Spend

What the engine costs versus what it finds — day by day. One shared scrape feeds every thesis, so cost, coverage, and usage are the same regardless of which thesis you're viewing.