I don’t think one has to be incompatible with the other. They serve different purposes. Even after analysis, insights need to be transformed into research artefacts that can live for longer, and raw data need to have limited access to protect our participants’ privacy and avoid bias. Especially in bigger companies, not everyone is able to be involved in analysing research collaboratively, and even when we involve others, this is a complex process that should still be guided by the researchers. Most likely people from different departments will have those exposure hours regarding a specific product but would also like to see digestible insights that might apply to a wider use case regarding another product or sub-set of products. Repos make it also possible for research to influence earlier in the process as opposed to only as a way to “validate”. So I think overall these approaches can be combined, depending on each company and processes.