AI advisors modelled on practitioners
@andrewfoxwell
AI interpretation, not the person. This advisor is modelled on Andrew Foxwell's publicly available work and documented frameworks. It is not Andrew Foxwell. It does not speak as Andrew Foxwell, it does not fabricate quotes attributed to Andrew Foxwell, and it will not reproduce copyrighted material. Every response cites the public source it draws from.
Paid social as a testing discipline: structure campaigns to learn, then scale what works
This is an AI interpretation of Andrew Foxwell's public work. Andrew Foxwell is a paid social specialist and co-founder of Foxwell Digital. His publicly documented work focuses on Meta and paid social advertising strategy, creative testing, and campaign architecture. He is known for systematic approaches to paid social that prioritise learning and scaling signal over spending.
Foxwell Digital's creative testing framework isolates variables across three creative dimensions: (1) Hook: first 2-3 seconds or headline, (2) Body: the core value proposition delivery format (UGC, talking head, text overlay, static), (3) CTA: offer and action. Testing all three simultaneously creates noise; the matrix approach tests one dimension at a time with controlled budgets, then combines winning elements. This surfaces the creative driver of performance rather than attributing results to the overall ad.
A summary of Foxwell Digital's published position on ROAS as a primary metric: ROAS measures the ratio of reported revenue to ad spend, but reported revenue is distorted by attribution windows, cross-device journeys, and organic activity attributed to the ad. At scale, ROAS targets lead to spend on the easiest-to-convert users rather than genuine new customer acquisition. CAC (customer acquisition cost) for genuinely new customers is a more reliable north star, even though it is harder to measure.
In domain
When a question falls within paid-social, paid-ads, facebook-ads and related areas, this advisor answers grounded in Andrew Foxwell's documented frameworks, with citations to the public source.
Out of domain
Out of domain, this advisor explicitly flags low confidence and defers rather than guessing. It never answers authoritatively outside its declared specialisation.
Example
“How do I set up a measurement plan and track micro conversions?”
This question is about analytics which is outside this advisor's declared domains.
Capabilities
2Schedule
Source scraping
onWeb research
onAdd to knowledge