CTV Measurement: how to track, verify, and prove campaign performance in India

Why CTV measurement is harder than linear TV

Linear TV measurement in India runs on BARC — a panel-based system that tracks what households watch on broadcast channels. It is imperfect, but it is standardised. Every buyer and seller is working from the same panel data, the same GRPs, the same reach curves. The disagreements are about interpretation, not methodology.

CTV has no equivalent. There is no single agreed panel for streaming viewership in India. Each OTT platform reports its own delivery numbers — impressions served, content hours watched, monthly active users — using its own definitions and without standardised third-party verification. When JioCinema reports 32 million concurrent viewers for an IPL match, that number comes from the platform itself. It may be accurate. It cannot be independently verified with the same rigour as a BARC GRP.

This matters for advertisers because the comparison of linear and CTV reach is therefore an apples-to-oranges exercise. A GRP on Doordarshan is not the same measurement unit as an impression on a streaming platform, even though both represent a person watching something on a screen. Until there is a unified audience measurement currency that covers both, planners will be making cross-media comparisons with structurally incomparable data.

The panel vs census debate

Audience measurement in CTV takes two broad approaches: panel-based and census-based.

Panel measurement tracks a representative sample of households and extrapolates their viewing behaviour to the total population. BARC is a panel system. Nielsen in the US is a panel system. The advantage is population-level reach estimates. The disadvantage is that panels are slow to update, struggle with fragmented streaming behaviour, and often undercount emerging demographics — young urban cord-cutters, for example.

Census measurement tracks actual viewing events on a platform — every stream started, every ad served, every completion. OTT platforms do this by default through their own analytics. Census data is accurate for what happened on a specific platform, but it cannot tell you unduplicated reach across platforms or compare audiences across different content environments.

The ideal measurement system combines both: census data from platforms to count actual events, calibrated against a representative panel to project unduplicated population-level reach. This is the direction the global measurement industry is moving, but India is still in the early stages of building such a system. For now, advertisers must triangulate — using platform census data for delivery verification and panel data (where available) for reach estimation.

Video completion rate: the primary CTV metric

In CTV advertising, video completion rate (VCR) is the closest thing to a universal primary metric. VCR measures the percentage of ad impressions where the viewer watched the ad to completion. On a 30-second pre-roll, that means the viewer sat through all 30 seconds.

CTV VCRs are significantly higher than mobile or desktop video VCRs for a structural reason: most CTV ad environments are non-skippable. If you are watching a show on an AVOD platform and an ad plays, you typically cannot skip it. This creates high completion rates by design — typically 90–95% on premium CTV inventory — but it also raises a question about what completion actually proves. A viewer cannot skip the ad; that does not guarantee they are engaged with it, or even looking at the screen.

VCR remains the most widely used CTV performance metric because it is measurable, standardised, and meaningful enough for brand campaign optimisation. But sophisticated buyers are moving beyond VCR alone — towards attention metrics, brand lift, and outcomes-based measurement that connects ad exposure to actual business results.

Viewability on CTV: different from display

Viewability in digital advertising means: was the ad actually seen? For display ads, the IAB standard is 50% of pixels in view for at least one second. For video, it is 50% of pixels for two seconds. These standards were built for desktop and mobile environments where ads can be below the fold, hidden by other content, or served in non-human traffic.

CTV changes the viewability equation. A TV screen is a dedicated viewing device — there is no below the fold, no other browser tab. An ad served on a CTV device is, by definition, displayed on the screen if the device is on and the content is playing. As a result, CTV consistently achieves near-100% viewability rates by standard IAB definitions.

This does not mean viewability measurement is irrelevant on CTV. Ad fraud — invalid traffic, emulator fraud, server-side ad insertion manipulation — can result in ads being reported as delivered when they were never actually displayed on a real screen. Third-party verification (DoubleVerify, IAS, MOAT) exists to detect these fraud signals and validate that reported impressions represent genuine delivery to real devices. In India, third-party verification adoption on CTV is growing but not yet universal across all publisher relationships.

The attribution gap: connecting CTV to business outcomes

Attribution is the hardest measurement problem in CTV. Linear TV has always struggled with attribution — you could not track whether a person who saw a TV ad then bought the product. CTV was supposed to fix this, because streaming happens on IP-connected devices, and IP addresses can theoretically be matched to other digital behaviour.

