market straight on localised cuts. If you must overlay a sale, keep the spoken reframe as the CTA and the sticker as a caption. Two numbers that disagree — 'less than a coffee' plus '50% off £80' — make the viewer do maths instead of click.

A rhetorical question at the end hands the decision back to the viewer, which can lift CTR on already-convinced traffic and die on cold traffic that still has objections. It is short, caption-friendly and easy to test as a last-line swap on an otherwise identical video. Treat it as a polish on a working body, not as a strategy. Use it as a last-line swap on a body that already proved the product, never as the only persuasion in the file.

The question has to be earned. After a tight demo plus a guarantee, 'why would you not' is cocky in a useful way. After a vague testimonial, it is a shrug. Stronger variants name the remaining friction: 'if the only thing stopping you is the price, look at the 2-pack' is a question-plus-reframe and usually beats a naked 'so, what do you think?'. Avoid fake engagement bait ('comment YES if you want the link') in paid ads — it confuses the optimisation event and looks dated. If you need a real question, ask which size or which bundle, so the click has a job.

On Reels and TikTok the question also works as on-screen text for muted plays, which is the real reason to keep this formula in the kit. Spoken-only questions vanish for a large share of impressions. Put the same five words on screen, hold them through the platform CTA, and you have a close that works with sound off. Do not let the question run past three seconds of end-card or you are paying to linger. Retargeting is the home for this formula; cold traffic still needs a next step they can obey. A question with no destination is a shrug you paid CPMs for.

It still converts in some accounts, which is why it is on the list. A large end-card, a burned-in discount, a countdown — these are the CTAs of catalogue and Direct-response 2018, and they can win on retargeting, on offer-led products, and on placements where people expect an ad (Facebook feed more than TikTok). They also strip the UGC of the thing that made it cheap to buy: native feel. If you run it, run it as a labelled test cell, not as the default last five seconds on every UGC file in the account.

Use it when the offer is the product. A genuine 24-hour restock, a bundle that only exists this weekend, a free-gift threshold — the overlay is then information, not decoration. Do not fake the clock. Platform review and comment sections both punish fake urgency, and the auction learns you as a discounter, which is expensive to unwind. If you need the overlay, mute the spoken hard sell; two shouts at once is how UGC becomes a TV spot. Match the overlay to the live offer every time you refresh, or you will train comments to call you a liar.

Ranked last because the same video with a rank-1 or rank-3 spoken close usually keeps more of the hook-rate advantage that UGC was bought for. Test it as a variant, especially on retargeting, not as the default on cold creator footage. If the only way your offer works is a sticker, the offer may be the problem, not the CTA formula. A clean split that keeps native feel: spoken afterthought, platform button hard, overlay only on the retargeting cut. Most accounts that 'need' the sticker only needed it on warm traffic and then spread it everywhere.