So imagine you are given the monumental task of ensuring the longevity of D&D fifth edition.
What do you do? What products become priority? What gets released, and more importantly, what doesn't?
It's not a difficult question, really. 'Sustaining mode' is a standard formula. You cut costs, you focus on existing customers, you drop R&D, you don't take risks, you maintain consistency, and you let it spin out as long as it'll go.
Top priority is cutting costs. Costs can't be too low.
The next priority is to put a single, identifiable, product out there and keep it on the shelves. That's the 3 traditional core books. Nothing else should even look like a core book. I wouldn't even have gone for hard-bound adventure-path style modules as too easily mistaken for a rulebook, but old-school, magazine-style ones with a map on the inside of the removable cover.
If practicable, support the AL program to keep people playing and bring in a trickle of new players (maybe even enough to replace existing ones as they lose interest and/or die) and maintain the perception that D&D is a real thing.
Put out an OGL to keep adventures and definitively-not-core-book product flowing without costing you money. Maybe the OGL would need to be focused on enabling D&D supplements, rather than other d20 games.
Encourage (don't actively discourage) fan activities. Fan sites, fan-authored material, meetups, etc... They don't cost you anything, and are only a danger if they somehow dilute the brand.
Things to avoid:
Any big, expensive push for new players. No mainstream advertising, for instance.
Expensive software or on-line service initiatives: too expensive to develop and keep current, not the core appeal of TTRPGs.
Any risks to the existing market: Avoid excessive controversy (a little controversy helps peek interest), don't violate customer expectations (don't even try to 'exceed' expectations, that's a continuous climb and would eventually raise costs). The mantra should be "keep doing what you've always done" (just do it cheaper).
Endangering brand Identity: Don't make changes to the existing product, even if they might theoretically be improvements. Don't issue errata (save it for a later printing), don't roll revs. Any new edition (new printing, really) should have nothing more than different art and cleaned up text.
