A giveaway trades prizes for engagement: follow, like, tag a friend, win something. It reliably spikes every number it touches, and reliably gives part of the spike back when winners are drawn, because an audience recruited by prizes contains people who came for the prize.
Entry conditions are engineered distribution: tags drag friends in, shares carry the post outward, follows are the toll at the door.
It's real reach, cheap, and platforms tolerate the honest version. Launches, milestones and collabs use giveaways precisely because the spike is dependable.
When winners are announced, the prize-motivated slice unfollows. That drop is public on your growth chart, and the audience that stays skews less engaged than one recruited by content, which your engagement rate then reports honestly.
The loop-giveaway variant (many accounts pooling one prize, entrants follow all of them) scales spike and hangover together, and platforms have cooled on it: mass follow-for-prize patterns read as manipulation to the systems that police everything else.
A share unfollows after the draw, and more go quiet. Relevance decides the kept fraction: niche prizes keep more, generic prizes keep least.
They deliver volume with the worst quality and the most platform risk of the family. If the goal is a number, compare prices against direct delivery; if the goal is audience, prefer single, niche-prized giveaways.
Honest ones aren't: platforms publish promotion guidelines (disclose, don't demand shares-to-DMs, follow local law). Mass-coordinated follow farms are where enforcement starts caring.
Every definition lives in the full glossary, and the services behind the vocabulary are on the services page.