Skip to main content

Market-neutral small-business guide

Single-Platform Dependency Risk

Measure channel contribution and cash exposure before funding diversification without inventing a disruption probability.

Fix the comparison boundary before calculating

Records and assumptions to align

  • Reconcile platform-attributable orders, refunds, fees and retained contribution for one period. (not complete)
  • Map receivable, reserve and payout timing separately from profit. (not complete)
  • Choose an interruption duration and recoverable-order assumption as labelled sensitivities. (not complete)
  • Price the fallback setup, duplicate operating cost, transition effort and spare capacity. (not complete)
  • Name non-financial dependencies such as catalogue data, reviews, customer access and fulfilment eligibility. (not complete)
Minimum evidence and unit contract
FieldRequired unitPreferred evidence
affected platform ordersorders/windowcompleted order record or labelled scenario
retained contribution per orderCU/orderreconciled marketplace order cohort
contribution recovered through the named fallbackCU/windowobserved test or labelled sensitivity

Build a reproducible scenario

Bounded platform-interruption contribution exposure

contribution exposure = affected platform orders ร— retained contribution per order - contribution recovered through the named fallback + incremental interruption costs

affected platform orders
Orders exposed during the selected interruption window (orders/window) โ€” completed order record or labelled scenario
retained contribution per order
Revenue less included product, fulfilment, payment, platform and return costs (CU/order) โ€” reconciled marketplace order cohort
contribution recovered through the named fallback
Only contribution supported by fallback capacity and an explicit conversion assumption (CU/window) โ€” observed test or labelled sensitivity

Show withheld or delayed payout cash separately. Do not multiply revenue by an outage probability supplied by Margin101.

Fictional four-week interruption scenario

The business supplies the interruption window and recovery assumptions; neither is a forecast.

Inputs, intermediate arithmetic and result
CaseDeclared inputsSubstitutionResult
Gross contribution exposed400 affected orders; 24 CU retained contribution/order400 ร— 249,600 CU
Fallback recovery120 recoverable orders; 21 CU contribution/order120 ร— 212,520 CU recovered
Net bounded exposure9,600 gross; 2,520 recovery; 800 transition/rework cost9,600 - 2,520 + 8007,880 CU exposure
CU means fictional neutral currency units. Every figure is an educational scenario, not a benchmark, quote or forecast.

Stress-test the uncertain inputs

Change one assumption at a time
VariableLower caseHigher caseWhat it tests
Interruption windowOne weekSix weeksTests duration concentration
Fallback capacityLimited tested capacityPre-built capacityTests recoverability rather than revenue concentration
Payout timingNormal dated receiptDelayed/reserved payout scenarioTests liquidity separately from contribution

Stop and review when

  • Platform order contribution cannot be reconciled after refunds and fees. (not complete)
  • Fallback recovery is asserted without capacity or conversion evidence. (not complete)
  • The cash and contribution horizons have been mixed. (not complete)

Turn the scenario into a controlled decision

  1. Reconcile the baseline to current records and name the evidence owner.
  2. Run the base case, then change one uncertain input at a time.
  3. Record the chosen response, approval limit, review date and stop trigger.
  4. Compare actual results with the original boundary before reusing the assumption.

Avoid these mistakes

  • Calling all platform revenue a loss when some costs are avoided or orders recover. (not complete)
  • Using an invented outage probability to create false precision. (not complete)
  • Ignoring the fixed cost and management load of a fallback channel. (not complete)

Questions to resolve before approval

Is a high platform revenue share automatically unsafe?
No. Concentration identifies exposure, not probability or unacceptable risk. Recovery capacity, payout timing, contribution and tolerance all matter.
Does the model value customer data or reviews?
No. Record those dependencies qualitatively or in a separately evidenced scenario; do not hide a speculative value inside contribution.

Methodology and source boundary

Change history

  1. โ€” Initial public release of the article after pre-launch factual, editorial, source and presentation review.