International comparisons in education are now regularly produced and consumed, deeply influencing Europe and nations across the world. But only a few decades back, establishing international comparability across diverse education systems appeared almost impossible. An enormous machinery of expertise, trust, mathematics and other resources needed to be mobilised to tame the monsters of diversity and uncertainty and impose calculability, comparability and predictability. This paper traces, historically, the production of the OECD education indicators, from the establishment of theInternational System of Education Indicators (INES) to the routine, annual publication of Education at a Glance, where diverse nations are assembled together in a single spatiotemporal frame, using a universal set of metrics.Using concepts from actor-network theory, it describes the ways in which varied constraints were addressed, interests translated, categories defined, classifications negotiated, frameworks agreed upon, choices made, methodologies established and protocols developed, as the indicators exercise moved from being nearly impossible to becoming routinely performed. However, the success of such a project itself appears to have created a monster – with such comparisons sometimes wildly misunderstood and misused. The paper ends with questions regarding how European and international scholars might respond to the challenges of runaway comparisons and standardisations in education.