Start with a dated cohort table
The dataset below is hypothetical. It is not a Nanocreator World customer story, benchmark, or promised result. Assume one campaign cohort, a defined 30-day reporting window, and rewards that may remain valid after the cutoff.
| Measure | Example | Interpretation |
|---|---|---|
| Joins | 40 | People in this cohort |
| Submitting participants | 24 | 60% of joiners |
| Approvals | 20 | 83.3% of submitters |
| Pending reviews | 2 | An unresolved operating task |
| Corrections requested | 2 | Still an opportunity to complete |
| Redemptions to date | 12 | 60% of approvals so far |
| Staff time | 3 hours at $20 | $60 modeled labor cost |
| Reward cost to date | 12 × $1.50 | $18 fulfillment cost |
| Other cost | $10 | For example, printing |
Calculate both observed cost and open obligations
Modeled cost to date is $60 + $18 + $10 = $88, or $4.40 per approval across 20 approvals. Eight approved rewards remain unredeemed. If all eight are still valid and redeemed at $1.50 each, they add $12, taking modeled cost to $100 before any additional labor.
The four submitted participants awaiting a final outcome could create more obligations. Report them separately instead of pretending the cohort is finished. Also distinguish cash outlay from the assumed value of owner time.
Use a repeat, revise, or pause framework
Repeat when the process is understood, staff can manage it, and the cost fits the goal. Revise when a particular step creates avoidable friction. Pause new recruitment when the business cannot review or fulfill its promises. These are management judgments, not universal performance thresholds.
| Observation | Next decision |
|---|---|
| Repeated questions about proof | Revise the evidence instructions |
| Pending reviews beyond the promised response | Repair staffing before recruiting more |
| Reward is frequently unavailable | Pause new promises and resolve existing ones |
| Smooth workflow, manageable cost, useful participation | Repeat with a defined learning question |
| Owner wants to know sales impact | Design additional purchase measurement; do not infer it from approvals |
Copy this monthly decision memo
Write one paragraph for each prompt: what we intended to learn; cohort dates and cutoff; observed participation counts; costs and outstanding obligations; customer and staff feedback; biggest uncertainty; decision and reason; one change; next review date.
For the hypothetical dataset, a defensible decision might be: “Resolve the four open submissions, clarify the most common proof question, and repeat the same offer with a comparable group.” It would not be defensible to announce a sales return from these numbers alone.
Follow people long enough to interpret the outcome
Keep late redemptions attached to the original cohort, and give new cohorts a comparable follow-up window. If recruitment source, reward, and action all change together, acknowledge that you cannot isolate which change affected participation.
Save the dated memo before beginning the next campaign. Over several cycles, these records can reveal recurring operating patterns. A publishable case study still needs a real business, permission, verified records, and an honest account of what did not work.
Model your outstanding reward costs.
Use your approved-participant count and fulfillment costs to estimate the fully redeemed scenario.
Check the reward obligationPlanning your spend? Use the reward-budget calculator. Ready to see the product? Walk through the campaign flow.