Co-viewing on CTV: why one impression is worth more than one
Mobile advertising reaches a person; Connected TV advertising reaches a room. Co-viewing — several people watching the same screen — is the quiet structural advantage of CTV, and it changes how you should count reach, choose audiences and write creative. This article explains the mechanic and its commercial consequences.
The mechanic: one screen, several viewers
Television has always been a shared medium, and streaming did not change that — it moved the shared screen from the cable box to the smart TV. Industry measurement convention counts a delivered CTV impression as one impression, while the actual audience in front of the screen is frequently two or more people. The gap between counted impressions and actual viewers is the co-viewing factor.
For an advertiser this means your real reach is systematically understated by the numbers in your dashboard. A campaign that delivered a million impressions reached more than a million people — how many more depends on content type, time of day and household composition, but the direction of the error is always in your favor.
What co-viewing does to your reach math
Media planning on CTV starts from household-level numbers, not person-level ones. Impressions bought translate into households reached, and households reached translate into persons reached at a multiplier greater than one. When you compare CTV's CPM against mobile's, the honest comparison divides the CTV price by the co-viewing factor before comparing cost per person reached.
The same logic applies to frequency: a frequency of three on TV means three household exposures, each potentially seen by several members. Effective per-person frequency is higher than reported, which is one reason CTV campaigns achieve recall at impression counts that would underdeliver on mobile.
Household-level targeting and attribution
CTV targeting operates at the household level: geography, household composition signals, viewing behavior, and matching against advertiser first-party data at address or device-graph level. This is a different targeting grammar than mobile's per-device IDs — coarser in one dimension, richer in another, because it captures the unit where many purchase decisions actually happen.
Attribution follows the same shape: exposure logged at the household is matched to actions on any device in that household — the phone install, the tablet purchase, the laptop site visit. Cross-device household graphs, maintained by MMPs and measurement vendors, are what turn CTV from a faith-based brand channel into a measurable performance input.
The categories that win from co-viewing
Co-viewing matters most where the viewing group overlaps the buying group. Entertainment and gaming apps benefit when several household members see the same ad and install together. Food delivery, streaming, travel and family finance products are decided in front of the TV as often as anywhere. Even single-user categories gain: a second viewer in the room is a second chance at recall, and word-of-mouth inside a household is the oldest retargeting mechanic there is.
The flip side: hyper-personal response formats — a QR offer meant for one specific user — should be designed knowing others are watching. What plays well to a room plays differently than what plays to a pair of headphones.
Creative consequences of a shared screen
Shared viewing raises the creative bar in specific ways. The message must work for the least interested person in the room, not just the target one. Humor and spectacle outperform intimate, personal tones. Sound-on is the default, so audio carries meaning that mobile creative usually buries in captions. And the call to action must be room-appropriate: a QR code with a clear spoken and visual reason to scan works; a wall of fine print does not.
Practical rule: watch your creative on an actual TV with one other person in the room before you ship it. The reactions you get in that ten-second exercise predict performance better than any panel score.
Measuring a household medium honestly
Because dashboards count impressions and households rather than people, CTV measurement should focus on outcomes: incremental installs in exposed households versus holdouts, branded search lift in exposed regions, site-visit lift on household-matched devices. Person-level last-touch attribution will undercount CTV's contribution by design; incrementality measurement is the correction.
We set up CTV campaigns with cross-device household attribution through your MMP and, where budget allows, geographic or household holdouts — so the report you get shows what the channel actually caused, not what it happened to be near.