Household Reach
Part of CTV campaign reviews
Reporting CTV reach when household matching is incomplete
Report measured household reach alongside unmatched CTV delivery, with clear units, inventory coverage and no invented whole-campaign total.
When a household-reach method covers only part of a CTV buy, report its estimate for the inventory and period it covers. Show delivery outside or unresolved by that method separately. Dividing measured reach by a match rate does not yield a valid whole-campaign household total: unlinked impressions may include households already counted and households the method has not identified.
Identify the coverage gap
Ask the provider what event enters its household-reach method, which identifiers and inventory are eligible, and what cannot be included. Distinguish an impression without a usable household association from a reach estimate withheld because the model lacks enough data. Ask what the provider means by a match rate, and name its denominator; uploaded records, eligible impressions, devices and households are different units.
Display & Video 360's help documentation covers Household Reach and separately covers Reach reports. Do not treat a household-level or user-level metric as a count of named people observed in a room. Confirm whether those reports are available for the actual Australian buy, and which inventory the returned report covers.
Put delivery and estimates in separate rows
Use one campaign period and impression definition for the delivery rows. Give each reach figure its own unit, method and scope.
| Report line | What to state |
|---|---|
| Total eligible CTV delivery | Impressions, dates, TV-screen rule and included inventory |
| Delivery eligible for the household method | Count and eligibility rule, if supplied |
| Delivery outside or unresolved by that method | Count and reason where available; otherwise state that the count is unknown |
| Estimated household reach | Estimate, covered inventory and dates, and any reported uncertainty or suppression |
| Other reach estimates | Device or modelled-person figures in separately labelled rows |
Do not equate measurement eligibility with successful matching unless the provider defines them as the same stage. Do not add household estimates from separate publishers or sum reach from separate report cuts unless a combined method accounts for overlap.
Check the timing and suppression rules of the actual household report separately, and do not treat a blank figure as zero.
If a provider relies on IP-derived associations, ask what other signals and validation support the method.
Write the conclusion at the measured level
State the estimated household reach with its covered inventory and dates, then state what delivery sits outside or remains unresolved by that method. Keep a person or device estimate under its own label. If the gap could change the campaign decision, seek a documented combined estimate or a better measurement design for the next flight. Until then, leave campaign-wide household reach unknown.
