The mobility of data has been variously described as data: flows, streams, journeys, threads, transfers, exchanges, and circulation. In each case, the mobility of data is conceived as a movement from here to there; that data moves along a chain of receivers and senders. However, we contend that the metaphor of a data flow (or stream, journey, etc.) does not reflect well the sharing and circulation of digital data. Rather, data replicate (copies), with the original source retaining the data and a new source gaining it, and data proliferates (multiplies) and diffuses across systems and sites when made openly available. As data replicate and proliferate they are transformed through processes of data cleaning, data wrangling, data fusion and enrichment, producing new incarnations of the source data. Moreover, data does not replicate and diffuse alone, but with companions, such as other data (e.g., metadata, derived data) and information (e.g., documentation, visualisations). The replication, proliferation and diffusion of data is facilitated by seams (interface/connection points between systems) and aided by metadata, standards and protocols, and hindered by frictions and vulnerabilities. We illustrate our argument through a case study of the planning data ecosystem in Ireland.