Impressions, reach, and frequency are three of the most fundamental metrics in any advertising campaign — and three of the most consistently misunderstood in CTV. Impressions count how many times an ad was delivered. Reach counts how many unique viewers or households saw it. Frequency counts how many times each unique viewer saw it on average. Getting these right in CTV requires understanding how digital tracking, device-level identity, and the fragmented India publisher landscape interact with each metric.
Impressions in CTV: what they count and what they don't
A CTV impression is recorded when the VAST impression beacon fires — the signal that the ad began rendering on the device. This is a technical event, not a confirmed human viewing event. One impression = one ad delivery to one device, in theory.
The complication: a device is not a person. A smart TV in a living room is typically a shared device. One impression to that device may represent viewing by one person, two people, or five people watching together. CTV impressions are device-level, not person-level — a critical distinction when comparing CTV reach numbers to people-based reach in other media.
Impression volume is the most basic delivery metric — the raw count of how many times your ad was served. It is necessary but not sufficient for evaluating campaign performance. High impression volume with poor reach means you are over-serving the same households repeatedly. High impression volume with strong reach means you are building broad coverage.
Reach in CTV: unique households vs unique viewers
Reach measures the unduplicated number of unique viewers or households exposed to your campaign. It is the foundational metric for understanding how broadly your message travelled — one impression to the same household ten times is one reached household, not ten.
Household reach vs person reach
CTV reach is almost always measured at the household level rather than the individual level. This is because CTV identity is anchored to devices and IP addresses — not individual user accounts. A household may have one JioCinema account watched by three family members. The platform can count one household reached; it cannot reliably distinguish three individuals without explicit user-level login data tied to individual profiles.
Some platforms are building person-level reach capabilities using logged-in user data (JioCinema with Jio account logins, YouTube with Google accounts). These are valuable where available but represent a minority of CTV impressions in India's open ecosystem. Most reach measurement is still household-level.
Cross-publisher reach deduplication
Here is where India CTV measurement breaks down significantly: there is no reliable way to deduplicate reach across publishers. If your campaign runs on JioCinema and Zee5 simultaneously, you cannot determine how many households saw your ad on both platforms. Each publisher reports its own reached households — and those reports overlap in unknown ways.
In the absence of cross-publisher deduplication, actual reach is typically lower than the sum of publisher-reported reach. The degree of overlap depends on which publishers you use and what content those audiences share. Industry estimates suggest overlap of 15–30% between major India CTV platforms, but these estimates are not verified with precision.
Frequency in CTV: the measurement gap India planners need to understand
Frequency is the average number of times a reached viewer saw your ad during the campaign period. Calculated as:
Average frequency = Total impressions ÷ Total unique reach
This formula works within a single publisher's reporting — but breaks down entirely when you add cross-publisher frequency. If a viewer watches JioCinema, Zee5, and SonyLIV in the same week and your ad runs on all three, that viewer may see your ad six times or more. The individual publishers each report a low frequency. The combined actual frequency experienced by that viewer is invisible.
The cross-publisher frequency problem in India
India has no industry-standard cross-publisher CTV frequency cap infrastructure. There is no universal ID that ties a viewer's exposure on JioCinema to their exposure on Zee5 — not in a way that is accessible to advertisers for frequency management. The result: India CTV campaigns routinely over-serve heavy CTV viewers while under-serving lighter viewers.
Heavy CTV viewers — who watch multiple platforms, spend significant time in ad-supported content — may see the same campaign twenty or thirty times over a two-week flight. Light viewers, who open one platform once or twice a week, may see it once or not at all. Without cross-publisher frequency data, you cannot see this pattern in your reporting, because no single publisher shows you the full picture.
This is not a solvable problem with current infrastructure. It is a known limitation that serious planners factor into campaign design by:
- Setting frequency caps within each publisher (even though cross-publisher caps are impossible)
- Using publisher mix strategically — running on three publishers at lower weights rather than one publisher at full weight reduces per-platform frequency without solving the total problem
- Weighting toward publishers with known lower overlap with others in your plan
- Monitoring in-platform frequency reports and pulling back on publishers where frequency is running high
How frequency is measured: the device graph approach
Within a single publisher or walled garden, frequency is measured by tracking ad exposures per device ID (or user ID, if login data is available). The platform maintains a record of how many times device X or user Y has seen campaign Z and applies a frequency cap accordingly.
For cross-publisher frequency, the only viable approach is a common identity layer — a shared device graph or universal ID that both publishers recognise and use to count cross-platform exposures. In mature markets, identity solutions like LiveRamp's RampID or The Trade Desk's UID2 are beginning to enable cross-publisher frequency management. In India, these solutions are nascent in CTV. The major India-specific identity layer for CTV remains Jio's account data (for Jio-connected devices), which covers a significant portion of the market but not all of it.
Practical guidance for India planners on reach and frequency
Set realistic reach expectations
India's CTV universe is approximately 40–50 million connected TV homes as of 2026 (industry estimates; exact figures vary by source). Not all of those homes are ad-supported, active viewers. A realistic addressable CTV universe for an advertiser is perhaps 20–30 million households on ad-supported inventory. A campaign reaching 5–8 million unique households is considered strong reach for most categories.
Use frequency caps as a tool, not a guarantee
Set frequency caps within each publisher — typically 3–5 exposures per viewer per week for brand campaigns. Understand that these caps only work within the publisher's own ecosystem. Cross-platform, a viewer may still see your ad far more often than your planned frequency. Treat per-publisher frequency caps as a partial mitigation, not a solution.
Prioritise walled gardens for frequency control
JioCinema and YouTube on TV have the most robust within-platform frequency management because they have logged-in user data and strong device graphs. If frequency control is a campaign priority — for example, avoiding creative fatigue for a high-budget brand campaign — concentrate spend with publishers who can enforce per-user frequency caps reliably, rather than spreading across open programmatic where frequency data is fragmented.
Co-viewing adjustment for reach reporting
CTV is a shared screen medium. Average co-viewing multipliers for India CTV are estimated at 1.5–2.5 viewers per device, depending on content type and household size. A campaign reaching 5 million unique devices may have reached 7–12 million unique viewers in practice. Some publishers apply co-viewing multipliers to their reach reporting; others do not. Always clarify whether reported reach figures include a co-viewing adjustment — and if so, what multiplier was applied and on what basis.
How CTV reach and frequency compares to linear TV
Linear TV in India is measured by BARC — which provides people-level reach and frequency data across cable and satellite television. CTV has no equivalent single-source measurement system. This creates a fundamental comparability problem for planners trying to evaluate CTV against linear on a reach-per-rupee basis.
Until India has a unified cross-platform measurement standard that covers both linear and streaming — which BARC is working toward with its streaming measurement initiative — planners must accept that CTV reach numbers and linear reach numbers are not directly comparable. CTV reach is device/household level; linear reach is people-level. CTV frequency is within-publisher only; linear frequency is measured at the total campaign level. Plan across both media, but do not try to add them on the same scale without significant assumptions.