AB Ship Lab

This week, we continued expanding SewBuddy’s core systems. Our database has grown significantly, and the product architecture is becoming much more sophisticated as we build the foundation for long-term scale. Our acquisition campaigns continue to show positive signals, and we’ve significantly improved our data tracking. We now monitor every scan in real time—from the input itself to whether the result was successfully generated—giving us much better visibility into product performance and user behavior. We’re increasingly confident in the product’s potential. We continue to deepen its capabilities by making more effective use of AI, structured knowledge, and specialized skill sets, allowing the product to become smarter and more capable over time. We see this as just the beginning. The deeper we go with our AI systems, content, and data, the stronger the product becomes. Early signals continue to reinforce our belief that investing in depth will lead to better long-term results. Our focus remains the same: keep strengthening SewBuddy every week and turn it into a product that’s increasingly difficult to replicate.

Building More Depth Into SewBuddy This week reinforced our confidence in SewBuddy. The marketing results validated the direction we’re taking, so we’re continuing to invest in the product and focus on adding significantly more depth. Our priority is expanding the experience with richer categories, trends, sewing patterns, and a much larger content database, while introducing features that improve retention and increase renewable purchases. It’s a substantial effort, but AI is allowing us to build and scale this foundation far more efficiently. We believe SewBuddy has reached a level where it stands among the strongest products in its category. From here, our goal is to build enough depth, proprietary content, and AI capabilities that it becomes exceptionally difficult to replicate. We’ll stay fully focused on SewBuddy for the rest of the tournament and keep raising the bar every week.

This week, we released several new SewBuddy versions with minor bug fixes and product improvements. On the creative side, we focused on developing stronger hooks for the first few seconds of our videos. While Scate AI has been producing a large volume of creatives, we concentrated on creating the first 5 seconds of videos tailored to the audience we have been learning about since we started development. On the Gemstone side, we expect to have clearer results next week. While working on SewBuddy creatives, we have also been building an internal AI agent that analyzes successful creatives and popular social media concepts, learns them as skills, and then helps generate new video ideas and assets more efficiently. Both SewBuddy and Gemstone are serving as the first projects where we are applying and refining this approach.

This week, we focused on bringing SewBuddy to a point where we feel confident in both the product experience and its monetization potential. To support this, we continued investing in the AI engineering side of the product, improving both low-cost pattern generation and visual sewing guides. The experience is already valuable today, and we see significant room for further improvements over time. On the Gemstone side, we launched a new batch of creative tests. We are excited to see the results and learn whether the new concepts can improve performance compared to our previous experiments.

This week we focused primarily on SewBuddy. SewBuddy is now live on the App Store, and we have officially released Version 1.0. Users can take a photo of a garment, fabric, or sewing project and receive AI-generated sewing guidance, pattern suggestions, and step-by-step recommendations. We believe we reached the scope aligned with our mentors and the product is now ready for purchase campaign testing. We're excited to start gathering real user data. On the Gemstone side, we focused on concept exploration and reviewing previous creative test results. Our goal is to launch a new test next week and share early purchase campaign learnings soon. Stay tuned — we hope to have meaningful results to share for both products over the coming weeks.

- We have submitted our new sewing app for review and are currently waiting for approval. Our goal is to have it ready for CPI testing this week. We have also finalized the update roadmap and will share it once the app is approved. - On the Rock app side, after testing alongside Skate, we continued running creative tests with our own concepts and achieved lower CPI results. This allowed us to explore angles we had not previously tested. - One of our main focuses was the "Real or Fake?" concept, as we noticed that many of the top-performing creatives in this category were built around this theme. We tested multiple variations and identified a strong-performing format. - Our plan is to double down on this winning angle and launch a new round of creative testing. The results have helped us narrow our focus, and we believe there is still room to improve both CPI and conversion performance. - We will continue evaluating the impact of our latest eMPC/CPI tests and will share further results and learnings over the coming weeks.

