Metrify | Devpost

Inspiration

E-commerce merchants struggle to optimize product pricing manually, juggling inventory levels, sales velocity, and cart abandonment rates without clear data-driven insights. We built Metrify to solve this by creating an intelligent pricing engine that connects to Shopify stores and automates pricing decisions based on real-time analytics.

What it does

Metrify is a Shopify analytics and dynamic pricing platform. It ingests product, inventory, and sales data from Shopify stores, then calculates key metrics like total sales, inventory levels, and abandoned cart ratios. Merchants control pricing strategy through four intuitive sliders: inventory impact, sales velocity, abandonment sensitivity, and pricing aggressiveness. The system automatically calculates optimal price adjustments and syncs them back to Shopify in real-time while maintaining complete price history.

How we built it

We built the backend with Node.js, TypeScript, and Express, using MongoDB Atlas for data storage. Our custom Shopify GraphQL client handles pagination and rate limiting. The architecture flows cleanly: ingestion fetches data from Shopify, the processing engine uses weighted scoring to calculate price adjustments, and GraphQL mutations update prices in the store. We used TypeScript, Express, Mongoose, Axios, and Shopify Admin API 2024-10.

Challenges we ran into

We struggled with Shopify API authentication, learning the difference between token types. GraphQL pagination initially used string replacement which failed, so we refactored to properly pass variables. We had to ruthlessly minimize our schema from bloated legacy models. Implementing smart queuing and throttling helped us avoid Shopify's strict rate limits. Balancing when to update MongoDB versus syncing to Shopify while handling failures gracefully was complex.

Accomplishments that we're proud of

We built a production-ready system that connects to real Shopify stores and ingests live data. We successfully implemented cursor-based pagination for thousands of products and created an elegant slider-based algorithm that gives merchants full control. The system has proper error handling, logging, and retry logic throughout, with atomic updates to both MongoDB and Shopify.

What we learned

We learned Shopify's Admin API, GraphQL queries, and authentication system. We discovered the importance of pagination and rate limiting with large datasets, how to design minimal schemas, and weighted scoring algorithms. TypeScript proved invaluable for type safety in complex pipelines. Most importantly, production-ready code needs proper logging, error handling, and retry logic, not just happy path coding.

What's next for Metrify

We plan to integrate machine learning for predictive pricing strategies, add A/B testing for different pricing approaches, and incorporate competitor pricing data. We want to support multiple stores from one dashboard, build an advanced analytics dashboard, and add automated scheduling for sales events. Profit margin optimization using cost of goods sold and packaging as an official Shopify App Store listing are also on our roadmap.