Yield optimisation in CTV is the practice of maximising the revenue a publisher earns from every available ad impression. For a CTV publisher, the goal is simple: fill as many ad slots as possible, at the highest achievable CPM, without harming viewer experience. In practice, achieving that balance requires managing floor prices, demand partners, ad pod structure, and fill rate simultaneously — pulling each lever affects the others.
This guide explains the core concepts, the key levers, and how yield optimisation in CTV differs from display or mobile web. India publishers face a specific set of constraints and opportunities that are covered in the final section.
What does yield optimisation actually mean in CTV?
In CTV, yield is typically expressed as revenue per thousand impressions (CPM) or, more usefully, as revenue per hour of content viewed (RPH). A publisher running two ad pods per episode, each with two 30-second slots, has four ad impressions to monetise per episode. Yield optimisation asks: how do you make each of those four impressions worth as much as possible?
The levers are interconnected:
- Floor prices: Set too low and you leave revenue on the table. Set too high and you lose fill, leaving slots empty.
- Demand partners: More demand sources increase competition for your inventory, which should raise CPMs. But each integration adds latency and complexity.
- Ad pod structure: The first slot in a pod commands a premium. Pod length and frequency affect completion rates, which affects CPMs advertiser willingness to pay.
- Fill rate: An unfilled slot earns nothing. Backfill strategies ensure revenue even when premium demand is absent.
- Audience data: Richer targeting signals let advertisers bid more confidently. Publishers who surface clean first-party data get higher CPMs.
How does CTV yield optimisation differ from display?
CTV and display yield optimisation share the same underlying logic — maximise revenue per impression — but the mechanics are different in important ways.
No cookies, no real-time bidding with full user profiles
Display yield optimisation benefits from third-party cookies that carry rich behavioural data into the auction. CTV uses device IDs and platform first-party data. The signals are different, the identity resolution is harder, and the cross-device matching is more complex. This affects both CPMs and how floor prices should be calibrated.
Ad pods instead of individual placements
Display serves one ad at a time. CTV serves ad pods — sequential groups of ads in a single break. Managing a pod requires deciding how many ads to include, what duration mix to accept, how to sequence them, and how to handle unfilled slots mid-pod. This is structurally different from optimising a single display impression.
Latency is more punishing
On desktop display, a slow ad load is an annoyance. On CTV, a delayed ad break is a buffering spinner on a 55-inch screen. Viewers notice. Completion rates drop. That makes latency management — particularly when running header bidding or multiple demand calls — a hard constraint that display publishers rarely face at the same severity.
Premium CPMs but limited liquidity
CTV commands higher CPMs than display globally — the lean-back, full-screen, non-skippable environment justifies a premium. But programmatic CTV liquidity is lower. Fewer DSPs bid on CTV, fewer advertisers have activated CTV budgets, and the demand-side infrastructure is less mature. In India, this gap is particularly pronounced.
The five key levers of CTV yield optimisation
1. Floor price management
A floor price is the minimum CPM at which you will sell an impression. Setting the right floor is the most direct yield lever. The trade-off: a higher floor means higher average CPM on impressions that sell, but a lower fill rate. A lower floor means more inventory fills but at weaker prices. The optimum floor is content-specific, daypart-specific, and audience-specific. Live sport commands a different floor than archive VOD content.
2. Demand partner diversification
A publisher with only one demand source — say, a single SSP — gets whatever that SSP's buyers are willing to pay. Adding more SSPs, direct demand, and private marketplace (PMP) deals increases competition for each impression. The first additional demand partner typically delivers the largest yield uplift. Returns diminish as you add more, while complexity grows. Most mature CTV publishers settle on two to four SSPs plus direct IO deals for premium inventory.
3. Ad pod strategy
Slot position within a pod matters. The first slot in a pre-roll pod commands a 20–40% premium over later slots in many markets — buyers know viewers are most attentive at the start of a break. Keeping pods short (two to three slots) maintains completion rates and viewer experience. Longer pods increase ad load but risk tune-out, which undermines the completion metrics that justify premium CPMs.
4. Fill rate optimisation
An unfilled slot is a zero-revenue slot. Fill rate management means having a backfill strategy — house ads, public service announcements, lower-CPM demand — that ensures every slot earns something, even if not at premium rates. Backfill should be priced carefully: setting your backfill floor too high leaves slots empty; setting it at zero devalues your inventory signal to the market.
5. First-party data activation
Publishers who surface audience attributes — content preferences, geography, device type, viewing frequency — into the bid stream get higher CPMs because buyers can bid more precisely. This requires clean data infrastructure, appropriate consent flows, and SSP integrations that support passing custom key-values or user data signals. In India, this is largely underdeveloped, which represents an opportunity for publishers who invest in it.
Unified auction vs waterfall: the architecture question
How you run your auctions matters as much as what floors you set. The traditional waterfall model sequentially offered inventory to demand sources one at a time — inefficient and value-destroying. Unified auctions (or header bidding equivalents in CTV) let all demand partners bid simultaneously, ensuring the highest bidder wins. Most mature CTV monetisation stacks have moved to unified auctions, but India's ecosystem is still transitioning. Publishers using waterfall architecture for CTV are likely leaving significant revenue on the table.
CTV yield optimisation for India publishers
India's CTV publishers operate in a market with three defining characteristics that shape yield strategy:
Demand-supply imbalance
Supply of CTV impressions in India is growing faster than advertiser demand for programmatic CTV specifically. Most India CTV advertising is still bought direct — brand teams and agencies buying sponsorships and packages directly from platforms like JioCinema or Hotstar. Programmatic CTV buying through DSPs is limited, which means smaller publishers have fewer demand sources and weaker auction dynamics. Yield optimisation in this environment means prioritising direct sales and PMP deals over open exchange.
Limited SSP options
Only a handful of SSPs actively support India CTV publishers: PubMatic, Magnite, and a few others have India CTV programmes. Google Ad Manager supports CTV ad serving but programmatic CTV fill from Google's demand side is uneven for India. This limits demand diversification options compared to US or European publishers.
CPM benchmarks are lower than global
India CTV CPMs are significantly below US benchmarks. Industry estimates suggest India CTV programmatic CPMs range from ₹200 to ₹600 ($2.50–$7.50) for standard content, with live sports commanding ₹600–₹2,000+ in direct deals. These ranges vary widely by content type, deal structure, and audience quality. Publishers who can demonstrate premium, verified audiences command the higher end of these ranges.
Sports-driven demand spikes
India's CTV advertising market is heavily IPL-correlated. Demand spikes sharply during cricket tournaments and drops during off-seasons. Yield strategy must account for this seasonality — floor prices, demand source mix, and pod strategy may need to be adjusted by quarter. Building non-sports content categories that attract steady advertiser demand is a long-term yield diversification play.
Common yield optimisation mistakes CTV publishers make
- Setting floors and forgetting them: Floors set in 2022 are wrong in 2026. Market CPMs shift. Floors need quarterly review at minimum.
- Optimising for fill rate instead of revenue: A 95% fill rate at ₹150 CPM earns less than an 80% fill rate at ₹400 CPM. Track revenue, not just fill.
- Too many ad slots per hour: Ad load that degrades viewing experience reduces completion rates and long-term audience retention — both of which hurt yield over time.
- No first-party data strategy: Publishers who cannot describe their audience to buyers in bid-stream signals rely entirely on contextual signals. This consistently underperforms audience-targeted inventory.
- Ignoring direct sales: In India's current market, programmatic yield is limited. Publishers who invest in direct sales relationships — building a deck, pricing packages, selling branded environments — typically outperform pure programmatic approaches on a revenue-per-hour basis.