Guide · Programmatic Buying

Programmatic CTV buying playbook for India

Programmatic CTV buying in India is structurally different from programmatic display or mobile video. The inventory is largely walled. The one platform with genuine open programmatic access — YouTube CTV — operates within Google's closed ecosystem. Private Marketplace (PMP) deals dominate what little programmatic CTV access exists outside YouTube. And the DSP landscape for India CTV is thin compared to the US or EU. This playbook covers how to navigate all of it: DSP selection, deal structures, bid floors, frequency management, and the mistakes that waste budget on India CTV programmatic buys.

India CTV programmatic: the honest picture

Before getting into execution, you need to understand the structural constraint: most India CTV inventory is not available programmatically. JioHotstar is direct-deal only for the majority of its inventory. SonyLIV is direct-deal only. Zee5 is direct-deal only. Samsung Ads is primarily direct with limited programmatic access. The programmatic CTV options in India as of 2025–26 are:

  • YouTube CTV — fully programmatic via DV360 or Google Ads. The largest single programmatic CTV pool in India. But it is Google's ecosystem, not the open web.
  • JioHotstar — limited programmatic access through select DSP partnerships. Not all DSPs can access JioHotstar inventory. Inventory available programmatically is a subset of total JioHotstar inventory.
  • Amazon DSP / Fire TV — Amazon's DSP provides programmatic access to Fire TV home screen and in-app inventory. Available in India with India audience targeting.
  • OEM/Smart TV platforms — Samsung Ads, LG Ads have limited programmatic pipes in India. Mostly direct deal but programmatic access is developing.

If you are expecting to run a single programmatic CTV campaign in India that reaches audiences across JioHotstar, SonyLIV, and Zee5 the way you might run a display campaign across publishers — that is not possible. India CTV programmatic means YouTube CTV + limited JioHotstar + Amazon DSP. Everything else is direct.

DSP evaluation for India CTV

Not all DSPs have meaningful India CTV inventory access. Before committing to a DSP for India CTV, verify these five things:

1. Actual India CTV inventory sources

Ask the DSP to show you the specific supply sources for India CTV — which SSPs are connected, which publishers have active deals, and what the estimated monthly India CTV impression pool looks like. A DSP that lists "CTV" as a category but whose India CTV inventory is primarily YouTube (accessible via any DSP through Google's pipes) is not adding value over Google Ads directly.

2. JioHotstar programmatic access

JioHotstar's programmatic inventory is only accessible through select DSP partnerships. Ask specifically: does this DSP have a direct supply deal with JioHotstar for India CTV? If yes, what inventory segments are available (live sports vs VOD vs news)?

3. Measurement and reporting capabilities

Verify the DSP can report CTV-specific metrics: video completion rate by quartile, household reach and frequency, device type breakdown (smart TV vs streaming stick vs game console), and content category targeting. A DSP that reports CTV impressions without completion rate breakdown is not equipped for CTV buying.

4. Identity and targeting on CTV

CTV does not use cookies. Ask what the DSP's identity solution is for CTV in India: IP-based household targeting, first-party data onboarding, or platform-provided audience segments. Understand what targeting actually resolves on device vs what is modelled at the segment level.

5. Minimum commitment and pricing model

Some DSPs have minimum monthly spend commitments for CTV access. Understand the pricing structure: platform fee (% of media spend), CPM add-on, or flat fee. Total cost of programmatic buying on CTV in India is often 20–35% above media CPM when DSP fees, data costs, and ad serving are included.

DSP India CTV access Best for Key limitation
DV360 (Google) YouTube CTV (full), limited JioHotstar Scale, audience targeting, measurement Google ecosystem only at full access
Amazon DSP Fire TV (full), Amazon Publisher Services Amazon audience data, Fire TV home screen Amazon ecosystem; India CTV scale is smaller
The Trade Desk Limited; depends on SSP partnerships Cross-channel planning, UID2 identity India CTV supply thin vs US; JioHotstar access varies

Deal types: open auction vs PMP vs programmatic guaranteed

India CTV programmatic operates across three deal structures, each with different inventory quality and price implications:

Open auction (RTB)

Open auction CTV inventory in India is thin and quality-variable. What appears as "CTV" in open auction can include long-tail publisher apps, sideloaded applications, and inventory with limited verification. On YouTube CTV bought via Google Ads non-reserved, open auction is the default and is high-quality. On other DSPs, open auction CTV in India should be treated with caution — verify supply path, app bundle IDs, and publisher names before scaling spend.

