Launched this week
Wisry
Clone the ads already winning in your market, at scale
536 followers
Clone the ads already winning in your market, at scale
536 followers
AI agents that run the whole play: scan the Meta and TikTok ad libraries for the ads already winning in your market, clone them into your brand as static and video ads, and launch to Meta and Google, optimized for ROAS. Winning ads, without the guesswork.
Interactive
Payment Required
Launch Team / Built With
This is an app I've been watching very closely since I have tried four separate apps for creating "AI ads" this year and I canceled them all within one month of using each since the excitement wore off. It looks like the key factor in this case is being able to start with an existing ad rather than with a blank canvas. It'll depend on how good their ROAS optimisation is in practice.
@bl560 Joy, that's the exact pattern we're trying to break. Four tools cancelled in a month usually means the novelty was the product. Starting from a proven ad instead of a blank canvas is our answer to that.
On ROAS: I won't promise a number here, and I'd be wary of anyone who does. What we do is build cuts that can be measured (one claim per scene, hooks as variants, not one guess), launch them to your ad account, and read the results back into the next round. Whether that beats what you have is something you should see on your own account, not take from me. Happy to set you up to test it properly.
How does Wisry deal with seasonal goods? Is there an option for the agent to choose from competitor's ads based on season changes?
@len5 Partly by default, partly by instruction. By default the research reads what's live right now, ranked by how long it's been running and how many variants are behind it. So in October it's finding the Q4 gifting ads, not last summer's. That's the seasonal signal built in.
On top of that, you can brief the agent in plain text before it researches: "find ads for the holiday gifting angle" or "summer launch, skip anything winter", and it filters on that. There's no dedicated season toggle yet, but the instruction does the job. What's the product? Some seasonal categories have a much deeper ad library than others.
When two brands copy the same ad, how dissimilar will the two outputs appear in comparison to each other?
@heovl Quite different, because everything visible comes from each brand, not from the source. Different products, palettes, type, footage or avatar, and a script written for each brand's buyer. What they'd share is the skeleton: same hook type, similar pacing, proof in the same place. Watching them side by side you'd notice the rhythm matches. Scrolling past them in a feed you wouldn't. And if you want more distance, you can tell the agent to change the hook or the format before you launch.
Does the system adjust the creative when the original winning ad targets a very different customer segment?
@amybradley Good question. Yes, partly by design and partly by you.
Wisry starts from your store, not the source ad. It reads your products and who you sell to first, and that brand profile is what the script, the avatar and the visuals get built against. So what carries over from the source ad is the structure (hook, pacing, proof, close), not the audience or the message.
That said, if you pick a source ad aimed at a totally different buyer, the skeleton might just not fit. That's why the research step ranks ads from your own market first. And if something's off, you don't re-edit by hand. You tell the agent in chat what to change, say "make this speak to first-time buyers, not repeat customers", and it rewrites that part while keeping everything else consistent: the script, the scene timing, the captions, the product claims. Is there a specific case you're thinking of?
Congratulations on the launch @ruta_jaskunaite But how does Wisry avoid recreating something close enough to trigger a copyright or trademark flag from Meta or Google, especially on video ads with recognizable footage or music?
@sidraarifali Thanks Sidra! Honestly, that's up to the brand, and the tool works both ways.
If it's your own ad and you want variations, you can keep it very close: same footage, same look, and just swap the hooks or the text. That's the fastest way to fight creative fatigue.
If it's someone else's ad, you tell the agent to adapt it to your product, and it regenerates everything: the footage, the characters, the music, the script. What stays is the structure (hook, pacing, proof, close), which isn't something copyright or trademark covers. You can also tell it explicitly to steer away from the original's music or style if you want more distance.
So the platforms have nothing to match against, because nothing from the source is in the file. Have you run into flags on generated video before? Curious what tripped it.
How much does "cloning" actually modify versus copying, in order for it to retain its originality without copyright infringement issues?
BTW, Congratulations @ruta_jaskunaite @kasrav & team @Wisry 🚀✌️
@kasrav @aymi_malik Thanks Muhammad! It's your call, and it depends on whose ad it is.
If it's your own ad, you can keep it nearly identical and just swap hooks or text. That's you making variations of your own work. If it's someone else's, the agent regenerates everything: footage, characters, music, script, product shots. What's kept is the structure (hook, pacing, proof, close), and structure isn't something copyright covers. You can also tell it to steer further away, like a different style or music, if you want more distance.
So it's not "how much do we modify a copy". Nothing from the source is in the file to begin with.
PaymentKit
Also, the Pinterest and YouTube growth strategy is one that I agree with as well. I would start off with YouTube, since that is where my clients' money is spent the most.
@hamza_aafzal Noted, Hamza. That's two votes for YouTube first today. We already launch to Google, so the gap is on the research side, reading what's winning on YouTube the way we do for Meta and TikTok. Useful to know that's where your clients' budget sits.