To price Twitch Channel Points rewards, start with the time you can give them during a broadcast. Decide how many requests you can fulfill, set costs around how accessible you want each reward to be, and add limits that keep the workload manageable.
A points price and a redemption cap do different jobs. The price affects how long someone may need to save. The cap controls how many requests you accept, including from regulars who already have a large balance.
The worked plan below is for custom rewards you fulfill yourself, with Channel Points already enabled on your channel. It uses a two-hour stream, three rewards, and a small allowance for interruptions. The figures are starting assumptions to adapt, not recommended prices for every streamer.
Set a workload budget before choosing prices
Suppose you want to spend no more than twelve minutes of a two-hour stream on reward requests. That is your planning allowance. It includes explaining the request, switching away from the main activity, doing it, and getting back into the stream.
Choose rewards with a defined end. “Ask one question about today’s project” can fit into a short pause. “Tell me anything to do” leaves the commitment open. If you still need suitable actions, the Channel Points reward ideas guide gives examples with clear boundaries.
For each reward, estimate the total interruption and decide its maximum number of uses. Use a slightly generous time estimate rather than counting only the few seconds when you perform the action.
Time per request × maximum uses = planned workload.
| Reward | Time allowed per request | Maximum uses this stream | Planned workload |
|---|---|---|---|
| Ask about today’s project | 2 minutes | 3 | 6 minutes |
| Choose an unlocked outfit for the next round | 1 minute | 2 | 2 minutes |
| Meet the desk mascot during a break | 2 minutes | 1 | 2 minutes |
| Total | — | 6 requests | 10 minutes |
This leaves two minutes of the twelve-minute allowance for overruns. The example assumes you already have the outfit choices and a desk mascot you can show. Change the actions to fit your stream and equipment.
The maximums are ceilings, not targets. A quiet stream with one request can still be a good stream. You are checking whether the menu remains workable if viewers claim everything you have offered.
If you shorten the broadcast, revisit the allowance before going live. A plan that fits a relaxed two-hour session might crowd out the main activity during a forty-minute stream. Increasing prices will not necessarily help if the people watching already have enough points.
Choose costs by how accessible the reward should be
Put the actions you are happy to repeat within easier reach. Set a higher saving goal for something you want to feel occasional, then keep its workload cap in place. You can make an action accessible without promising unlimited uses.
For a sense of scale, Twitch’s viewer guide lists 10 points per five minutes of non-subscriber watching and 50-point click-to-claim bonuses at fifteen-minute intervals. A two-hour model with 24 watch awards and eight claimed bonuses produces 240 + 400 = 640 points from those sources.
That is a calculation under stated assumptions, not a universal session payout. Streaks, other bonuses, subscription multipliers, missed claims, and existing balances affect what a viewer has available. Use the current earning table when revisiting your prices.
Here is one possible price ladder for the worked menu:
- 300 points for a project question: A relatively accessible way to join the conversation, with the three-request stream cap still protecting the schedule.
- 900 points to choose an unlocked outfit: A longer saving goal for a choice that affects the next part of the broadcast.
- 1,800 points to meet the desk mascot: An occasional feature, offered once during a suitable break.
Under the 640-point model, a viewer starting at zero could reach those amounts in roughly one, two, or three complete sessions if they saved everything. Someone with an existing balance could redeem immediately. The prices do not guarantee a particular frequency.
The mascot request takes the same planned time as a project question but costs more in this example because you want it to happen less often. Pricing does not have to be proportional to seconds of work. The action’s place in the show, its appeal to viewers, and the saving goal you intend also matter.
If every reward requires several broadcasts of saving, a newcomer may have nothing they can try soon. Keep at least one choice within the reach you intend for occasional viewers, and describe what they receive clearly enough to judge whether it is worth saving for.
Give each limit a specific job
Twitch’s creator guide documents cooldowns, per-stream limits, and per-user limits for custom rewards redeemable while live. These controls solve different problems:
| Control | What you are controlling | Example use |
|---|---|---|
| Cost | Points needed for one request | A project question costs 300 points. |
| Cooldown | Spacing between redemptions of that reward | Leave time between project questions. |
| Per-stream limit | Total uses of that reward during the stream | Accept at most three project questions. |
| Per-user limit | Uses by one viewer during the stream | Let each viewer claim one project question. |
For the worked plan, start with a twenty-minute cooldown on questions and a thirty-minute cooldown on outfit choices. Use a one-per-viewer limit on each reward, alongside the stream caps of three, two, and one. These are example settings to test against your actual pace.
A cooldown spaces redemptions; it does not schedule when you will fulfill them. If a question arrives during a difficult round, the description can still say you will answer at the next break.