Private Marketplace (PMP)

PMP deals are the primary programmatic model for non-YouTube India CTV. A publisher (JioHotstar, a premium OTT app, an OEM) creates a private deal with a floor price and specific inventory segments. The buyer receives a Deal ID to activate in their DSP. PMP gives the buyer access to inventory that is not in the open auction, with better brand safety guarantees and more transparent supply path. Most JioHotstar programmatic access for India buyers is structured as a PMP.

Programmatic Guaranteed (PG)

Programmatic Guaranteed combines the workflow of direct deals with the infrastructure of programmatic: a fixed impression volume, fixed CPM, and guaranteed delivery — but executed via DSP rather than a manual insertion order. PG deals are rare for India CTV currently but are the direction the market is moving for larger direct buys. YouTube Reserve (Masthead) is the closest India example of PG-style guaranteed CTV buying.

Structuring a PMP deal

When you access India CTV inventory through a PMP, the negotiation and setup has specific components that differ from a direct IO. Here is how to structure it correctly:

Negotiate the deal parameters upfront

  • Floor price: the minimum CPM the publisher will accept in the auction. Your bids must clear this floor. Publishers set floors at or above their direct-deal CPMs — do not expect PMP to be cheaper than direct.
  • Inventory segment: specify exactly what inventory the deal covers. Live sports only? All VOD? Specific content genres? Unspecified PMP deals can deliver against lower-value inventory within the publisher's catalogue.
  • Impression forecast: ask for an estimated monthly impression volume at the floor price. This tells you whether the deal can actually scale to your budget.
  • Frequency cap: negotiate a frequency cap at the deal level, not just at the DSP campaign level. Publisher-side frequency capping is more reliable on SSAI platforms where DSP-side cookies cannot count exposures.

Deal ID activation

Once terms are agreed, the publisher provides a Deal ID. Activate this in your DSP under the Inventory section (exact location varies by DSP). Create a campaign or line item that targets the Deal ID specifically — do not mix PMP and open auction inventory in the same line item or you cannot isolate PMP performance.

Test before scaling

Run the PMP deal at low budget for 48–72 hours and verify: bid response rate (are your bids winning?), CPM actually clearing (is spend pacing against the floor?), and creative serving (are completion rate signals coming through?). PMP deals sometimes have configuration issues at activation — catch them before committing full budget.

Bid floors and CPM strategy

Bidding strategy for India CTV programmatic differs from display. CTV inventory is scarce relative to demand, especially premium content inventory. Under-bidding to "find efficiency" usually results in zero delivery on quality inventory — the floor price clears the market and if you bid below it, you simply do not serve.

Practical approach:

  • For YouTube CTV non-reserved: start with a CPM target 20–30% above the category average to ensure competitive delivery. Reduce once you have data showing where impression volume actually clears.
  • For PMP deals: bid at 1.1–1.3× the floor price. Bidding exactly at the floor risks being outbid by other buyers in the same deal. Bidding significantly above the floor on a first-price auction wastes budget unnecessarily.
  • For IPL or live sports PMP: expect floor prices of Rs 300–600+. Budget accordingly — there is no negotiating the market during live inventory.

First-price vs second-price: India CTV programmatic auctions are predominantly first-price. You pay what you bid, not the second-highest bid plus one. This makes overbidding directly expensive — set bid ceilings carefully and use bid shading where your DSP offers it.

