Hard to overstate how revelatory this paper was for me — one takeaway I had from it was "small fixed costs can really deter people from optimizing." Which made me think about a few other recent papers where fixed optimization costs are a key ingredient: