A jewellery brand owner used AI tools Pixverse, Invideo, and Adtest to scale video production from 4 to 30 videos monthly, leading to a 10% sales increase through enhanced creative testing.
Owner of a jewellery brand here. Decent sized, but we were hitting a bottleneck in creative production. We had quite a big product catalogue, but producing video for every product was just too expensive. At the beginning we were making maybe 4 product videos a month because that was roughly what our small production setup could handle. The problem was that we had way more products than that and a lot of them were basically just sitting there with product photos. Started experimenting with an AI workflow. Here are all the tools I used to convert product photos into finished, final ads. Pixverse Used this to generate standalone, hero shots in extreme detail. Uploaded the product photos and reworked the backdrop, with a custom camera movement. The shots were visually stunning and the products had their characteristics intact. Invideo became useful for the production/assembly side. Jewellery is where intricate details matter a lot: color, lightning, exact product shots on the models wearing them from diff angles. Took the hero shots and started shaping the whole ad in it, experimenting with the messaging and different angles we wanted to test. Variations are much easier to make as the tool remembered everything about the product. Referencing some hooks from ads that performed well really helped. Spun iterations of each to have a variety of ads. Adtest Took the final outputs and fed them to ad test. Defined the audience for each variations and then it basically gave a score on which ad is likely to perform better. Ended up alloting more budget to those. The result was going from roughly 4 videos/month to around 30. Having 30 videos meant more data, which would lead us to testing way more hooks, products and creative angles instead of putting most of the budget behind a handful of creatives. Over the period we were tracking, sales were up roughly 10% as well. I obviously wouldn't attribute that entirely to AI/video because there were other things changing in the account, but the increased creative testing definitely helped. Would be interested to hear how other ecommerce brands are handling this. Are you using AI mainly for generating the actual footage, or more for speeding up the production/editing pipeline?
The author built a custom AI video editor using models like Gemini and GPT to automate video editing for SEO and link building content, significantly boosting productivity but incurring high costs.
The article discusses the challenges of AI video generation, emphasizing that while tools like PixVerse can produce clips, creating a full video with a narrative still requires significant human effort and creativity.
Invideo AI, an India-based startup, launches a multi-agent video creation platform built on OpenAI models (GPT-4.1, o3, gpt-image-1, text-to-speech) that enables users to generate professional-quality videos 10x faster from natural language prompts. The system uses specialized AI agents for planning, scripting, research, content moderation, visual generation, and narration, now serving over 50 million users creating 7 million videos monthly.
A solo creator used AI tools to build a photorealistic luxury AI influencer in 30 days from scratch with zero budget, landing a real brand deal. Key lessons include the importance of face consistency and the surprising speed of brand deals.