Fix the comparison boundary before calculating
Records and assumptions to align
- Use one currency, market, tax basis, plan and accessed/effective date for provider terms. (not complete)
- Define the same checkout, hosting, email/community, learner support and analytics boundary. (not complete)
- Reconcile fixed fees, per-transaction fees, payment fees and feature add-ons without duplication. (not complete)
- Value self-hosted setup, maintenance, support and incident time at an explicit loaded rate. (not complete)
- Test migration and step costs separately from recurring per-learner cost. (not complete)
| Field | Required unit | Preferred evidence |
|---|---|---|
| N | learners/period | completed cohort or labelled scenario |
| fixed platform/stack cost | CU/period | current provider contract/statement |
| loaded maintenance/administration | CU/period | time record × explicit loaded rate |
Build a reproducible scenario
total cost(N) = fixed platform/stack cost + N × per-learner transaction/payment/support cost + loaded maintenance/administration + named step costs
- N
- Paid learners in the matched period/cohort (learners/period) — completed cohort or labelled scenario
- fixed platform/stack cost
- Recurring fixed tools needed for the declared capability boundary (CU/period) — current provider contract/statement
- loaded maintenance/administration
- Internal/contractor time required to operate the selected model (CU/period) — time record × explicit loaded rate
Do not use a current provider rate outside its market, currency, plan or fee base. Treat volume as a sensitivity and keep migration cost outside recurring cost unless amortisation is declared.
Fictional 200-learner monthly comparison
No real provider is represented; fees and time are invented inputs.
| Case | Declared inputs | Substitution | Result |
|---|---|---|---|
| Managed platform | 300 fixed; 5 CU/learner fees/support; 200 learners | 300 + 200 × 5 | 1,300 CU/month |
| Self-hosted stack | 650 fixed stack; 1.50 CU/learner; 30 hours × 25 CU maintenance | 650 + 200 × 1.50 + 750 | 1,700 CU/month |
| Self-hosted at 500 learners | 650 fixed; 1.50 CU/learner; 750 maintenance | 650 + 500 × 1.50 + 750 | 2,150 CU/month |
| Platform at 500 learners | 300 fixed; 5 CU/learner | 300 + 500 × 5 | 2,800 CU/month |
Stress-test the uncertain inputs
| Variable | Lower case | Higher case | What it tests |
|---|---|---|---|
| Paid learners | Completed lower cohort | Capacity-limited scenario | Tests fixed versus variable cost |
| Maintenance/support hours | Stable observed case | Incident/high-support case | Tests hidden self-hosting labour |
| Provider fee base/plan | Current eligible terms | Named step/add-on case | Tests contract-specific cost steps |
Stop and review when
- Provider terms lack market, plan, currency, fee base or date. (not complete)
- The compared capability/service boundaries differ. (not complete)
- Self-hosted labour, reliability or migration work is omitted. (not complete)
Turn the scenario into a controlled decision
- Reconcile the baseline to current records and name the evidence owner.
- Run the base case, then change one uncertain input at a time.
- Record the chosen response, approval limit, review date and stop trigger.
- Compare actual results with the original boundary before reusing the assumption.
Avoid these mistakes
- Comparing a platform percentage with server hosting alone. (not complete)
- Copying a provider price into a global default. (not complete)
- Treating projected enrolment as evidence for the cheaper case. (not complete)
Questions to resolve before approval
- Is self-hosting cheaper at high volume?
- Not automatically. Fixed stack, maintenance, support, payment, reliability and step costs determine the current scenario.
- Does the model value data control or uptime?
- No. Score those criteria explicitly with evidence rather than hiding speculative values inside the cost arithmetic.
Methodology and source boundary
- Platform Plan Break-even methodology — Margin101: Product-owned fixed/variable plan crossover boundary.
- Payment Processing Comparison methodology — Margin101: Product-owned user-entered percentage/fixed fee boundary.
- Make vs Buy methodology — Margin101: Product-owned internal time, fixed and external-cost comparison.