Make's If-Else & Merge Modules: Build a Zero-Credit Data Normalizer

In this hands-on tutorial, Volker walks you through building a reusable, zero-credit Data Normalizer from scratch using Make's new If/Else and Merge modules. If you've ever had to wrangle customer data from multiple platforms into one consistent format, this is the blueprint you've been waiting for. 🔍 Inside the Build: * The Normalizer Concept: Understand why a normalizer is essential when pulling data from multiple sources (CSV exports, HubSpot, Stripe) that each return fields in completely different structures — and how to unify them into a single, predictable output. * If/Else Module as a Router: See how the new If/Else module replaces the need for duplicating logic across router branches, with each path handling the unique field mapping for its source (CSV, HubSpot, Stripe) before converging cleanly. * The Merge Module Magic: Learn how the Merge module collects all three paths into one unified output (first name, last name, email, ID) — eliminating the old workaround of setting variables on each branch and reloading them. * Zero-Credit Architecture: The entire normalization layer runs without consuming any Make operations — the If/Else and Merge modules are free to use, making this pattern extremely cost-efficient at scale. * **Bonus — Cross-Platform Lookup Table:** A practical extension of the build: using the normalized output to maintain a live lookup table that maps HubSpot IDs, Stripe IDs, and emails to the same customer record across tools. This video is an ideal reference for anyone building multi-source data pipelines who wants a clean, maintainable, and scalable pattern without reinventing the wheel every time a new integration is added. 🛠️ Featured Stack: * If/Else Module (Conditional Routing) * Merge Module (Path Consolidation) * Variables & Data Store (Lookup Table) * HubSpot (Mock Contact Data Source) * Stripe (Mock Customer Data Source) * Custom CSV (Flat File Data Source) 💡 Pro-Tip from Volker: "Just throw in the whole bundle." Rather than mapping individual fields as inputs to the normalizer scenario, passing the entire data bundle keeps the call generic and future-proof — if you need to add fields later, you only update the normalizer itself, not every scenario that calls it.