Most prize decisions get made backwards. A streamer decides what they can afford, buys the most impressive thing in that range, and gives it away. The prize is treated as a reward for the audience, when what it is actually doing is deciding which audience turns up.
That is the whole argument of this page, and it is worth stating plainly before the lists start: you are not buying a thing for a viewer. You are buying the composition of the room. Every other question — budget, quantity, digital or physical — follows from deciding who you want in it.
Expensive prizes attract the wrong people
A graphics card is worth entering for regardless of whose stream it is on. So it brings in the people who enter every graphics-card giveaway on the internet, most of whom will not remember your name the following week, and it brings enough of them that your regulars notice their odds got worse.
There is a second effect that is easy to miss: a big prize can feel unwinnable. Four hundred people entering for a card makes a viewer do the arithmetic and conclude it is a lottery. Twelve people entering for a custom emote feels like a raffle at a school fair — worse odds are less discouraging than they sound when the prize feels human-scale.
Free, and better than it sounds
A custom emote or a badge named after the winner. A Discord role nobody else has. Your VOD-editor’s time on a clip of their choosing. A character named after them in the next playthrough. A game night where the winner picks what you play and joins the lobby.
These share the property that makes them work: they are worthless to a stranger and valuable to somebody who watches you. That is exactly the filter an expensive prize fails, and it is why small channels often run better giveaways than large ones.
Prizes you already have
Before buying anything, look at what is already sitting in your account. Bundle keys you bought for one game and never redeemed. Codes a publisher sent you for a sponsored stream and told you to hand out. A spare peripheral. A second copy of a game you own on two stores.
These are free in the sense that matters — the money is already spent — and they are usually better prizes than something bought to be a prize, because they are things you actually chose once. A key for a game you loved enough to buy twice comes with a recommendation attached, and that recommendation is most of what makes the segment interesting to watch.
Cheap, and the workhorse
A game key from a bundle, which costs a few pounds and often means more to a viewer than something dearer. A key for the game you are playing that night, so the giveaway and the stream are the same thing. A mousepad or a sticker pack. A month of something the audience already uses.
Keys are the most efficient prize in streaming and the reason is delivery: they arrive by email in seconds, nobody hands over an address, and you can give away twenty of them in one draw instead of running twenty segments.
Physical prizes cost more than they cost
The sticker price is the smallest part. A physical prize means collecting an address from a stranger, packing something, paying postage that is sometimes a meaningful fraction of the prize, and finding out whether your winner is in a country you can affordably ship to. It also means holding somebody’s home address, which is a responsibility rather than a step.
None of that is a reason never to do it. It is a reason to decide before you announce it rather than after somebody has won — including which territories you can actually post to, because the rules page will say the giveaway is open to a list and that list is displayed rather than enforced.
Merch, and the drop it can launch
Merch is a good prize for the same reason it is a difficult one: it is worthless to a stranger, which is the filter you wanted, and it has all the shipping problems above. If you already have stock, giving a piece away costs you the unit rather than the retail price.
The more interesting use is as a launch. A drop announced on its own is a sale; a drop with the first piece given away on stream is a segment, and the draw gives everyone a reason to be there at the moment the store opens. The giveaway is the ad for the thing, which is the whole argument for running giveaways at all.
How many, and how often
More small prizes beat one big one on almost every measure that matters. Three winners feel like three moments, three names get read out, and three people leave with a reason to come back — against one person leaving with something nicer.
There is a rough sizing rule hiding in that. If you can afford one fifty-pound prize a month, three ten-pound prizes across three streams will be remembered by more people and cost you less — and the three draws give you three reasons to tell your audience something is happening, which is the scarce resource rather than the money.
And a giveaway worth farming will be farmed. If your prize is valuable enough that entering is worth somebody’s while as an activity rather than as a moment in your stream, you have bought a different audience than the one you wanted, at your own expense.
Which is the test to apply to any prize idea before the budget question is even asked: would somebody who has never heard of you enter this? If the answer is yes, that is who will fill the entry list. Sometimes that is exactly what you want, on a sponsored stream or a launch night. The mistake is arriving there by accident, on an ordinary Tuesday, and then wondering why the numbers went up and the chat stayed quiet.