Brand safety in CTV is structurally different from brand safety in display or mobile advertising. The threats are different (app spoofing, not domain spoofing), the viewability question has a different answer (CTV is inherently 100% on-screen), and the content adjacency risks are managed differently (publisher-level curated content, not URL-level brand safety tools). Understanding these distinctions is essential for buyers setting up brand safety requirements for India CTV campaigns.
Invalid traffic in CTV
Invalid traffic (IVT) in CTV is categorised by the MRC (Media Rating Council) and IAB into two types:
- GIVT (General Invalid Traffic): Non-human traffic that can be identified through routine means — known data centre IPs, bot signatures, and traffic patterns that clearly indicate non-human activity. In CTV, GIVT is less common than in desktop display because CTV apps run on physical devices in real households rather than in browsers on data centre servers. GIVT rates on premium India CTV publishers (JioHotstar, SonyLIV) are typically below 1%.
- SIVT (Sophisticated Invalid Traffic): Harder-to-detect fraud requiring more advanced analysis. The primary SIVT vector in CTV is app spoofing: a fraudulent app misrepresents itself as a premium publisher by declaring a false
app.bundlein the OpenRTB bid request. A fraud farm running an unknown app can declareapp.bundle = "com.hotstar.tv"and receive premium CPM bids from buyers who think they are buying JioHotstar. This is the most significant IVT risk in India CTV open exchange buying.
The defence against app spoofing is app-ads.txt (for apps) and sellers.json (for SSPs). App-ads.txt is a file that CTV app publishers host in their app store listing or domain, declaring which SSPs are authorised to sell their inventory. Buyers should configure their DSPs to only buy inventory from sellers declared in app-ads.txt files. In India, app-ads.txt adoption among CTV publishers is high for major publishers (JioHotstar, SonyLIV, Zee5) but inconsistent for smaller FAST channel and AVOD publishers.
IVT rates on open exchange CTV inventory in India: IAS and DoubleVerify data consistently show India CTV open exchange IVT rates of 5–8%, driven primarily by app spoofing on small publishers. Premium publisher PMP deals typically run at 0.5–1% IVT.
CTV viewability standards
CTV viewability is categorically different from display viewability. The MRC's standard for display viewability (50% of pixels in view for 2 continuous seconds) does not apply to CTV because CTV ads always occupy 100% of the screen. A CTV ad impression, by definition, is fully viewable as long as the ad plays.
The relevant CTV viewability standards are therefore different:
- Completion rate (VCR): The primary viewability proxy for CTV. An ad that completes (fires the 100% quartile beacon) is considered fully delivered. VCR on India CTV for non-skippable in-stream inventory typically runs 85–95%.
- Active View for CTV: Google's Active View measurement for CTV defines a viewable impression as one where the ad plays for any duration on a connected TV screen (since it is always 100% of the screen by definition). This is distinct from Active View for display.
- MRC CTV Impression Measurement: The MRC released updated CTV impression measurement guidelines in 2023 that define when a CTV impression is counted as valid — the player must be active, the stream must be active, and the ad must begin rendering. The guidelines exclude impressions where the TV was turned off during delivery (using ACR data on smart TVs to detect this).
Content adjacency and brand suitability
Content adjacency — ensuring your ad does not run adjacent to content that conflicts with brand values — is managed differently in CTV than in web display. In web display, brand safety tools scan URLs and page content for unsafe content. In CTV, the content is a video stream. URL-level brand safety tools do not work for CTV content.
CTV content adjacency control relies on:
- Content metadata targeting: Publishers expose genre, content rating, and IAB content taxonomy categories in the OpenRTB bid request's
contentobject. Buyers can exclude specific content categories (e.g. crime/violence content, adult ratings) using these fields in their DSP targeting. This only works if the publisher populates the content object accurately — which varies by publisher and SSP in India. - Publisher-level curated lists: Buyers can restrict campaigns to specific approved publishers (JioHotstar, SonyLIV, Zee5) and exclude unknown or low-quality inventory. For most India CTV brand campaigns, running on a curated allowlist of 3–5 premium publishers with direct deals provides more reliable content adjacency control than running open exchange with content category exclusions.
- Contextual segments: IAS and DoubleVerify offer contextual targeting products for CTV that classify content by genre and sentiment, allowing positive targeting (only sports, only family content) or negative exclusion. India CTV contextual coverage is limited to the publishers where IAS/DV have server-side integrations.
Brand safety in India CTV
Practical brand safety setup for India CTV campaigns:
- Always activate IAS or DoubleVerify pre-bid filtering: Both platforms offer pre-bid segments that block known fraudulent apps and spoofed inventory before the bid is placed. This is the single most impactful brand safety investment for India CTV open exchange campaigns.
- Use an allowlist for premium inventory: Restrict open exchange buying to publisher app bundles from your approved list. For India, this typically means JioHotstar, SonyLIV, Zee5, Voot, Amazon Prime, and selected FAST channel bundles. DSPs allow app bundle allowlists in targeting settings.
- Verify app-ads.txt compliance: Before activating any publisher's inventory at scale, confirm the publisher has a current app-ads.txt or ads.txt file that declares the SSP through which you are buying. DSP-level app-ads.txt enforcement should be enabled for all India CTV campaigns.
- Set content category exclusions in DSP: Even for allowlisted publishers, add IAB content category exclusions for sensitive categories relevant to your brand (adult content, gambling, alcohol if your brand is in a related category).