Before making LIVE a recurring channel, a brand should be able to answer two questions: what does a completed, retained sale contribute, and can the team deliver the promises made during the session? Neither answer comes from the number of orders received alone.
Give the session a customer job
The fictional business in this example, Plain Weave, is an existing Malaysia home-textile shop. Assume it has separately verified its shop, product and chosen LIVE feature eligibility under its applicable current terms. This example does not establish eligibility for another seller, market or product.
Plain Weave plans to demonstrate two cushion-cover fabrics. Customers struggle to judge texture and care requirements from photographs. The session's job is to make those differences visible, including limitations. A presenter will compare the materials under the same light, show the reverse and fastening, explain the actual care instructions and answer questions about the displayed variants.
That is a useful reason to go live. “We need a big sales event” gives the team less guidance about what to demonstrate or what a customer should understand. TikTok's Malaysia content policy requires accurate, useful and verifiable content. A running order count does not justify unsupported claims about a fabric's performance.
Calculate contribution after the order has an outcome
All figures below are invented planning assumptions in Malaysian ringgit, not platform fee quotes or reported brand results. The shop assumes a sale price of MYR80 per cover. For a retained sale, product cost is MYR28, platform and payment charges MYR6, packing and delivery MYR10, and variable creator/service work MYR6. That leaves MYR30 per retained sale before session fixed costs and separate failed-order costs.
The fixed session cost is MYR900. With no failed-order costs, thirty retained sales would cover that fixed cost: 900 ÷ 30 = 30. This is a conditional break-even calculation, not a promise that thirty placed orders make the session profitable.
Plain Weave then examines a hypothetical completed group of forty paid orders. Four cancel before fulfilment and, for this illustration, leave no net cost. Thirty-six are delivered; four of those receive full refunds, leaving thirty-two retained sales. The four refunded cases each leave a net loss of MYR20 after any recoveries and reversals. That separate MYR20 assumption includes all remaining costs of each refunded case; the retained-sale costs are not also applied to those cases.
| Item | Calculation | MYR |
|---|---|---|
| Contribution from retained sales | 32 × 30 | 960 |
| Net losses from fully refunded cases | 4 × 20 | −80 |
| Fixed session cost | One session | −900 |
| Session result under these assumptions | 960 − 80 − 900 | −20 |
The live order counter showed forty paid orders. The model still loses MYR20 before other business overhead and any applicable taxes. Pending deliveries cannot be quietly classified as retained sales to improve the result. Partial refunds, replacement shipments or cancellation costs would require their own treatment if they occurred.
The fixed cost also has to mean what it says. Here, MYR900 covers the separately budgeted preparation, host, production and allocated fulfilment-readiness time. The MYR6 variable amount covers order-related creator/service work; no hour or payment is included in both. Actual contracts and work records must replace these assumptions before a real spending decision.
Dashboard terminology needs the same care. A Malaysia Seller Center analytics manual published in 2025 illustrates that LIVE revenue metrics can include refunded orders. It is a reason to inspect the definitions and exports in the shop's current account, not to assume every historical dashboard definition remains unchanged. Accounting for the session requires reconciling paid orders, adjustments and retained outcomes.
Plan the busiest minute, not just the average hour
Plain Weave limits this exercise to single-cover orders and forty allocated units across the two displayed variants. The limit is an exposure control, not a sales forecast. It must be enforced through the shop's available controls and reconciled with other channels.
The host demonstrates; a separate producer monitors questions and stock. Neither is counted as packing capacity during the session. Afterward, two trained packers each have two protected hours. A rehearsal establishes twelve simple orders per person-hour:48 theoretical orders, with eight order-equivalents reserved for checks and exceptions, leaving forty. A named service owner is available for the following support period and can approve the appropriate remedies. These are fictional staffing assumptions; real teams need to time their own tasks and confirm delivery requirements.
A forty-order cap can still fail if thirty questions arrive together and the producer cannot answer accurately. The presenter therefore pauses selling prompts when a repeated uncertainty emerges. Questions are collected, answered with the actual product and added to the product page afterward. “Keep the show moving” is not a reason to guess.
Use a run-of-show that teaches the distinction
| Session time | Customer question | Demonstration and operating action |
|---|---|---|
| 0–5 minutes | What exactly is being sold? | Name the two variants, contents and scope of the demonstration; producer verifies live stock |
| 5–12 minutes | How do the fabrics differ? | Show both under the same light, including reverse and fastening; avoid unverified performance claims |
| 12–20 minutes | Which fits my needs? | Compare texture and actual care instructions; explain situations in which neither is suitable |
| 20–30 minutes | What remains uncertain? | Answer collected questions; producer flags uncertainty and variant limits before another purchase prompt |
| 30–35 minutes | What happens after I order? | Restate the actual offer and confirmed delivery information; close with reconciliation and handover |
The schedule does not authorise promises beyond the shop's confirmed arrangements. If a displayed variant becomes unavailable, the team removes the corresponding prompt and follows the applicable customer process for existing orders. It does not silently substitute a different cover.
After the session, the team records unresolved product questions, reconciles stock and moves accepted orders into the actual fulfilment workflow. Watching the replay cannot replace that handover. Follow-up analysis waits for meaningful order outcomes without delaying any customer's remedy.
Decide whether a second session deserves to exist
In the worked economics, repeating the MYR900 format on the same assumptions is unattractive. One alternative is a simpler format costing MYR600. If every other outcome were identical, the model would produce MYR280: 960 − 80 − 600. That is a sensitivity test, not a forecast. A cheaper format may change audience response, staffing or demonstration quality, so those assumptions need checking.
Another option is to use a recorded comparison if customers mainly need the same demonstration and do not benefit from real-time questions. A third is to pause the offer if retained-sale contribution is uncertain or the delivery route cannot support the session. LIVE is not automatically the best answer to a product-information problem.
The evidence that justifies repetition is therefore specific: customers understood the distinction, accepted orders were handled reliably and retained sales supported the actual cost of the format. A busy session is an event. A viable operating channel earns the next session through those outcomes.