# Event ad creative test plan and results log

Copy this sheet for each experiment. It is an organizer worksheet, not a platform experiment or a statistical guarantee. Keep the original record after deciding what to run next.

## 1. Decision to learn

| Field | Entry |
| --- | --- |
| Event, market, ticket/registration window | |
| Campaign and platform | |
| Owner and approver | |
| Audience and offer held constant | |
| Hypothesis: if we change **one creative element**, then **which attendee action** may change, because **why**? | |
| Control: approved headline/image/video/CTA | |
| Variant: exactly what changes | |
| Creative files and approval links | |
| Landing page and ticket destination (check both variants) | |
| Ticket availability and price shown on the page | |

## 2. Comparison and measurement

| Field | Entry |
| --- | --- |
| Platform-supported experiment or variation method | |
| Allocation setting and what is actually comparable | |
| Planned start/end or stop event (ticket sellout, deadline, policy issue) | |
| Approved spend ceiling and pause owner | |
| Primary outcome chosen **before** launch | |
| Measurement event and where it is recorded | |
| Guardrails: spend, registration quality, complaints, accessibility, policy | |
| Minimum evidence needed for a decision; what makes result inconclusive | |
| Tracking/consent and hosted-checkout boundary | |

Prefer completed registrations or purchases when the event and measurement volume make that defensible. If using landing-page views or clicks, label them diagnostic; they are not ticket sales. A before/after comparison is exploratory unless delivery and other changes are controlled. Platform traffic splits may still produce unequal impressions and spend.

## 3. Launch gate

- [ ] Both ads accurately state date, place, price/offer and availability.
- [ ] The only deliberate creative difference is written above.
- [ ] Both destinations and registration/checkout handoffs work for a public visitor.
- [ ] Audience, placement, bid, budget and schedule are as comparable as the platform permits.
- [ ] Primary outcome and guardrails are recorded before results are viewed.
- [ ] Creative, destination, tracking and spend have named approvers.

## 4. Results and decision

| Observation | Control | Variant | Caveat / evidence link |
| --- | --- | --- | --- |
| Delivery period and material changes | | | |
| Impressions and reach, if available | | | |
| Spend | | | |
| Clicks or landing-page views | | | |
| Completed registrations/purchases, if observable | | | |
| Primary outcome rate or cost using the chosen denominator | | | |
| Guardrail result | | | |

**Decision:** adopt / retest / revise / inconclusive / stop.

**Reason, including allocation imbalance, low volume, tracking gaps and ticket-window changes:**

**Next creative question, owner, date:**

Do not call a variant a winner merely because its click count is higher; check the preselected outcome, delivery and evidence limits. If the event sells out or the page changes during the run, record that change rather than hiding it.