Co-viewing — multiple people watching the same CTV screen simultaneously — is structurally more significant in India than in any other major CTV market. Average India household size is 4.4 people. When one person streams JioHotstar on the living room TV, 2–4 family members are likely watching alongside them. A single logged-in account generates one impression but reaches multiple people. Advertisers who do not account for co-viewing systematically undercount their true reach and overestimate their cost per person reached.
Why co-viewing matters more in India
Co-viewing is not unique to India — it happens on CTV everywhere. What makes India distinctive:
- Household size: India average household = 4.4 persons. US = 2.5. UK = 2.4. More people per household means more co-viewers per impression by default.
- Joint content consumption culture: TV watching in India is a family activity more than a personal one. Hindi GEC content (drama serials, reality TV, cricket) is explicitly designed for joint household viewing. Appointment viewing around cricket and IPL brings 4–6 people to a single screen.
- Single-TV households: Many India households have one primary TV — the living room CTV device. A single OTT subscription shared across the family is the economic norm, not an exception. This concentrates viewership on the shared screen.
- Single-account families: Unlike US households where each adult may have a personal streaming account, India families frequently share one account. One logged-in user identifier = 4+ actual viewers.
How co-viewing is measured
Co-viewing cannot be measured directly by ad tech systems — no pixel, VAST tag, or DSP log captures how many people are in the room. The methods used are indirect:
Automatic Content Recognition (ACR) panel surveys: Smart TV manufacturers (Samsung, LG) embed ACR technology that detects what is on screen. Some manufacturers layer this with periodic surveys asking how many people are watching. This provides a direct co-viewing signal but is limited to ACR-enabled smart TV households and has low survey response rates.
Publisher survey research: JioHotstar and SonyLIV have commissioned periodic audience studies that survey logged-in users about their viewing behaviour ("how many people typically watch with you?"). This produces a co-viewing multiplier for different content categories. The numbers are self-reported and may overstate co-viewing for lower-engagement sessions.
BARC peoplemeter data: For linear TV, BARC's 50,000-home panel captures per-person viewership within metered households. This produces verified co-viewing data for linear content. BARC Streaming (SDK-based) does not yet have a per-person layer — it counts streams, not viewers. BARC linear co-viewing data is used as a proxy benchmark for CTV co-viewing estimates.
Third-party audience measurement (Nielsen, Kantar, Ipsos): Post-campaign surveys and audience panels can capture co-viewing at a segment level. Nielsen's DAR (Digital Ad Ratings) product attempts to measure total audience (including co-viewers) for digital campaigns in India, but CTV coverage is limited compared to the US.
Publisher co-viewing multipliers
India CTV publishers apply co-viewing multipliers when presenting reach data to advertisers. These multipliers express how many people are estimated to have seen each impression:
| Content type | Estimated co-viewing multiplier | Notes |
|---|---|---|
| IPL / live cricket | 3.0–4.5× | Peak co-viewing; multi-generational household event |
| Hindi GEC (drama, reality) | 2.5–3.5× | Family appointment content; evening prime time |
| Bollywood films | 2.0–3.0× | Weekend family viewing; lower on weeknights |
| News / current affairs | 1.5–2.5× | Primarily adults; lower co-viewing than entertainment |
| Original series / dramas | 1.5–2.0× | More individual viewing behaviour on originals |
| Kids content | 1.5–2.5× | Parent often present; but fewer adults per screen |
These multipliers are publisher-provided estimates based on their own research. They are not independently verified by any third party in India as of 2026. Treat them as indicative, not certified. The IAB US has developed a co-viewing methodology for CTV; India has not yet adopted a standardised equivalent.
Co-viewing and frequency capping
Co-viewing creates a frequency cap problem. A DSP frequency cap of 3 impressions per device per day prevents the same account from seeing your ad more than 3 times. But if 3 people are watching that account, they have collectively seen 9 impressions — 3 each. From the advertiser's perspective, the effective per-person frequency is already 3× higher than the device cap suggests.
There is no clean solution to this with current technology. Options:
- Reduce device frequency caps for large-household audience segments (SEC B/C, Hindi language, non-metro) where co-viewing is highest. A 2-per-device cap in a 4-person co-viewing household still delivers 8 impressions across the household.
- Account for co-viewing in frequency planning: If your target frequency is 3 per person and the co-viewing multiplier is 2.5, set device frequency cap to 1–2 impressions to hit per-person frequency of 2.5–5.
- Accept the measurement gap: For brand awareness goals, co-viewing is a feature not a bug — you reach more people per impression. For precise frequency management, co-viewing limits accuracy.
Co-viewing and brand lift measurement
Brand lift surveys are served to logged-in users on the platform — one survey per account, not per viewer. If a household with 3 co-viewers receives a brand lift survey, only the account holder responds. The other 2 co-viewers are not captured. This means:
- Brand lift results may understate awareness lift because co-viewers who actually saw the ad are not in the survey base.
- Brand lift sample sizes based on impression counts underestimate true person-level reach — the denominator should be impressions × co-viewing multiplier, but survey platforms use raw impression counts.
- For co-viewed content (IPL, Hindi GEC), brand lift ad recall numbers may appear lower than expected because the survey response pool is a subset of actual viewers.
Planners running brand lift studies for campaigns heavy on co-viewed content should note this caveat in reporting and consider it when interpreting recall and awareness scores.
India co-viewing benchmarks
Based on available publisher research and industry estimates for India CTV (2024–2026):
- Overall India CTV average co-viewing multiplier: approximately 2.3–2.8×
- JioHotstar self-reported co-viewing: 2.4× average; 3.5–4.5× during IPL
- Non-metro households: higher co-viewing (fewer screens, larger families) — estimated 3.0–3.5× average
- Metro SEC A households: lower co-viewing (more personal devices, smaller families) — estimated 1.8–2.2×
- Tamil/Telugu GEC content: high co-viewing (2.5–3.5×) — family viewing culture strong in South India
Planning implications
Four practical actions for India CTV planners accounting for co-viewing:
- Adjust reach calculations. If a publisher reports 10M impressions at 2.5× co-viewing, true reach is ~25M person-exposures. Use this in cost-per-reach calculations rather than raw impressions.
- Set lower device frequency caps for high-co-viewing content. Divide your target per-person frequency by the estimated co-viewing multiplier for your content mix to get the device-level cap.
- Factor co-viewing into CPM value assessment. JioHotstar's ₹200–400 CPM includes co-viewing reach. On a per-person-reached basis, a ₹250 CPM with 2.5× co-viewing = ₹100 effective CPM per person. This makes CTV significantly more competitive vs linear TV (where co-viewing is built into GRP methodology).
- Interpret brand lift with caution. Survey-based brand lift undercounts co-viewer recall. Add a co-viewing correction factor when comparing CTV brand lift to display or mobile benchmarks.