1. Frequently Asked Questions

Aggregate Reach Totals vs. Line Item Reach Totals and Why They Do Not Match

There are 3 factors driving the sum of individual line item reach not matching aggregate reach:

  1. Reach goes through a deduplication process when aggregated
    1. A user downloads an episode containing Ad 1 and Ad 2 or a user downloads multiple episodes - Ad 1 in Episode 1 and Ad 2 in Episode 2.
    2. Each of those ads "reached" one household and it counts as one reach per line item in the table view.
    3. In the aggregated reach number for all line items in a campaign, Ad 1 and Ad 2 both "reached" the same household so this counts as only 1 household reached for a campaign and therefore, it doesn't sum the reach from both ads.
    4. Summary
      1. The table view showing individual line items in a campaign displays distinct households reached per ad/line item
      2. Aggregated reach is the total number of distinct households reached across all ads in a campaign
  2. There are different data sources for aggregated vs. line item breakdown
    1. Aggregated data is pulled from Big Query and Line Item breakdown is pulled/calculated from a combination of Spanner, BQ, Postgres
    2. Big Query is populated once per day and Spanner is populated hourly causing some lag in data available to Big Query
  3. Reach for a line item is an approximate calculation using impressions and frequency because we don't calculate reach on an hourly basis like we do for impressions.
    1. Total Impressions and Total Reach (over the campaign) for a line item are used to generate a frequency value
      1. Frequency = Impressions / Reach
    2. This frequency value is used to generate an estimated reach value for a time interval using an impressions count over that interval divided by frequency
      1. Reach = Impressions / Frequency
    3. The frequency for a particular line item will be different than the aggregate frequency of all line items across a campaign
    4. Summary
      1. Therefore, the calculation will not produce consistent values across different levels of granularity (overall campaign vs 1 line item). The sum of line item reach could be less than or greater than overall campaign reach depending on how far the time frame impressions vary from the average.