GuideThread
Back to implementation guides

Implementation guide

Ecommerce AI Creative Production Pipeline

Turn product evidence and brand context into a repeatable AI-assisted workflow for producing and reviewing ecommerce ad creative.

3 ordered steps

  1. Step 1

    Evidence-led creative direction

    Research

    Claude Listing Creative Workflow

    Created by Raphael Guilhem

    Use this to turn a product packshot, product-page context, features, benefits, and differentiators into Amazon ad concepts with visual hooks, messaging approaches, and consistent styles. Carry the selected direction into Step 2.

    How to use it

    1. Add the product packshot, product URL, and verified brand context.
    2. Review generated features, benefits, and differentiators for accuracy.
    3. Select one concept, hook, message, and visual style for Step 2.

    Open resource View original source post

  2. Step 2

    UGC video prototype

    Production

    MakeUGC Ad Factory Workflow

    Created by Stijn Feijen

    Use the Step 1 direction to script, voice, edit, and review a UGC-style video prototype in MakeUGC. MakeUGC execution is paid after trial access; treat performance claims as unverified. Carry approved script and assets into Step 3.

    How to use it

    1. Translate the Step 1 concept and message into a UGC script.
    2. Generate the voice, presenter, and first video cut in MakeUGC.
    3. Review claims, likenesses, disclosures, and the final cut before Step 3.

    Open resource

  3. Step 3

    Creative variant production and review

    Execution

    Vibelets MCP Workflow for Meta Ad Creatives

    Created by Nipun Taneja

    Use the approved direction and prototype assets to plan and generate additional Meta-ad variants through Claude and Vibelets. Generation uses Vibelets credits; review every asset before any separate campaign launch.

    How to use it

    1. Connect Vibelets to Claude and load only authorized reference assets.
    2. Generate variants from the approved hook, message, and visual direction.
    3. Review claims, likenesses, brand fit, and disclosure needs before launch.

    Open resource View original source post