MVA / FIELD GUIDES / Marketing reporting
When a tracking repair changes the conversion chart.
A jump in recorded conversions may mean the measurement improved. The report needs that context before the change is credited to marketing.
The collection changed. The interpretation should too.
Suppose a website was accepting inquiries that its analytics tracking missed. A repair starts recording more of them. The conversion line rises, even if the repair had no effect on how many people contacted the business.
A repair can also remove duplicate events and bring a count down. Either change needs an explanation before it is described as a change in marketing performance.
An example with the context kept in view.
A change in the measurement.
Tracking repaired in week four. The solid line shows recorded analytics requests; the dashed line shows accepted requests.
The explanation belongs beside the chart. Here, part of the increase follows a repair in collection.
View illustrative data and explanation
| Week | Accepted requests | Analytics requests |
|---|---|---|
| 1 | 38 | 22 |
| 2 | 45 | 28 |
| 3 | 50 | 35 |
| 4 | 58 | 58 |
| 5 | 66 | 66 |
| 6 | 76 | 76 |
Counts agree from week four in this illustration. This comparison alone does not validate every event or resolve attribution.
The accepted-request series is a reference in this fictional example. For a client report, the comparison needs a suitable business record: the same action, period, and set of forms. Matching totals alone do not establish that every form works correctly or that channel attribution is complete.
What belongs in the change note.
ILLUSTRATIVE MEASUREMENT CHANGE NOTE
- When
- Week four in the example above.
- What changed
- Tracking was repaired to record the accepted inquiries in scope.
- What it affects
- Comparisons of recorded requests before and after the repair.
- How it was checked
- The form submissions and their recorded results were reviewed.
- What to avoid
- Describing the full increase as improved marketing performance.
For the actual report, include the implementation date, affected events or metrics, how the change was checked, and anything still unresolved. Put the short explanation beside the chart so readers can understand the result without opening a separate technical document.
Choose a comparison the measurement can support.
If the old tracking missed events, a before-and-after percentage can combine marketing changes with tracking changes. Explain that break in the series. Use a sufficiently comparable business record when one is available, or assess a later period measured under the same setup.
Historical counts may not be recoverable. If an estimate is useful, show its method, assumptions, and uncertainty, and label it as an estimate.
A repair addresses the tracking problem in scope. Other differences may come from definitions, reporting periods, filters, consent choices, or attribution. The report reconciliation worksheet helps separate those questions.
The chart is the visible part of the work.
A useful reporting handoff includes metric definitions, an explanation of measurement changes, and context for the comparisons the team will use.
If the reason for a change is unclear, an analytics audit can investigate it. An understood problem may lead straight to tracking repair. When the measurement is usable and the question is interpretation, the starting point may be reporting and analysis.