What do firms do when they choose NOT to innovate with IT when they face performance problems and why? Though counterintuitive, we suggest that studying this question enables us to: explore alternatives to the dominant rational maximizer logic prevalent in innovation research; overcome the pro-innovation bias critique; and more carefully study the very earliest time period of the innovation process. We answer this question through an exploratory, qualitative study of what firms do when they encounter problems (specifically performance gaps) and we pay special attention to what they do when they elect not to adopt new information technology solutions to solve them. Through systematic analysis of our qualitative data set and application of a problemistic search/sensemaking lens, we discover that common actions include tinkering, stretching, making do and going back to prior IT. Because firms wrestle with complex environments with multiple, weak and ambiguous cues which hinder the identification of performance problems, and feel pressed to economize on scarce resources, they develop smaller and more familiar solutions rather than attempting unfamiliar and more resource-demanding innovations. When these solutions work, they become long-term alternatives to IT innovation. We conclude with a discussion of how our approach succeeds in going beyond the dominant paradigm, informs an agnostic stance towards innovation and promotes a call for a new frontier of research into the pre-innovation phase of IT innovation research.