Challenge 3: Choosing a Release Track — Solution

(a) The very first build, wanting only the two-person dev team to try it
    -> INTERNAL TESTING. This track is meant exactly for a small,
    manually-specified list of testers (the dev team here), reaches them
    near-instantly with no Google review delay, and is the natural
    starting point before any wider audience sees the app at all.

(b) A build ready for broader feedback from 500 opted-in volunteers
    -> OPEN TESTING (BETA), assuming the 500 volunteers signed up via a
    public opt-in link (which matches "opted-in volunteers" — a group
    that isn't a small manually-curated list, but also isn't the general
    public). If this were instead a smaller, specifically-invited group
    (e.g. via an email list or Google Group the developer controls
    directly), CLOSED TESTING (ALPHA) would fit better — the key
    distinguishing factor is whether testers self-selected via a public
    link (open) or were specifically invited by the developer (closed).

(c) A routine update to an already-live app, being cautious about a
    risky change
    -> PRODUCTION, using a STAGED ROLLOUT (e.g. starting at 10% of
    users). This is already a production update, not a pre-release test
    — but the "being cautious about a risky change" detail is exactly
    what a staged rollout percentage is for: reaching a fraction of the
    real user base first, catching a serious bug's impact before it
    reaches everyone, before increasing the rollout percentage over the
    following days.

Notes:
- The deciding factors throughout: how large/curated is the intended
  audience (internal < closed < open < production), and — separately —
  for a production release specifically, how risky is the change (which
  determines whether a staged rollout percentage is warranted, rather
  than shipping to 100% immediately).
- (a) and (b) are both PRE-release testing tracks (the build isn't live
  to the general public at all yet); (c) is a PRODUCTION decision about
  HOW to roll out an already-decided release — these are genuinely
  different kinds of decisions, not points on one single scale.