The versatility of digital technologies relies on a capacity to represent and subsequently manipulate algorithmically selected physical processes, objects or qualities in a domain. Organizationally real digital representations are those that, beyond the mere capacity to, actually get woven into everyday work practices. Empirically, we draw on a four-year case study of offshore oil and gas production. Our case provides a vivid illustration of Internet of Things (IoT) based visualizations and data driven predictions characteristic for efforts of digitally transforming industrial process and manufacturing enterprises. We contribute by identifying and discussing three mechanisms through which digital representations become organizationally real: (i) noise reduction (the strategies and heuristics to filter out signal from noise), (ii) material tethering (grounding the digital representations to a corresponding physical measurement) and (iii) triangulating (in the absence of a direct correspondence, corroborating digital representations relative to other representations).