# Example: A Missing Day Is Not Zero Spend

**Synthetic illustration, not a customer case or an official platform export.** No advertising account was queried, no ad was changed, and these numbers must not be used as product performance evidence.

| Scope | Value |
| --- | --- |
| Account | A synthetic example, not a provider account |
| Dates | 1–3 January 2026, completed source-account days |
| Reporting calendar | Asia/Hong_Kong; not re-bucketed to UTC |
| Currency | USD |
| Data coverage | Two observed dates; one missing date |
| Result | Partial; full-period reconciliation is blocked |

| Date | Synthetic reference spend | Observed spend | Reference clicks | Observed clicks |
| --- | --- | --- | --- | --- |
| 1 January | $40 | $40 | 20 | 20 |
| 2 January | $60 | $60 | 30 | 30 |
| 3 January | $50 | Missing | 25 | Missing |

The two observed days contain $100 and 50 clicks. Those are a covered-date subtotal, not a complete three-day total. The synthetic reference contains $150 and 75 clicks across the requested period. The period's actual spend and clicks remain unknown until the missing day has been checked.

There is no evidence here that spend fell to zero, the account stopped running, or the result needs a budget change. A reporting gap does not establish a campaign diagnosis. Do not generate full-period percentages or turn the missing date into a zero row.

Next, check the original task's source coverage and independent platform result for the same reporting calendar, filters and attribution definition. Keep the original result and task reference; disclose any collection needed to fill the gap instead of automatically querying the advertising platform. Only after comparable evidence is complete should the period be reconciled.

In a real session, the team would show the participant how to return to that saved result and how to stop access. This example grants no access, records no activation, and does not demonstrate a particular client's App UI.

Source for this illustration: `synthetic-input.json`. The accompanying machine result is generated by the existing GravityAds reconciliation contract and intentionally remains blocked for a complete-period conclusion.