We have submitted our new sewing app for review and are currently waiting for approval. Our goal is to have it ready for CPI testing this week. We have also finalized the update roadmap. On the Gem ID side, after testing alongside Scate, we continued running creative tests with our own concepts and achieved lower CPI results. This allowed us to explore angles we had not previously tested. One of our main focuses was the "Real or Fake?" concept, as we noticed that many of the top-performing creatives was in this category. We tested multiple variations. Actually let me show you some examples now. Our plan is to double down on this winning angle and launch a new round of creative testing

So far: • We are no longer continuing with the previous application in a focused way. • We finalized the concept of our new app and are currently in the development phase. • Our main goal is to make it ready for CPI testing as soon as possible. • The app is again focused on the hobby space. • There are very few apps in this category, and we aim to differentiate ourselves with deeper features and experiences that existing apps don’t offer. • We can share more details about the concept next week as we further refine it. • We also launched some narrative tests for our Rock app to validate a few ideas we wanted to experiment with. We’ll share the results in the coming weeks once we have clearer data.

So far: • We are no longer continuing with the previous application in a focused way. • We finalized the concept of our new app and are currently in the development phase. • Our main goal is to make it ready for CPI testing as soon as possible. • The app is again focused on the hobby space. • There are very few apps in this category, and we aim to differentiate ourselves with deeper features and experiences that existing apps don’t offer. • We can share more details about the concept next week as we further refine it. • We also launched some narrative tests for our Rock app to validate a few ideas we wanted to experiment with. We’ll share the results in the coming weeks once we have clearer data.

This week, we have a bit of difficult news to share. Over the past period, we had been doing deep research around the gemstone space and continuing to build the product further. However, after testing the concept together with our mentors and Scate, we saw that the CPI evolution was too high, and we started to feel that this idea did not have a strong enough market to justify going deeper. Because of that, we decided to park the Gemstone idea for now. The good news is that we already have multiple new directions in mind. We will be narrowing them down this week and immediately starting development on the one we believe in most. So while this is the end of one direction, it is also the beginning of a new and exciting one.

we already have multiple new directions in mind. We will be narrowing them down this week and immediately starting development on the one we believe in most. So while this is the end of one direction, it is also the beginning of a new and exciting one.

This week we started preparing for creative CPI testing after aligning on a clearer product direction together with our mentors. One thing became very clear during this process: we don’t want to build “just another scanner app.” We want to understand what people actually care about in the gemstone world and build around those behaviors and emotions. So recently, we’ve been deeply researching the audience itself — watching long-form YouTube videos, analyzing highly engaging TikToks, exploring Reddit communities, and trying to understand what truly creates interest and connection. Is it the excitement of finding stones in real locations? The polishing and shaping process? Daily usage? The emotional meaning behind gemstones? Or the collecting journey itself? Our goal is to identify real user interests and pain points, then turn those insights into meaningful features and experiences inside the app. Still early, but excited about the direction we’re heading.

This week we started preparing for creative CPI testing after aligning on a clearer product direction together with our mentors. One thing became very clear during this process: we don’t want to build “just another scanner app.” We want to understand what people actually care about in the gemstone world and build around those behaviors and emotions. So recently, we’ve been deeply researching the audience itself — watching long-form YouTube videos, analyzing highly engaging TikToks, exploring Reddit communities, and trying to understand what truly creates interest and connection. Is it the excitement of finding stones in real locations? The polishing and shaping process? Daily usage? The emotional meaning behind gemstones? Or the collecting journey itself? Our goal is to identify real user interests and pain points, then turn those insights into meaningful features and experiences inside the app. Still early, but excited about the direction we’re heading.

We kicked off this testing phase by using our onboarding and welcome bonus, jumping straight into the core experience with a real use case. The stone you see here is something we genuinely found outside—we didn’t place it or stage anything—so this is as close as it gets to a real user moment. From there, we scanned it to understand what it is, explore its properties, and see how it connects to us on a more personal level, like zodiac compatibility or potential use cases. At the same time, this wasn’t just about the scan itself. We’re actively testing the latest version of the product, which we refined after multiple discussions with our mentors. We received a lot of valuable feedback around experience, clarity, and depth, and this version reflects those improvements. So this moment is both a product test and a validation step for the direction we’re taking. It’s been a really exciting process so far. This is just the beginning—we’ll continue iterating, improving the experience, and building a product that people can come back to every day.

We tested our latest update by scanning a real stone we found outside—exploring its properties, value, and zodiac connection. This is just the beginning.