Neilson Financial
My specific role and responsibilities
AI Prompt Generator, Paid Social Media Designer, Scripting Q/A.
Who else was involved
Worked closely with a Paid Social Production Director and Content Manager.
- Worked with CSV data, dynamic templates, and content automation to scale personalization across multiple compliance regions (Canada, Australia, UK).
- Produced 15,000+ creative assets monthly at scale, optimizing for platform-specific formats and performance.
- Leveraged performance data to inform creative iterations, drive engagement, and surpass monthly KPIs by 300%.
- Leveraged GenAI tools (ChatGPT, Ideogram, MidJourney, Generative Fill) to create authentic UGC-style social media assets, streamlining production workflows and significantly increasing output efficiency.
Prompt:
- a woman aged 45. standing in a professional pose, holding a white sign board above her head with both hands. The background is a blurred street of London. A big street billboard at the back, The overall mood is professional, yet with a hint of uncertainty
Weekly Objectives:
- Produce creative that resonated with audiences in a fast-paced dynamic environment, targeting by age group, location and gender.
- KPIs were set at ~800 assets a week, ~3,200 a month. Focus on AI asset generation.
Key Highlights:
- I blew my targets out of the water, surpassing my KPIs by over 300%, consistently hitting close to 15,000+ a month.
- Created both UGC-style and non-branded content.
Top Performing Social Ads
The creatives I developed and iterated were some of the top performing and had the lowest CPAs and cost per clicks, generating high OKRs.
Using AI organize my work
I developed an AppleScript that streamlined the management of high-volume asset production, improving productivity and organization across the workflow for myself and my teammates. It renamed files then sorted them by their role, age and orientation.
Developing assets at scale
I assisted in QA for a custom Adobe After Effects plugin that dynamically rendered static and motion assets based on CSV data input.
We implemented that further with a custom Figma plugin. Rendering thousands of dynamic assets at a time.