Reach & Frequency
Part of Reach and frequency planning
Estimating audience duplication across channels
Estimate cross-channel audience overlap without double-counting reach, and show how uncertain duplication changes the combined figure.
Audience duplication counts the people reached by more than one channel in the same period. Align the audience, dates and reach threshold, then obtain comparable channel reaches and a credible overlap estimate. Where overlap cannot be measured, use labelled scenarios. Adding channel reach figures yields combined unique reach only when there is no overlap.
Establish one boundary
Define the eligible people, locations, campaign dates and minimum exposure level. Check whether each supplier counts people, households, accounts or devices, and whether its figure describes the proposed ad buy rather than everyone using the service. Do not insert unlike units into a people-level duplication calculation.
For two channels with comparable people-level figures:
Combined unique reach = Channel A reach + Channel B reach − people reached by both.
Suppliers may deduplicate within their own systems without a shared way to identify the same people across systems. A panel, matching method or cross-media model can estimate overlap, but you must state the inventory, population, period and uncertainty it covers.
Show what uncertain overlap changes
Suppose two hypothetical channels reach 6,000 and 5,000 people in the same defined population and period. If 1,500 people are reached by both, combined reach is 9,500. If 4,000 overlap, it is 7,000. These are sensitivity scenarios, not findings about an Australian campaign.
| Assumed overlap | Combined unique reach | People reached by B but not A |
|---|---|---|
| 1,500 | 9,500 | 3,500 |
| 4,000 | 7,000 | 1,000 |
For these channel totals, overlap cannot exceed B’s 5,000 people. The actual eligible population may further constrain a plausible scenario. The last column describes B’s additional people relative to A; it does not establish whether exposure in both channels was useful.
Check what a combined estimate covers
Ask which campaigns and placements are included, whether audience and dates match, and how cross-device use and people watching together are handled. Establish whether the figure concerns forecast ads, delivered ads or general content use. Name excluded channels and unmeasured exposure.
Check the documentation for Google’s Cross-Media report to establish what channels, audience and overlap measure it covers; do not assume it applies to an Australian advertiser or arbitrary supplier pairs.
Without suitable overlap evidence, retain each channel’s reach separately and present the scenarios with their assumptions. Do not call a scenario midpoint measured reach. If the decision changes substantially across plausible assumptions, investigate overlap before claiming that one channel adds a specific number of people.
After delivery, revisit the estimate using the dates and placements that actually ran. Changed spend, inventory or timing can make the original overlap assumption less relevant.
