Guide
The same video rarely does both jobs well. A cold audience needs to be introduced; a warm audience needs to be answered. Ship one file to both and you underpay one of them.
A cold viewer is a stranger. They have never heard of you, they owe you nothing, and they will decide whether to keep watching in the time it takes to recognize your video as an ad. Everything that matters for cold creative happens in the opening: a hook that stops the scroll and a fast answer to "what is this and why should I care." Cold is where the hook rate is decided, because cold is the only audience that hasn't already agreed to listen.
A warm viewer is not a stranger. They've seen the product, maybe visited the page, maybe added to cart — and then didn't buy. They don't need to be told what it is. They need the one thing that closes the gap between "interested" and "bought": the objection handled, the proof shown, the reason to act now. Warm creative that re-introduces the product is answering a question they stopped asking days ago.
The most common mistake isn't bad creative — it's good creative pointed at the wrong temperature. A cold ad that's crushing it spends its whole runtime on introduction, and that's exactly why it can sag on a warm audience: the warm viewer already watched the intro, already understood the offer, and swiped past the buy anyway. Playing them the same reel is repetition, not persuasion. Nothing new entered the conversation, so nothing changes.
It runs the other way too. A polished warm ad that opens on "Still thinking about it?" assumes a relationship the cold viewer doesn't have — so it lands as noise, because it's answering an objection the stranger never raised. Both ads are fine. Both are mis-cast. The temperature of the audience is the casting decision, and it's the one most easily skipped when you duplicate an ad set and change the targeting but not the film.
Cold creative front-loads the hook and the what-is-this, keeps the claim broad enough to catch people who didn't know they had the problem, and treats every second of introduction as earned rather than assumed. Warm creative can open in the middle of the story — skip the setup, name the hesitation out loud, and spend the runtime on the thing the cold ad had no room for: the review, the guarantee, the side-by-side, the "here's what you were worried about." The two files often share footage and share nothing about structure.
This is also the cleanest way to cut one shoot into multiple ads: one raw session, two edits with two different jobs. Retargeting is the warmest end of this spectrum and deserves its own treatment — retargeting ad creative is where the "handle the specific objection" logic gets most literal, because the audience is small, expensive, and defined by a purchase they almost made.
If you only pre-screen one version, screen the cold one. Cold is where attention is scarcest — you're fighting for a first glance from people with no reason to give it, so the opening either works or the whole spend is wasted on impressions nobody watched. Warm audiences forgive a slower start because the viewer already has a reason to care; cold audiences forgive nothing. A predicted attention read tells you whether the cold cut clears that first gate before you pay to find out live.
Just be honest about what the screen sees. It reads whether a cut holds the watch, not whether the message matches the audience's state of mind. A warm ad can score beautifully on attention and still miss the objection that actually stopped the sale — the score can't know what your warm viewer was worried about. Use the screen to protect the cold opening and to compare cuts; live results still decide whether the warm message closes.
Upload your cold-audience ad and see whether the opening holds against a benchmark of top-performing ads — up to 8 free credits.
Score Your AdUsually not, because the two audiences arrive in different states. A cold viewer has never heard of you and owes you nothing — the creative has to earn the first second and explain what this even is. A warm viewer already knows the product and stopped short of buying, so the creative can skip the setup and answer the specific hesitation that held them back. Running one file against both wastes the cold impression on context the warm viewer doesn't need, or wastes the warm impression re-introducing a product they already recognize.
A strong cold ad spends its energy on introduction — the hook, the problem, the what-is-this. A warm viewer has already sat through that once and swiped past the buy, so replaying the intro tells them nothing new and gives them nothing to reconsider. The cold ad isn't bad; it's answering a question the warm viewer stopped asking. Warm creative has to move past awareness to the objection, the proof, or the offer.
Upload each cut and PreTestAds predicts how attention holds across the runtime and scores it against a benchmark of top-performing short-form ads. That read is most useful on the cold version, where winning the first three seconds against strangers is the whole game. It predicts attention, not intent — it can tell you whether a cut holds the watch, but not whether a warm viewer's specific objection gets answered. Attention is the shared gate; the messaging fit is still your call.