Bullsift

Inspiration 👴🏻

Small businesses and creators struggle to understand what makes social media ads effective versus misleading or low-quality. After seeing friends' small businesses waste money on poorly targeted campaigns, we were inspired to create a tool that helps users critically evaluate ad content quality and learn what drives successful advertising on platforms like FaceBook and TikTok.

What it does 🚨

Bullsift acts as a "quality scanner" for social media advertising content. It assesses the credibility and effectiveness of ad creative, alerts users if content uses misleading tactics, and offers an interactive game to explain what makes ads successful or unsuccessful, providing users with insights into advertising best practices. With this tool, even aspiring marketers can evaluate ad quality like seasoned growth product managers!

How we built it 🔧

We developed a Chrome extension that seamlessly integrates into your browsing experience, offering real-time ad content analysis for quality and credibility assessment. Powered by Cohere's Command R Model, our multi-layer detection system ensures users are equipped with tools to evaluate the effectiveness of advertising content. Each layer of detection is carefully engineered to deliver comprehensive insights, ensuring a smooth and intuitive user experience. The layers include:

Engagement Signal Detection: Used to identify persuasive language and hooks within ad content Brand Credibility Verification: Cross-referencing advertiser sources to evaluate trustworthiness Campaign Pattern Analysis: Comparing ad tactics to successful campaign strategies from trusted brands Quality Scoring System: Providing a detailed "effectiveness score" based on messaging clarity, targeting signals, and creative quality

Challenges we ran into 🏁

Defining and grading ad effectiveness was complex, requiring accurate detection of persuasive techniques and audience alignment Integrating multiple API calls synchronously for the multi-layer detection system was technically challenging Making the experience fun and engaging, especially for aspiring marketers and SMB owners

Accomplishments that we're proud of 🌟

Developing a highly accurate detection system that evaluates ad content quality in real-time using Cohere's powerful AI Creating a fun and engaging learning experience that encourages critical thinking about advertising effectiveness and empowers users to identify high-quality campaigns Designing a user-friendly, interactive, and aesthetically pleasing UI/UX that simplifies complex ad analysis processes

What we learned 📖

Using Cohere's LLM models in different ways to achieve the ad quality assessment system Grinding with 2 hours of sleep!

What's next for Bullsift 📈

Our next step is to implement a feature that tracks campaign performance metrics across multiple social media platforms, helping businesses and creators understand which ad strategies drive the highest conversion rates and optimize their advertising spend.

Built With

• cohere
• css
• express.js
• html
• javascript
• llms
• node.js
• postman
• rag