Targeting parameters that actually work on India CTV

CTV targeting in India is more limited than desktop or mobile. No cookies. Limited third-party data infrastructure. Platform-provided segments are the most reliable. What actually works:

  • Content category targeting: most reliable signal on CTV. Target inventory adjacent to sports, entertainment, or news content. Publisher-provided content categories are deterministic — the publisher knows what content is playing.
  • Device-level IP targeting: household-level targeting via IP address. Not user-level but strong for household income and geography inference. Works on smart TVs and streaming sticks.
  • First-party data onboarding: upload your CRM or customer email list, match to household IPs via your DSP's identity graph, and target existing customers or lookalikes on CTV. Match rates in India are typically 15–35% — lower than desktop but meaningful for large CRM datasets.
  • Google audience segments (DV360/YouTube CTV): in-market, affinity, and custom intent audiences on YouTube CTV are the most sophisticated targeting available on India CTV. These work because Google has deterministic login data across devices.
  • Amazon audience segments (Amazon DSP): Amazon's purchase and browsing data is highly accurate for product-adjacent targeting on Fire TV. Best for e-commerce, FMCG, and consumer electronics brands with relevant Amazon purchase signals.

What does not work well on India CTV

Third-party cookie-based audience segments, demographic segments derived from mobile/desktop behaviour without device graph mapping, and real-time contextual segments dependent on page-level signals — none of these transfer cleanly to CTV. A segment that delivers strongly on mobile may have poor match rates on smart TV inventory. Always validate audience reach estimates for CTV specifically, not pooled across all digital environments.

Frequency management on India CTV programmatic

Frequency management is the most commonly mishandled element of India CTV programmatic buying. The walled garden structure means frequency caps set in a DSP only govern the inventory the DSP can see — which on India CTV is typically YouTube CTV + a PMP deal or two. A viewer watching on JioCinema (direct deal, no DSP visibility) is invisible to your DSP frequency cap.

Practical frequency rules:

  • Set DSP-level frequency caps at 3–4 impressions per household per week for programmatic CTV line items
  • Set separate caps per deal ID when running multiple PMP deals — do not let a single household accumulate impressions across multiple deals in the same DSP
  • For YouTube CTV: use Google's household frequency capping, which is more accurate than cookie-based caps because it uses Google login data across devices
  • Accept that total cross-platform frequency (programmatic + direct) is unmanageable in real time — build this into your plan by treating programmatic and direct as separate reach pools

Common India CTV programmatic mistakes

  1. Treating CTV as premium mobile video. CTV completion rates are higher and CPMs are higher. The creative, targeting, and measurement logic is different. Repurposing a mobile video campaign for CTV without adjusting specs, targeting, and measurement will underperform.
  2. Accepting "CTV" inventory without verifying supply path. In open auction, inventory labelled as CTV can include low-quality app environments. Always check app bundle IDs, publisher domain lists, and supply path transparency reports before scaling.
  3. Setting bids at floor price. On a competitive PMP, bidding exactly at the floor often means losing to other buyers. Add a 10–20% bid buffer above floor on deals where delivery is important.
  4. Mixing PMP and open auction in the same line item. You cannot compare performance or diagnose issues if premium and remnant inventory are aggregated. Always separate deal IDs into dedicated line items.
  5. Expecting third-party verification to work like display. IAS, DoubleVerify, and MOAT cannot fire client-side pixels on SSAI-served inventory (most of India's premium CTV). Do not benchmark CTV verification against display verification rates.
  6. Ignoring VAST chain depth. On India CTV, VAST redirect chains longer than 2–3 hops cause timeout failures and missed impressions. Confirm your VAST setup with the platform before launch, especially for programmatic delivery through multiple intermediaries.
  7. Scaling too fast before verifying delivery. Run a 48-hour test at low budget before scaling to full flight. Configuration issues — wrong Deal ID, VAST timeout, creative rejection — are much cheaper to catch early.

Pre-buy checklist

Before activating a programmatic India CTV buy

  • DSP verified for actual India CTV supply access (not just CTV category claim)
  • JioHotstar programmatic access confirmed if JioHotstar is in plan
  • Deal ID received and activated in DSP for each PMP deal
  • PMP deals set up in isolated line items (not mixed with open auction)
  • Floor prices confirmed and bids set at 1.1–1.3× floor
  • Inventory segments specified in each deal (not all-inventory)
  • Frequency caps set per deal ID and at campaign level
  • VAST tag tested — chain depth confirmed at 3 hops or fewer
  • Creative spec confirmed per platform (1080p, H.264, 15s or 30s)
  • Third-party tracking tags confirmed with publisher (or noted as not supported on SSAI)
  • 48-hour test run at low budget before full campaign launch
  • Reporting setup confirmed: completion rate by quartile, household frequency, device breakdown