Co-viewing is the most underappreciated factor in CTV audience measurement. A CTV device — a smart TV, a Fire TV Stick, a gaming console — generates a single ad impression each time an ad plays. But in many households, multiple people are in the room watching simultaneously. If two people watch the same pre-roll ad together, the ad server logs one impression. Both people were exposed to the ad, but only one was counted. This gap between device-level impressions and actual person-level reach is the co-viewing problem, and it affects every CTV campaign in every market — including India, where household group viewing is especially common.
What co-viewing means for reach measurement
If the average CTV viewing session in India involves 2.3 people watching simultaneously, then a campaign that delivers 10 million impressions actually reached approximately 23 million viewers — not 10 million. But if you report campaign reach as 10 million, you are significantly undercounting your actual audience.
Conversely, if you are using impression counts to manage frequency (targeting "no more than 3 impressions per person"), and 2.3 people are watching each impression, then a household that has received 3 impressions has actually received 3 individual exposures per person — which means you are appropriately capping frequency. Or if only one of those 2.3 people is the primary decision-maker for your product, you may have reached them far more often than the impression count suggests.
The direction of the error depends on what you are optimising for. But in all cases, treating device-level impressions as person-level exposures without a co-viewing adjustment produces inaccurate campaign analysis.
How co-viewing is estimated
Because CTV devices do not inherently know how many people are watching, co-viewing must be estimated. There are three main approaches:
1. Panel-based co-viewing measurement
Audience research panels (like BARC India's TV panel or Nielsen's US panel) can measure co-viewing directly. Panel homes have measurement devices that log when people are in the room and watching. The resulting data provides average co-viewing rates by content genre, time of day, and demographic segment.
For linear TV in India, BARC data implicitly captures co-viewing because the measurement is person-level (panel members register themselves). For CTV, this panel data is used to produce co-viewing multipliers — statistical adjustments applied to CTV impression counts to estimate person-level reach.
2. Platform-based co-viewing signals
Some streaming platforms have indicators that suggest co-viewing — for example, a family profile on a shared account, household-level profile groupings, or UI interactions that suggest multiple people are engaged. These are imperfect signals but can inform co-viewing adjustments for platform-specific measurement.
3. Proprietary measurement vendor models
Vendors like Nielsen and Comscore apply co-viewing adjustment factors based on their proprietary research. Nielsen's co-viewing factors for CTV are derived from its US panel data and published annually. These are US-calibrated figures — directionally useful but not directly applicable to India.
Co-viewing rates: what the data shows
In the US, Nielsen has published CTV co-viewing rates of approximately 2.0–2.5 viewers per impression for prime-time CTV viewing. This means a campaign that delivers 10 million CTV impressions in prime time may have been seen by 20–25 million people.
There are no comparable publicly available co-viewing benchmarks for India CTV specifically. However, there are structural reasons to expect co-viewing rates in India to be higher than US benchmarks:
- Household size: India's average household size is approximately 4.4 persons, compared to 2.5 in the US. More people per home means more potential co-viewers.
- Joint family viewing: Multi-generational households are more common in India. TV watching tends to be a family activity, particularly in the evenings.
- Single TV per household: Many Indian households have one television that the entire family uses. In US households, multiple TVs in different rooms reduce co-viewing per device.
- Premium content events: IPL matches, major movie releases, and live events drive especially high co-viewing. These are also major CTV ad inventory moments.
Industry estimates for India CTV co-viewing typically range from 2.5 to 3.5 viewers per impression, but these are not based on systematic measurement. Treat them as directional guidance, not established benchmarks.
Co-viewing and frequency management
Co-viewing creates a specific challenge for frequency management. CTV campaigns typically apply frequency caps at the device or user ID level — for example, "show this ad to a maximum of 4 times per week per device." If 3 people are watching each time, those 3 people collectively receive 12 exposures from 4 device-level impressions. Whether that is over-exposure depends on the campaign objective.
For brand-building campaigns where broad exposure is the goal, co-viewing is a bonus: you are reaching more people than the device count suggests, at the same cost. For performance campaigns where precise frequency and targeted reach are important, co-viewing creates noise in the measurement that makes it harder to connect ad exposure to individual behaviour.
How advertisers should factor co-viewing into planning
The practical guidance for India CTV planners:
- Apply a co-viewing multiplier when reporting reach: If your platform reports 5 million impressions, apply a co-viewing factor (industry directional: 2.5x for India prime-time CTV) to estimate actual person reach: approximately 12.5 million people. Document the methodology so the number is defensible.
- Do not apply co-viewing multipliers to frequency: Frequency caps are set at the device level for a reason — you want to control how often a household sees your ad, regardless of how many people are in the room. Do not inflate frequency figures by the co-viewing multiplier.
- Distinguish impression CPM from person CPM: If you paid CPM 350 for 5 million CTV impressions, your effective cost per person exposed is much lower once co-viewing is factored in. This is a legitimate argument for CTV's value relative to linear TV CPM comparisons — but make the co-viewing assumption explicit.
- Acknowledge uncertainty: There is no audited, India-specific co-viewing benchmark for CTV. Any co-viewing adjustment you make is based on estimation. Be transparent about this in campaign reports.
India household viewing context
CTV in India is disproportionately an evening, household-level activity. Research from platform surveys and third-party studies consistently shows that prime-time CTV viewing (8pm–11pm) in India involves higher co-viewing than daytime or weekend morning viewing. Content categories with highest co-viewing include:
- Live cricket matches (IPL, India internationals)
- Family drama and soap operas
- Bollywood film premieres
- Reality and competition shows
Niche content — individual-oriented genres like romance, thriller films, or comedy specials — tends to have lower co-viewing rates than sports or family drama. Planners should consider content environment when applying co-viewing assumptions.