Insights - Understanding Unique Listeners & Frequency

The Insights reporting tool offers additional reporting information not found in the built in Ad Stats reports in the AmperWave console.  Reach (Unique Listeners) and Frequency are some of the more useful stats available exclusively in Insights.

Frequency rates/goals set by the client can depend on and are affected by several factors; competing campaigns, stations the ad is targeting and their cadence of commercial breaks, audience targeting, geo targeting etc. An ad with broader targeting be it geo, station or audience segments will likely end up with a lower frequency as it can reach more people. Where the more narrow the targeting is the less people are eligible and the more often those eligible will likely hear the ad.

This guide aims to explain how Uniques and Frequency are calculated, how changing your reporting query can change those results and why you cannot SUM Uniques or Frequency.

The system records each listener with a unique fingerprint. This fingerprint tracks all their listening sessions and any ads they heard. Then the database uses a query built by the user in Insights to pull the exact information the report asks for (dates within a date range, impressions to the ad media id or campaign id, unique fingerprints in this case) and presents that data to you in a spreadsheet. How you build your report will determine the complexity of your results.

Let's explore some examples of Unique/Frequency reporting.

We have 3 listeners, Bob, Jane and Frank. We ask for several reports for the overall delivery on 1 campaign that contains 3 Ad lines, A, B and C.  How you structure your report will determine your results.

In this example we will run 

1. One report for the Campaign with no added Dimensions.
2. One report where we include the Ad TItle in the Dimensions.

Bob hears:  

• Ad line A, 2 times
• Ad line C, 1 time

 Bob's overall (no date grouping) reach is calculated as:

1. Campaign - 3 Impressions, 1 Unique listener, Frequency of 3
2. Ad Line A - 2 Impressions, 1 Unique listener, Frequency of 2
   Ad Line C - 1 Impression, 1 Unique listener, Frequency of 1

Jane hears:

•  Ad line A, 3 times
• Ad line B, 1 time
• Ad line C, 3 times

 Jane's overall (no date grouping) reach is calculated as:

1.  Campaign - 7 impressions, 1 unique listener, Frequency of 7
2. Ad Line A - 3 Impressions, 1 Unique listener, Frequency of 3
   Ad line B - 1 Impression, 1 Unique listener, Frequency of 1
   Ad Line C - 3 Impressions, 1 Unique listener, Frequency of  3

Frank hears:

•  Ad line B - 5 times.

 Frank's (no date grouping) reach is calculated as

1. Campaign - 5 Impressions, 1 Unique Listener, Frequency of 5
2. Ad Line B - 5 Impressions, 1 Unique Listener, Frequency of 5

Since these examples are all just one person the uniques look to add up from the ad line level to the campaign level. But we don't report on just one person at a time, we report on Bob, Jane and Frank together.

So in that case let's add those up so they are all counted in your reports 

If we report by Campaign as a whole. (No additional Dimensions, Filter on Campaign ID)

• Campaign - 15 impressions (3-Bob, 7-Jane, 5-Frank), 3 unique listeners, Frequency of 5

If we report the same for the ad lines broken out (add Ad Title to your Dimensions, Filter on Campaign ID, OR all 3 Media IDs)

• Ad line A - 5 impressions (2-Bob, 3-Jane), 2 unique listeners, Frequency of 2.5
• Ad line B - 6 impressions (1-Jane, 5-Frank), 2 unique listeners, Frequency of 3
• Ad line C - 4 impressions (1-Bob, 3-Jane), 2 unique listeners, Frequency of 2

If we report for Lines A and C, maybe they targeted the same Geos, but don't break them out (don't add Ad Title to your Dimensions, FIlter on the 2 Media IDs) 

• Ad Lines A,C - 9 Impressions (3-Bob, 6-Jane), 2 unique listeners, Frequency of 4.5

Now Ad lines B and C but don't break them out (don't add Ad Title to your Dimensions, FIlter on the 2 Media IDs) 

• Ad Lines B,C - 10 Impressions (1-Bob, 4-Jane, 5-Frank), 3 unique listeners, Frequency of 3.3

You can see how 3 users' listening experiences are recorded and how it creates a different total for each way a report is run.

If you then change the Date grouping to be by Month, Week or Day instead of the overall date range the number will then change again. Jane could have been listening only one week during the flight, where Bob could listen every week and so the numbers would then move around. 

Example : Bob listens to the radio one time in January, from 10pm to 2am on 1/1/23 into 1/2/23.
We run a report for the first week of January.
If we choose to group by Day, Bob will count as 2 Unique listeners (one listener on 1/1 and one listener on 1/2) because his fingerprint was seen on both days.
If we choose to group by Week, Bob will only count as 1 Unique listener for 1/1 through 1/7 because his fingerprint was seen just one time in that time period.

Another example: Monica listens to the radio 3 times in March, on the 3rd, 4th and 31st.
We run a report for the month of March.
If we choose to group by Day, Monica will count as a unique listener for each of those days, adding 3 uniques to the total.
If we choose to group by Week, Monica will count as a unique listener for 2 of the weeks (first and last week) adding 2 uniques to your total.
If we choose to group by Month, Monica will count as a unique listener for the month, adding 1 unique to your total.

Now consider adding other Dimensions like Station or Geo locations and the numbers will spread out even further. Or add more Ad lines to the Campaign and again, you're going to have numbers that will be different each way you build your report.

For simplicity's sake it is best to choose what way you want to report to a client, and stick to that method.

• The most popular By Date Groupings are either; Overall (none), or by Month.
• If there is only one ad line in the campaign the line's totals and the campaign's totals will be the same, so often the report pulled is just for the Campaign's totals.
•  If there are multiple lines, or there are differences in targeting by lines it can be useful to do as we did above and show the client both the Campaign's reach and the reach of the different lines (1,2,3,4 etc.) or the different targeting segments' lines together (1-6, and 8-13 etc).  This shows the client how well each line or targeting group did, and then how well the campaign did as a whole.  

If you need to explain to the client when asked "why don't the uniques/frequency for the ad lines add up to equal the Campaign's unique/frequency?" let them know one listener can hear more than one ad within the campaign, so the uniques/frequency will not add together. 

If you have any questions about Uniques and Frequency please reach out to AmperWave Support.