de
Spend a few weeks around e-learning and you will hear it sooner or later: SCORM is outdated, xAPI is the future. It comes up at conferences, in vendor decks and in tenders where xAPI suddenly appears as a mandatory criterion. The sentence is half right, and that is exactly why it leads to expensive decisions.
What SCORM actually does
SCORM is not a file format for videos or slides. It is an agreement on how a learning package talks to a learning management system. The package is uploaded, the LMS launches it, and while it runs the package reports back whether someone started, completed or passed, with what score and how long it took. Nothing more.
In practice you meet two versions. SCORM 1.2 is the older one and still the most widely supported. SCORM 2004 can do more, such as controlling sequence and branching inside a course, but some systems implement it only partially. If in doubt, export to 1.2 and most LMSs will leave you in peace.
Why its reputation is so poor
The criticism of SCORM is not invented. The standard comes from a time when learning happened at a desk in a browser. It measures what happens inside the package and nothing else. A conversation with the team lead, a simulation in a separate application, a short video on a phone on the shop floor: SCORM sees none of it.
Then there are technical quirks L&D teams have cursed for years. Packages that run in one LMS and not in another. Completions that never arrive because the browser window was closed too early. Anyone who has lived through that a few times understands the wish for something new.
What xAPI does better, and what it costs
xAPI solves exactly the measurement problem. Learning activities are stored as small statements following the pattern “person did something”, regardless of where the learning took place. They are stored in a learning record store, or LRS.
That is the catch. An LRS is an additional system that has to be run, connected and analysed. Many LMSs now include one, not all, and even then someone has to define which statements get recorded and what happens to the data. Without that work xAPI mainly produces data nobody reads. And works councils will rightly look more closely at tracking that reaches beyond the LMS.
For launching content from an LMS there is also cmi5, an xAPI profile meant to combine the strengths of both. Support is growing, but it is not yet widespread enough to rely on blindly.
What holds instead
The question is not which standard is more modern. The question is what you need to measure.
- Mandatory training with proof requirements, where completion and date count: SCORM is enough and the more robust choice.
- Learning across several places and applications that you want to analyse as a whole: then xAPI pays off, with a clear plan for the LRS.
For most organisations that means content should export as SCORM because the LMS requires it today. Whoever needs xAPI as well should derive it from a concrete analysis goal, not from a tender template.
More important than the standard is how quickly content can be changed. A SCORM package that has to be re-produced and re-uploaded for every small correction costs more time than any standards question. With AI-assisted production, effort per training hour typically drops by up to 80%, and corrections become a matter of minutes. What a training hour costs the traditional way is covered in How much does a training video cost?, and how production grows without extra staff in the piece on scaling training production.
If your next tender lists xAPI as mandatory, ask who in the organisation will analyse the data. If nobody raises a hand, the answer is clear.
