Report unavailable creator or post metrics as missing, with a reason and a coverage count beside each total. Keep confirmed zero, not applicable, not collected and unavailable separate. A report should tell the reader what you measured, which records you could not measure, and which decisions those gaps prevent.
Missing influencer campaign data needs a reporting rule before it needs another spreadsheet formula. The legend and completeness table below are recommended working conventions. Platforms do not share one universal set of missing-data labels.
Give each missing cell a meaning
Keep the numeric value and its status in separate columns. Leave the numeric field empty when you do not have a usable value. Use these labels in the reader-facing report:
| Display | Meaning | Treatment |
|---|---|---|
| 0 | A confirmed measured zero for this metric and period | Include in totals and the count of measured records |
| N/A | The metric does not apply to this record under the declared scope | Exclude from the eligible-record denominator |
| NC | The metric applies, but your team has not collected it | Count as a missing eligible record |
| UN | The metric applies, but the chosen source cannot supply it at the reporting cutoff | Count as a missing eligible record; record the limitation |
| A number | A usable value from the named source and period | Include, with source and capture time |
For example, a requested creator export that has not arrived is NC. A report that explicitly withholds a breakdown is UN. A photo deliverable has N/A in a column reserved for YouTube video watch time. These examples describe your report's treatment of the record.
Do not use N/A for an inconvenient gap. If a metric applies but access failed, it belongs in the missing count.
Also check what a displayed zero means. YouTube's limited-data documentation says that dashes, zeros or absent data can result from privacy or system limitations. A zero visible in an export is therefore insufficient evidence by itself. Check report notices and scope before classifying it as a measured zero. If the meaning remains unresolved, mark it UN and explain the uncertainty.
Establish the records you expected
Start with the publication roster, including posts whose metrics you lack. If you build the roster from successful analytics responses, missing posts disappear before you calculate completeness.
For each metric, define:
- The unit, such as one published video.
- Which publications qualify for that metric.
- The reporting dates and time zone.
- The source and collection method.
- The capture deadline and acceptable age of the data.
Keep planned but unpublished deliverables in a separate delivery count. An unpublished video should not silently become a published video with zero views.
The Government Data Quality Framework treats completeness as the presence of expected records and essential values. It also separates completeness from accuracy. A filled spreadsheet can still contain wrong figures.
Use a campaign data dictionary to agree on those definitions before people fill the report. Then choose a reporting window so late exports do not introduce a different measurement period.
Put a completeness table beside the results
This hypothetical campaign has ten published deliverables: eight YouTube videos and two photo posts on another platform. The table covers only the YouTube video metrics. The two photos are outside that scope.
A usable value must match the chosen period and definition. Confirmed zeros count as usable values.
| Metric | Eligible videos | Measured | NC | UN | N/A deliverables | Completeness |
|---|---|---|---|---|---|---|
| YouTube views | 8 | 5 | 2 | 1 | 2 | 62.5% |
| YouTube watch time | 8 | 4 | 3 | 1 | 2 | 50% |
Calculate completeness as measured eligible records divided by all eligible records. For views, that is 5 / 8 × 100 = 62.5%. Exclude the two N/A deliverables from the denominator. Keep NC and UN records in it.
Count coverage separately for each metric. Having a video's view count does not establish that you have its watch time. YouTube defines watch time in hours as the time viewers spent watching the video.
Post coverage also differs from creator coverage. If you show both, define a fully measured creator as one whose eligible posts all have the required metric for the period. A creator with three posts and one export is partly measured. Do not count that creator as complete.
Label totals and averages with their coverage
Suppose the five measured video view counts in the hypothetical campaign are 1,000, 2,000, 0, 1,500 and 500. Assume the zero has been confirmed as a measured count, with no limitation notice.
Their sum is 5,000. The mean among measured videos is 5,000 / 5 = 1,000. Report it this way:
Recorded views: 5,000 across 5 of 8 eligible videos. Two videos await collection; one has unavailable data. Mean among measured videos: 1,000 views. Full-campaign total and mean remain unknown.
Dividing 5,000 by all eight videos gives 625, which treats the three missing values as zero. Multiplying the measured mean by eight gives 8,000, which assumes the missing videos match the measured ones. Neither number belongs in the observed results.
Coverage of 62.5% of videos also does not mean coverage of 62.5% of campaign views. The missing videos may have different performance. Do not rank a partly measured creator as the lowest performer because their recorded subtotal is smaller.
If a rate needs two fields, calculate it only for records with both fields available and a valid denominator. Name that subset. Do not combine a numerator from five posts with a denominator from eight. The engagement-rate denominator guide helps specify that calculation before comparing creators.
Preserve platform totals without inventing missing rows
A platform-reported total and the sum of visible rows can differ legitimately. YouTube explains that limited rows can remain included in a report total.
Keep the platform total, its filters and its date range. Label your own sum as the subtotal of visible rows. Do not distribute the difference among creators or categories without supporting data.
Also preserve metric boundaries. YouTube excludes external websites and apps from registered thumbnail impressions. That exclusion is part of the metric definition. It is not a missing value you should fill with an estimate of external exposure.
If you remove filters or expand dates to investigate a gap, save that result separately. It answers a different query from the original report.
Give each recoverable gap an owner
Modash's reporting guide describes delayed screenshots and captures that omit required metrics. A collection request should therefore name the post, metric and reporting period, rather than ask for analytics in general.
Keep a gap log with the publication ID, metric, status, reason, last request date, owner and next review date. Preserve the original report cutoff when late data arrives. Issue a revised report with the changed values and coverage, rather than replacing the old figures without explanation.
Before sending the report, inspect every empty metric cell. Give it a status, recalculate the eligible denominator, and place the resulting coverage beside the number a reader will use to make a decision.
