It's very sad because the kind of difficulties Marriott faces are very well-defined and well-understood as generalized versions of them come up everywhere in software. While I've never worked on creating a hotel reservation and inventory management system before, I have worked on analogous projects of much greater complexity and flexibility than what Marriott needs to support, and I believe that a relatively small team of skilled engineers could build a much more reliable and flexible replacement for MARSHA in under two years along with a process to migrate the information for the existing system. Whoever Marriott contracts for their IT is honestly very inept.

For example, I have been involved with data migration efforts before and the way they handled "merging" of SPG and Marriott accounts was one of the most hilariously broken, untested, and poorly conceived processes I've seen of that sort. Try reconciling many sources of media with varying levels of inconsistency in their metadata, and matching things appropriately between a user's own tracks, store tracks, streaming, etc into a unified cloud library. That's a level of complexity way beyond the merging of the Marriott and SPG accounts which both are very well-defined in what information they will have and how it's structured.

When companies like Marriott truly realize the potential value of leveraging data about their customers to offer personalization in a way that gets them to spend more money, they're going to regret not having technology as a core competency. I feel like some banking institutions in the world are already feeling the pain of this, whether it be because of outsourcing or just outright lack of investment.