Despite wide use, thermodynamic conventions are almost never presented explicitly as a topic in textbooks, lectures or publications covering metrology and uncertainty. Their meaning, and the particular need for them, is typically just taken for granted. Consequently, an uncritical acceptance and lack of understanding of their implications can often start in the classroom and persist long after, with unfortunate consequences that undermine some important metrological definitions and practices. Thermodynamic conventions generally create reference scales that allow properties which can only be measured as differences (such as enthalpy and electrode potentials) to be expressed in absolute terms for practical convenience. However, the arbitrary nature of these choices made when selecting and implementing any such convention is not only puzzling to students but can also confuse thermodynamic or metrological specialists. If, as a consequence, thermodynamic conventions are not properly understood, theoretical and experimental progress can become stuck within prevailing paradigms. In this work, we identify two such cases relating to pH and Gibbs energies of reaction, and show how more nuanced understandings of thermodynamic convention as a concept enable better choices and lead to improved scientific outcomes.