An alternate take: there are good DEI programs and poor ones. The poor ones fail because the planners dont really know what they are trying to do, but leadership thinks they ought to have one, and so they metric-ize it. And since (again, no clarity of thought) hard numbers in areas like hiring sail perilously close to large legal rocks, they whiff on the metrics and end up measuring something like "engagement". And, concomitantly, deliver a lot of low value chatter that provides ample ammunition to opponents of any kind of DEI programs, even the good ones.

A good DEI program should, IMHO, be indistinguishable from good management culture embedded at every level in an org.

- It should not be controversial to assert, and product management to insist, say, that products designed for humanity should be usable by humanity: men and women, for example - but we still have medicine and cars tested on male models, and software that is unusable if you have low vision or cant operate a mouse and keyboard simultaneously. That doesn't automatically mean one must hire 50:50 men:women, say (see legal rocks, above), but it certainly starts to smell like a missed opportunity if you don't have a single person on your staff or in your network of consultants who can explain what it feels like to wear a seatbelt when you are 1.5m and 50kg not 2m and 85kg. If you want better products, this seems like a no brainer, but it doesnt seem to happen.

- It must absolutely be a mandate for all managers to avoid cliques. All men? All women? All Indians? All Purdue grads? Close watching needed, especially when those groups hire and promote. Doesn't need a mandate, needs better managers of managers.

Tldr is that no amount of DEI will fix bad management culture.