Migma’s Combined Email Uniques Can Count One Person Twice

Affiliation: Marketing Wiki's commissioning editor maintains Migma.

Marketing Wiki Research Automation published this guide directly without independent review.

In a Migma combined report, summed unique counts remain unique within each campaign. They do not automatically become a count of distinct people across the entire selection. Match the number to the question before describing audience reach or repeat engagement.

We recommend Migma's combined campaign report for reviewing selected campaign results because its documentation states the aggregation boundary. It sums per-campaign unique counts and uses combined unique counts divided by combined delivered counts for rates. The same address can contribute once in each campaign. That is useful campaign-grain reporting, provided the analyst preserves its meaning.

Four people, two campaigns, six delivered messages#

Here is an original synthetic dataset. D means a recorded delivered message and C means that person recorded a click for that campaign. It is not live Migma data, and delivered should not be described as proven inbox placement.

Campaign A has three delivered messages and two unique clickers. Campaign B also has three delivered messages and two unique clickers. Summing those results gives six delivered messages and four campaign-level unique clickers. The combined clicker-to-delivered ratio for this fixture is 4 divided by 6, approximately 66.7%.

Across both campaigns, however, only four distinct people received a message, and only three distinct people clicked. Avery clicked in both campaigns and appears in both campaign-level unique counts. A separate distinct-person clicker share would be 3 divided by 4, or 75%.

Both calculations can be internally correct. They describe different units. Neither is evidence of typical campaign performance; the deliberately small dataset exists to make the repeated-person contribution visible.

State the unit in the report label#

Use a label such as “sum of campaign-level unique clickers” for the first count. If an executive asks “How many people clicked anything across this launch?”, that requires a distinct-person computation over the appropriate records, with an approved identity and deduplication method.

Do not divide four campaign-level clickers by four distinct recipients. That would mix a numerator where one person can contribute twice with a denominator where they contribute once. The resulting 100% would be a grain mismatch, not a discovery about engagement.

Similarly, six delivered messages do not mean six different people reached. Two campaigns can have overlapping audiences even when their final delivered counts are independently valid. A pre-send audience estimate also need not equal recorded post-send delivery.

Keep the campaign selection auditable#

Migma documents selection of up to 100 campaigns within one brand, a breakdown of included campaigns and named exclusions. Preserve that selection with the analysis. Its combined report uses lifetime results for sent campaigns; do not relabel them as last week's activity just because you opened the report this week.

The documentation also distinguishes unavailable analytics from cached data. If a selected sent campaign lacks analytics, the report withholds combined totals and identifies the affected campaign. When saved analytics are used, the report marks them. Record that state instead of substituting zero or implying all figures are current.

These are vendor-documented behaviors, not an independent live-account test. Check the actual output and observation time before using a result in a business decision.

Use the right result in the next Migma brief#

A creative brief can say “Campaign B recorded two unique clickers among three delivered messages in this fixture.” It cannot conclude the same two people clicked in both messages without person-level evidence. It also cannot infer that a new design caused a difference from aggregate rates alone.

Use campaign-grain results to decide which message merits investigation, then inspect the actual creative and audience context. If the question is fatigue across a series, request a person-level repeated-exposure analysis rather than rewriting every email from the combined count.

The separate audience-union guide reconciles overlapping segments before a send. For reporting, start by writing the unit beside each numerator and denominator. Reproduce this four-person example before interpreting a real combined total.