Hey @gregisenberg, I asked Hannah to validate the ad attribution idea from @ideabrowser. here's what she found. (+ some screenshots) the market: $24B in annual influencer spend globally. 71% of it - roughly $17B - runs without effective ROI measurement. the attribution software market on top of that is $4.74B growing to $10.1B by 2030. only 6% of marketers use any form of multi-touch attribution for influencer campaigns. MrBeast's own words: "my videos don't make money. even when i do a brand deal, i still lose money." he charges $390K-$3M per deal depending on integration type. brands pay it anyway. nobody can tell you if it worked. the competitive landscape: tools like Tagger (Sprout Social), Captiv8, and Klear (Meltwater) come closest to pre-deal prediction. all of them use the same approach - creator's historical engagement rate extrapolated forward. none of them run simulation. none have cross-client benchmark data. they tell you how big and real an audience is. they cannot tell you what ROAS your specific product will get from that specific audience. Hannah's take on the moat: it's not the simulation engine. the engine is buildable. the moat is proprietary cross-client campaign truth data - what did brand X pay creator type Y, and what did they actually get? no single brand can build that alone. the simulation only becomes defensible when it's trained on real outcomes across hundreds of campaigns. whoever builds that data flywheel first wins. the model gets smarter with every deal that runs. the agency angle is the most underappreciated threat: agencies manage 15-30% of all influencer spend. if brands can simulate ROI before signing, they need less of the agency's "we know from experience" value prop. agencies will either buy the tool or fight it. best go-to-market according to Hannah: mid-tier DTC brands doing $500K-$5M annually in influencer spend, and influencer marketing agencies. both have the pain, neither has internal data science, and agencies multiply the value across every client they run.