Compare the full operating fit
Headline fees are only one input. For each plausible venue, compare buyer fit, cost shape, owner time, cash timing, service fit, and how much control remains if the venue changes its terms or disappears. The archetypes make those trade-offs easier to scan:| Venue archetype | Where demand comes from | Operating load and cost shape | Main constraint to test |
|---|---|---|---|
| Marketplace | Platform search and existing buyer intent | Per-transaction costs plus listing and fulfillment work | Buyer fit, current costs, and payout lag |
| Direct storefront | Traffic the business brings | Site overhead and payment costs; ongoing demand work | Whether measurable traffic becomes orders |
| Live selling | Platform discovery plus a scheduled audience | Live hours, preparation, moderation, and platform costs | Whether contribution justifies scheduled selling time |
| In-person | Event or local foot traffic | Table or travel costs, setup, negotiation, and a fixed selling window | Whether buyer fit and fully loaded hours work |
Eliminate venues that break a hard limit
Begin with the lane and cash window from Chapter 3.2. Remove a venue if its buyers do not fit the inventory, its current costs push modeled contribution below your floor, its payout lag exceeds the cash limit, or its recurring work exceeds the hours available. Chapter 2.4 owns the price floor; do not invent a new one for the venue test. Until Chapter 4.4 defines the full service policy, use a provisional gate: a reliable handling cadence, maximum orders per ship run, and sustainable response time. Remove a venue whose current rules exceed it, then revisit the gate after writing the full policy. What remains is the test set. If two venues still look plausible, choose the one whose largest uncertainty can be measured with the least cash and irreversible setup.Test one venue long enough to see the workflow
Bring forward the provisional venue, cohort, completed orders, owner time, and friction record from Chapter 1.4. Add the cost, contribution, cash-lag, and pricing inputs you built in Part 2. This is the baseline; do not discard it and start the selling test again. Run or extend one primary-venue window until it covers 30 days with the measures below. Orders from the earlier work-fit test may count when they used the same cohort and rules and captured the required measures. Do not reconstruct a missing exposure or time record after the fact. Thirty days is an experiment boundary, not proof across every season or scale. Write the complete confirmation test before launch or extension: Venue and cohort: Primary venue ___; defined inventory cohort ___; price policy ___. Current inputs: Verified on ___ from ___; modeled venue costs ___. Activity and exposure: I will complete ___ and measure ___ as exposure. Operating limits: Owner-hour cap ___; maximum sale-to-cleared-cash lag ___; provisional handling and response limits ___. Decision controls: The test is valid only if ___. Review on ___. Confirm if ___; switch if ___; add another venue only if it has this distinct job: ___. Otherwise extend or rerun the test. Use a cohort you can count. If inventory is added during the test, record it separately so a changing denominator does not make demand look better or worse.Measure demand, economics, and effort separately
Add one venue-specific exposure measure: qualified store visits, live viewers, event conversations, or listing impressions when available. Define it before launch so zero sales can be separated from zero exposure. Cohort sell-through = completed cohort units sold under your return convention ÷ cohort units offered Contribution per owner hour = total modeled contribution from the cohort ÷ venue-specific owner hours Cash lag for an order = cleared-payout date − sale date For sell-through, use the same cohort unit in the numerator and denominator. Both ratios require positive denominators. If no owner hours were recorded, contribution per hour is not applicable, not zero. Before launch, declare when a sale counts as completed and how cancellations or returns change the numerator. Model contribution with the established cost model rather than gross sales, and record every venue-specific hour. Finalize cash lag only after payout clears. At review, keep uncleared orders visible as pending. A pending order that has already exceeded the maximum lag fails the limit; one still inside it leaves the cash condition unresolved. These are signals, not isolated diagnoses. Sell-through combines exposure, inventory, price, listing quality, and buyer fit. Contribution per hour describes observed economics and work; cash lag shows when completed sales became usable cash.Decide the branches before the result
First test whether the experiment itself is valid. If planned activity was not completed, the cohort or rules changed materially, or a key measure remains unavailable, mark the result inconclusive and extend or rerun it. Do not force weak evidence into a business decision. Then apply the precommitted branches:- Confirm when the test was completed and buyer fit, modeled contribution, owner hours, cash lag, exposure, and service performance all clear the thresholds you wrote, applying the pending- payout rule above.
- Switch when the test was completed faithfully but a hard venue constraint still fails. Name the failed constraint and what the next venue must solve.
- Add only when the first venue works within capacity and a second venue has a distinct job, such as reaching a buyer segment the first does not serve. Addition is not a substitute for repairing a failed first channel.