Goose Ads Remixer

Remix the ads already winning in your niche

Marketing

Artificial Intelligence

GitHub

Video

Visit WebsiteSee on Product HuntGithub ⧉Twitter ⧉

congrats on the launch! love that it starts from ads already working instead of a blank prompt, feels way smarter than the usual generic tools. quick q, is it meta only for now or are google/tiktok formats coming later? anyway wishing you guys a great launch 🙌

Really amazing to use, made creating ads for a baking business a way easier when it was proving to be a bottleneck. The concept of not creating ads from zero but instead using already proven to win ads to make more is simply brilliant.

"Remix the ads already winning" raises a straightforward question about intellectual property, if Goose is extracting hooks, copy structures, and layouts from competitor ads and generating new creative that closely follows those patterns, where does inspiration end and reproduction begin? Has there been any legal review of the remixing methodology, especially for ads that use distinctive creative approaches that could be considered trade dress?

@himanshu_bamoria1 Like the idea of learning from winning ads instead of starting from scratch. Upwards and onwards 🚀

Congratulations on the launch! Do you support creating video ads, for focusing on image ads for now?

Congrats on the launch. One thing I'm curious about, the Meta Ad Library shows you which ads are running and for how long, but not spend or performance. So how do you decide which ones are actually winning vs just live?

when Goose learns the pattern from a winning ad in my niche, where's the line between remixing the structure and just recreating that competitor's ad with my logo swapped in - is that boundary a manual review step or does the model handle it?

the nano banana / gpt-image-2 / opus checking-each-other's-work loop is the interesting bit - when the three disagree on whether an output passes the brand check, is there a tiebreaker or does it just regenerate until they converge?

I think if everybody is learning from the same winning ads then the meta feeds would really look similar over time.

the brand-drift question above is the UX risk, but there's a legal one underneath it too: "remixing patterns" is fine, but if the tool is looking at a specific competitor's ad to generate from, how do you make sure it's not echoing their actual copy phrasing rather than just the structure? a hook and a layout aren't protectable, but a distinctive line of ad copy can edge into IP territory if it comes out too close to verbatim

Just signed up for Goose Ads Remixer and tried it for my website ToolJunction. I’ve used a few AI ad creative tools before, but goose feels more polished and practical at this price point.

Starting with ads that are already working in your niche, instead of generating from a blank prompt, makes a noticeable difference. And, the outputs felt much closer to something I could actually run, and the overall experience was smooth.

The only issue I noticed was that it picked the wrong color palette for my website. Great work, team!

Most AI ad tools generate noise.

This actually helps teams iterate on what's already working. Thats a much stronger approach.

Congrats on the launch

Specifically to the presentation, I really like the UX of the geese. How do you make sure the new ads stay unique?

This is cool. I've used Claude and Remotion some for making little videos, more product explainers and ads. I think one of the valuable things here is that you have a library of ads that are already performing well in the user's niche. I definitely think, by default, AI is not good at copywriting, so having this verified high-quality content that steers the AI seems really useful.

Smart approach — remixing proven winners instead of generating from a blank prompt sidesteps a lot of the "AI slop" problem. Are you pulling structural/pacing patterns from the source ads, or more the visual style? Curious because that distinction has been a big design decision for us on the video-generation side too.

Congrats on the launch! I tested it out, very smooth! Does Goose learn from actual campaign performance over time, or is it mainly learning from the existing ad library? Feels like closing that feedback loop could become a huge advantage.

Congrats on the launch! The idea looks valuable — does it search for ads in a niche automatically? Basically in 1 prompt telling you the winning format