How to use the scorecard
Score each answer from 0 to 3: 0 for no answer, 1 for a claim without evidence, 2 for evidence with gaps, 3 for evidence with definitions, dates and a demonstration. Weight each section for your program shape; a suggested default is shown. Require every number to carry a definition, a scope and an as-of date, or score it as a claim.
Section 1: Move Fast (default weight 15%)
| # | Question | Evidence to require |
|---|---|---|
| 1 | Show a marketer taking a new promotion from idea to production. What waits on engineering? | Live session; list of change types that are self-service versus engineering |
| 2 | How are segments, promotions, challenges and tiers defined? Can a segment built for a promotion be reused in a challenge? | One eligibility framework demonstrated across mechanics |
| 3 | How is a change tested against program history and costed before publication? | Pre-launch simulation with projected cost and affected members |
| 4 | What does the AI layer do, and what does it not do? | Named workflows; statement of what requires human approval |
| 5 | How do engineers extend the platform when a requirement exceeds configuration? | Programmable core with versioning and audit; example of a client-built mechanic |
Section 2: Change Safely (default weight 15%)
| # | Question | Evidence to require |
|---|---|---|
| 6 | Which environments exist and how are they kept at the same release level? | Environment model; promotion path from stage to production |
| 7 | How is configuration versioned, exported and reviewed before publication? | Versioned, exportable configuration; approval workflow |
| 8 | What is written to the audit trail for a rule change, a rule execution and an approval? | Audit entries shown for each |
| 9 | How is a change rolled back, and how long does it take? | Rollback demonstrated |
| 10 | How are access rights scoped by role, brand, region and folder? | Role- and attribute-based access shown |
Section 3: Operate with Certainty (default weight 20%)
| # | Question | Evidence to require |
|---|---|---|
| 11 | What is the response-time commitment at the transaction, and how is it measured? | SLA with percentile and measurement method |
| 12 | What is the production availability record, from when, under what definition of downtime? | Start date, definition, exclusions, as-of date |
| 13 | What annual transaction volume and member base run in production today? | Governed figures with definitions and dates; references at comparable scale |
| 14 | How does the platform behave at the register or pump when it cannot answer in time? | Degradation behaviour documented |
| 15 | How are reversals, adjustments and account merges handled, and what unwinds with them? | Reversal demonstrated with tier, challenge and funded-discount effects |
Section 4: Migration (default weight 20%)
| # | Question | Evidence to require |
|---|---|---|
| 16 | Describe the migration method step by step. Is the new platform run against production traffic before cutover? | Parallel-run method documented |
| 17 | On what share of the member base are balances, tiers and outcomes reconciled, and what happens to differences? | Full-base reconciliation; discrepancy handling |
| 18 | How are historical transactions, expiry schedules, partner balances and open settlements carried? | Per-asset migration approach |
| 19 | How is legacy rule behaviour audited against documented program terms? | Rule audit method |
| 20 | Does rollback remain available after cutover, and for how long? | Rollback policy; migration references with downtime record |
Section 5: Financial Integrity (default weight 15%)
| # | Question | Evidence to require |
|---|---|---|
| 21 | Does every point carry its originating rule, currency, brand, partner and expiry? | Point-level lineage shown |
| 22 | How is burn order controlled, and how is a redemption attributed to the party that funded it? | Burn-order configuration; attribution report |
| 23 | How are partner-funded promotions, wallets and settlement handled when rates change after the fact? | Date-relative billing demonstrated |
| 24 | Can finance reproduce the liability rollforward from transaction-level exports? | Prior liability + earned − redeemed − expired = outstanding, from exports |
| 25 | How is a mixed-brand basket accounted for? | Line-item resolution to each brand's ledger |
Section 6: Security and Assurance (default weight 15%)
| # | Question | Evidence to require |
|---|---|---|
| 26 | Which independent attestations and certifications are held, what do they cover, who issued them, and when do they expire? | Reports and certificates with scope and dates |
| 27 | How is personal data handled, and can the platform run on tokenized identifiers only? | Privacy modes documented |
| 28 | How is access authenticated for staff and for systems? | SSO via federation; API authentication model |
| 29 | What is the availability architecture and the tested recovery process? | Multi-zone design; business continuity and disaster recovery evidence |
| 30 | Who owns the outcome when a problem sits between the platform and an integration? | Delivery and operating model; escalation path |
Scoring sheet
| Section | Questions | Default weight | Raise the weight when… |
|---|---|---|---|
| Move Fast | 1–5 | 15% | Marketing calendar drives the program |
| Change Safely | 6–10 | 15% | Many teams or brands configure the same instance |
| Operate with Certainty | 11–15 | 20% | Transactions happen at a physical lane |
| Migration | 16–20 | 20% | A live program with balances is moving |
| Financial Integrity | 21–25 | 15% | Partners fund rewards or several brands share a program |
| Security and Assurance | 26–30 | 15% | Regulated industry or in-country data requirements |
Questions per promise