In practice, CTV attribution is complicated by several factors. First, CTV devices are shared — a household's smart TV has multiple viewers whose individual identities are unknown. Second, the connected TV device and the conversion device are often different — someone sees a CTV ad, then later buys on their mobile or desktop. Matching those two events requires identity resolution across devices, which depends on either deterministic matching (a logged-in user ID that appears on both devices) or probabilistic matching (IP address plus other signals).

The approaches available include: last-touch attribution (crude and inappropriate for CTV — it almost always credits a performance channel, not the TV ad that drove awareness), multi-touch attribution (more sophisticated, but requires cross-device data), incrementality testing (the gold standard — comparing outcomes for exposed vs unexposed audiences), and media mix modelling (statistical modelling that allocates credit across channels without requiring individual-level data).

In India, most CTV attribution is still primitive — campaign delivery reports from the platform, combined with brand lift studies run by the platform's own measurement tools. Independent attribution that connects CTV exposure to purchase outcomes across the full funnel is available, but requires investment in methodology and is not yet standard practice for most advertisers.

BARC's role in India CTV measurement

BARC India (Broadcast Audience Research Council) is the audience measurement body for television in India. It operates a panel of approximately 50,000+ households across urban and rural India, metered to track channel viewership. BARC data is the official currency for TV advertising in India — every GRP, reach, and frequency number in a linear TV plan comes from BARC.

BARC's mandate has expanded to include digital and streaming measurement, but the progress is slower than the market would like. BARC Streaming Ratings, which aim to provide standardised OTT audience data comparable to television, have been in development and rollout for several years. As of 2025–2026, BARC streaming data is available for some platforms and content, but coverage is not comprehensive, and market adoption by buyers and sellers varies.

The challenge for BARC in CTV measurement is structural: streaming platforms are commercially motivated to control their own data and are not uniformly willing to share granular viewership data with a shared industry body. Unlike broadcast, where all content goes out over a common infrastructure and can be metered at the household level, streaming content delivery is platform-controlled end-to-end. Getting OTT platforms to participate in a shared measurement system requires industry alignment that takes time.

For India CTV planners today, BARC linear data remains the planning currency for broadcast. CTV delivery is measured through a combination of platform-reported data, third-party ad server verification, and — where available — independent measurement tools. The gap between what BARC measures and what streaming delivers is real, and any media plan that treats them as equivalent is working with a flawed model.

What is reliable vs what is not in India CTV measurement today

After years of industry conversation, here is an honest assessment of what India CTV measurement can and cannot deliver:

Reliable: Ad impression counts (when verified by a third-party ad server or verification vendor), video completion rates on platform-managed inventory, device and format-level delivery breakdowns, audience demographic data tied to logged-in platform users, and brand lift studies run by established measurement vendors.

Unreliable or unavailable: Unduplicated cross-platform reach, independent verification of platform-reported content viewership (MAUs, concurrent viewers), standardised audience age/gender measurement comparable to BARC TV data, deterministic attribution connecting CTV exposure to offline purchase, and fraud detection on server-side ad insertion (SSAI) streams where third-party pixels cannot fire.

The practical implication: treat CTV delivery reports from platforms as starting points, not endpoints. Wherever possible, run a third-party verification tag (DoubleVerify or IAS) alongside platform reporting. Accept that cross-platform reach will be estimated, not precisely measured. And build campaign evaluation frameworks that match what the medium can actually prove — brand lift and awareness metrics — rather than demanding last-click conversion data that CTV cannot reliably deliver.

The direction of travel: what is coming

India CTV measurement will improve. The commercial pressure is there — advertisers want accountability, publishers want to prove value, and the industry needs a shared currency to grow the total CTV advertising market. The building blocks include: wider BARC streaming coverage, more third-party verification integration at Indian OTT platforms, ACR (automatic content recognition) data from smart TV manufacturers becoming available for audience measurement, and improved cross-device identity resolution as logged-in user bases on streaming platforms grow.

The timeline is uncertain. Measurement infrastructure tends to lag the medium by several years — the same lag happened with digital and with programmatic. India CTV is at an earlier stage than the US or UK, and the measurement ecosystem will catch up, but not overnight. In the interim, the practitioners who understand the current limitations — and build their campaign measurement frameworks accordingly — will make better decisions than those who assume the numbers they receive are complete and comparable.

What you will find in this tower

The four hubs below cover the full measurement landscape. Start with video metrics if you are new to CTV performance measurement. Go to attribution models if your challenge is proving business outcomes. The India measurement hub is essential reading for anyone planning or buying CTV in India — it covers what works, what does not, and where the gaps are.