Skip to main content

Market-neutral small-business guide

Good-Better-Best Packages: Test the Margin Mix

Reconcile each package, test the assumed customer mix and define rollout checkpoints without treating the mix as a demand forecast.

Define the packages before testing the mix

Inputs and boundaries that must be explicit before comparing scenarios
Input or boundaryUnit and sourceDecision use
Package scopeNamed deliverables, usage or service boundary; business recordShows what the customer receives and prevents a cosmetic tier.
Package priceCU per package on one indirect-tax basis; user scenarioProvides the revenue side of package contribution.
Variable cost to serveCU per package; reconciled record or labelled assumptionIncludes only costs that belong to the declared package boundary.
Customer mixCustomers or orders per period; observed record or explicit scenarioWeights package contribution without pretending to forecast demand.
Capacity and support boundaryHours, units or service commitments per period; operating recordStops a package that cannot be delivered at the modelled cost.
Keep every row on one currency, unit, period and indirect-tax basis. Scenario values are user assumptions, not market benchmarks.
Core calculation

package contribution = package price - package variable cost; period contribution = ฮฃ(package contribution ร— package customers)

package price
Entered price for one package on the declared tax basis (CU/package) โ€” user input
package variable cost
Cost that changes with delivery of the package inside the stated scope (CU/package) โ€” business record or labelled assumption
package customers
Observed count or explicitly labelled mix scenario (customers/period) โ€” business record or user scenario

The sum is a contribution comparison, not complete accounting profit. Shared fixed costs, implementation cost and capacity constraints remain separate unless explicitly assigned.

Worked example: compare two package-mix scenarios

The business tests 100 customers in one period. Prices and costs use the same neutral currency and indirect-tax basis.

Reproducible fictional scenario in neutral currency units
StepInputs and arithmeticResult and interpretation
1. Calculate Good contributionCU100 price - CU55 variable costCU45 contribution per Good package.
2. Calculate Better contributionCU140 price - CU75 variable costCU65 contribution per Better package.
3. Calculate Best contributionCU190 price - CU110 variable costCU80 contribution per Best package.
4. Baseline mix(CU45 ร— 50) + (CU65 ร— 35) + (CU80 ร— 15)CU5,725 period contribution for a 50/35/15 scenario.
5. Revised mix(CU45 ร— 35) + (CU65 ร— 45) + (CU80 ร— 20)CU6,100 period contribution for a 35/45/20 scenario: CU375 higher than baseline.
6. InterpretCU6,100 - CU5,725The revised assumed mix improves contribution in this model; it does not prove customers will select it.

Test what would change the decision

Sensitivity and boundary cases
CaseRecalculation or evidenceDecision consequence
Best support cost rises by CU20Best contribution becomes CU190 - CU130 = CU60; revised mix becomes CU5,700.The revised mix now trails the CU5,725 baseline; re-scope, re-price or stop.
Mix remains 50/35/15Use the original CU5,725 result.A new ladder creates no modelled mix benefit without observed movement or another business reason.
Capacity is consumed by Best supportRecord support hours and the work displaced before assigning a value.Do not call the higher package profitable when the model omits its binding capacity cost.

Move from package design to a bounded rollout

Use this sequence

  • Name the customer job and make each package scope materially different and explainable. (not complete)
  • Reconcile price, variable cost, support and capacity on one package and period basis. (not complete)
  • Calculate contribution for every package before considering the customer mix. (not complete)
  • Label the mix as observed, proposed or downside; never present a proposed mix as demand evidence. (not complete)
  • Test cost, mix and capacity sensitivities and assign a stop condition. (not complete)
  • Run a bounded review, compare actual records on the same basis and revise the ladder or assumptions. (not complete)

Mistakes that invalidate the comparison

  • Adding cosmetic tiers whose cost and customer job are not materially different.
  • Leaving premium support, onboarding, refunds or fulfilment outside the Best-package cost.
  • Comparing monthly customer counts with annual or per-order costs.
  • Treating a provider-described benefit or an assumed upgrade path as a forecast.
  • Counting shared cost in every package and again at period level.

Use the registered calculation owner

Questions and limitations

Must a Good-Better-Best offer have exactly three packages?
No. Three is the scenario in this guide, not a universal rule. Use only the number of packages that can be distinguished, costed, delivered and reviewed clearly.
Does the package with the highest contribution always deserve promotion?
No. Customer fit, capacity, service quality, acquisition, refunds and the observed mix can change the decision. Compare the complete declared boundary.
Can the assumed mix be used as a sales forecast?
No. It is a sensitivity input until matched business records or a bounded market test provide evidence.

Sources and evidence scope

  • Choose a pricing strategy โ€” business.gov.au: Primary small-business guidance for reconciling cost, goals, market evidence, customer value and capacity. It supplies no universal margin, package mix, discount band or demand response.
  • Tiered pricing 101 โ€” Stripe: Provider and competitor context for common tier structures only. It does not support a guaranteed sales, conversion or margin outcome.

Change history

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