Household Reach
Setting practical frequency expectations when identity is incomplete
Incomplete identity makes a strict “one household, one cap” promise unreliable.
When identity is incomplete, set a cap as a limit within a defined measurement environment—not as a promise of one household’s total exposure. Write down three things: the time window, the identity key being counted, and the inventory included.
The counted identity might be a signed-in account, a device ID or a request group. Some server-side podcast measurement may use an IP address and user agent; controlled players may provide a signed-in account or device identifier. The available signals vary, so a signed-out viewer or someone switching devices may not be joined to the same count, while a shared television may represent more than one person.
Name the cap level and included inventory: for example, a campaign on one buying platform, or the publishers participating in a buy. State whether the rule applies at advertiser, campaign, flight or ad level. Do not imply that delivery outside the named platform or inventory is deduplicated.
A concrete example is a cap of 3 impressions per recognised user per day. A seven-day window is another usable planning choice, but it needs its own stated cap; the window alone sets no limit. Treat these as configuration examples, not industry benchmarks, and remember that a cap is a ceiling rather than an expected average.
Read the platform’s reported reach and frequency distributions after its normal processing delay. Record the chosen identity key and window alongside the results, and distinguish impressions matched to that key from those that could not be matched. If identity coverage is incomplete, measured frequency may understate repeat exposure; without match-rate information, a blanket uplift cannot be calculated.
Use the measured subset to make adjustments within that same subset, and retain creative rotation and publisher-level checks. Document which inventory remains outside the count and revisit that coverage before expanding spend.
Setting Practical Frequency Caps with Incomplete Identity
- Identify the identity keye.g., signed-in account, device ID, IP address and user agent
- Specify included inventorye.g., one buying platform or participating publishers
- Document unmatched impressionsTrack exposure not matched to an identity key
Key Considerations for Frequency Measurement in Australia
- Identity coverage limitation
- Incomplete identity may understate repeat exposure
- No universal benchmark
- Caps are configuration limits, not industry standards
- Platform-specific rules
- Frequency caps apply within defined platforms and inventories only
Pros and Cons of Using Identity-Based Frequency Caps
- ProsEnables targeted frequency control; supports better creative rotation and publisher-level checks
- ConsIncomplete identity coverage leads to undercounted repeat exposure; no reliable uplift can be applied without match-rate data
