A detail buried in most AI product launch announcements, easy to skip past, turns out to be a reasonably good predictor of whether a new feature is still around and actively maintained a year later: whether it launched as an opt-in option or was turned on by default for existing users.
Why this distinction matters
A default-on feature gets exposed to a company's full existing user base immediately, generating usage data and feedback at scale from day one. An opt-in feature only reaches users who actively sought it out, which is a much smaller and more self-selected group, generating a slower trickle of both usage and feedback.
How this affects a feature's survival odds
Features with slow, opt-in-only adoption are more vulnerable to being quietly deprioritized or discontinued, because the visible usage numbers used internally to justify continued investment never reach a threshold that makes them a priority relative to other initiatives competing for the same engineering resources.
Why companies choose opt-in for features they're not fully confident in
Launching as opt-in is a lower-risk way to ship a feature that isn't fully validated - it limits exposure if something goes wrong, but that same caution limits the feature's chance to generate enough usage to justify its own continued existence. It's a reasonable short-term risk management choice that sometimes has a real long-term survival cost for the feature itself.
What this means for evaluating a new launch
A default-on launch signals real confidence, or a decision to force adoption regardless of confidence. An opt-in launch signals caution that may be entirely appropriate - but it also means the feature is on a real clock to prove its value before it's quietly cut, a dynamic worth watching if you're planning to build a workflow around it.