Also plan for different rewards arriving together. A question and an outfit request are separate rewards, so do not treat their cooldowns as one shared waiting period. Twitch’s custom reward reference defines the limits and cooldown for each reward. The combined workload still needs your attention.
The one-per-viewer rule in this example spreads opportunities across people. It also means a stream with only two interested viewers might not reach the three-question cap. That is an acceptable result if sharing opportunities matters more to you than using every available place.
Write the promise viewers are actually redeeming
Set up the custom menu in Creator Dashboard → Viewer Rewards → Channel Points → Manage Rewards. Before enabling a reward, write down the action, the timing, and any choice the viewer must make. The creator guide above covers custom reward setup and optional text input.
For the question reward, a usable description would be:
Ask one question about today’s project. I’ll answer at the next natural break, for about two minutes. One request per viewer and three available this stream. If I cannot cover it today, I’ll reject the request and return your points.
For the outfit reward, list only choices already available in the game and say when the change starts. Decide how long it lasts so a second request does not immediately undo the first. One option is “from the next round until that round ends.”
For the mascot, name the planned break. Pause the reward when that opportunity has passed rather than leaving a promise you cannot fulfill before sign-off.
Plain titles help viewers tell these offers apart. Your themed points name can stay the same across the menu; the individual reward titles should explain the actions. The Channel Points name and icon guide handles that visual and naming decision separately.
Twitch describes Smart Costs for default rewards in its creator guide. This worked plan uses manually chosen costs for custom actions; it does not assume a default reward’s automatic pricing will manage your personal workload.
Treat the queue as a list of unfinished promises
Keep requests that need later attention in the requests queue. Twitch’s Making the Most of Channel Points explains completing requests, rejecting them with a refund, and the option for suitable rewards to skip the queue.
For this menu, leave the queue enabled because the question, outfit change, and mascot appearance all require you to act. Marking a request complete should follow fulfillment. A redemption notification tells you that someone made a request; it does not mean the promised action happened.
Check the queue at the natural pauses already built into your stream. When two different rewards arrive together, tell the viewers when you will handle them. In the example, the outfit choice might apply to the next round while the project question waits until the following break.
If the remaining stream cannot fit the pending work, pause the relevant rewards and review what is already owed. Twitch’s Channel Points FAQ confirms that rewards can be paused. Pausing stops further requests; it does not resolve the ones already waiting.
Fulfill a request you can still honor, or reject and refund one you cannot. Do not mark it complete merely to clear the list, and do not let an unplanned delay become an indefinite promise for a future broadcast.
If a moderator helps, agree who checks requests, who tells viewers about a delay, and who resolves an unfulfilled request. Avoid a situation where both of you assume the other person handled it. Twitch’s creator guide describes the queue tools moderators can use.
Test a burst of requests before expanding the menu
Walk through the busiest plausible case with the settings you intend to use. For the worked plan, imagine that a question and an outfit request arrive together, followed by the mascot request during the next break.
Check the sequence:
- Can you identify each request and the information needed to fulfill it?
- Does the viewer know whether the action happens now, next round, or during a break?
- Can you handle the three requests without dropping the activity people came to watch?
- If the stream ends early, can you find and resolve anything still unfinished?
The workload table protects the total allowance, but this exercise tests clustering. Ten minutes spread across two hours feels different from several interruptions in the same segment. You may need fewer available rewards during focused sections, even when the arithmetic fits.
Review the menu again when the format changes. A reward that is easy during a building session may be awkward during a tournament round or a story scene. Enable it when there is a real opportunity to deliver what the description promises.
Adjust the cause of the problem after a trial stream
Keep a short record of the reward, number of requests, approximate time spent, and anything left unfinished. Record what you can observe; Twitch’s FAQ says creators cannot see an individual viewer’s Channel Points balance. You do not need a balance leaderboard to see that four requests waited until sign-off.
Use the record to make a specific change:
- A reward was claimed rarely, but viewers asked what it meant: clarify the offer before lowering the price.
- Requests arrived faster than you could finish them: reduce the cap, increase spacing, narrow the action, or pause it during focused sections.
- One action repeatedly took longer than estimated: revise its scope and recalculate the workload allowance.
- The queue was manageable, but regular viewers said the saving goal felt too long: trial a lower cost while keeping the capacity limits.
Change one relevant setting and keep the previous value in your notes. A simultaneous price cut, new reward description, and extra availability make it harder to understand why behavior changed.
Before the next broadcast, you should be able to state what each reward costs, the maximum work it can create, when you will fulfill it, and what happens if you cannot. Keep that plan beside the queue. It is a more useful operating rule than copying another channel’s prices.