Bid shading is an algorithm that reduces the bid a buyer submits in a first-price auction to avoid overpaying. In a first-price auction — the standard for India CTV — the winning bidder pays exactly what they bid. Without bid shading, a buyer who bids ₹300 CPM and wins against a second bidder at ₹160 CPM pays ₹300 — a ₹140 premium over what was needed to win. Bid shading uses historical auction data to estimate the lowest bid likely to win and submits something closer to that figure, reducing the effective CPM the buyer actually pays.
The first-price auction problem
In a second-price auction (the old standard for display programmatic), you bid your true value and pay the second-highest bid plus ₹0.01. There is no incentive to shade your bid — you win at market price regardless of what you bid above the clearing price.
First-price auctions changed this. You now pay exactly what you bid. If you overbid relative to the clearing price, you waste budget. If you underbid, you lose impressions you could have won. The optimal bid is as close as possible to the clearing price — high enough to win, low enough not to overpay. Determining that optimal bid requires knowing what others are bidding, which you cannot see directly.
CTV programmatic in India moved fully to first-price auctions with the migration of Google Ad Manager (used by JioHotstar), Magnite (SonyLIV), and PubMatic (Zee5) to first-price models from 2019–2021. Since then, bid shading has been a built-in feature of every major DSP.
How bid shading works
Bid shading algorithms use historical auction data to predict what price is likely to clear a given impression. The inputs typically include:
- Historical clearing prices for the same publisher, ad format, time of day, and audience segment
- Recent win/loss signals — what this buyer bid when they won and lost on similar impressions
- Floor prices set by the publisher (bid shading will never go below floor)
- Current competition signals — how many DSPs are typically bidding in this auction and at what range
The algorithm outputs a shade factor — a multiplier between 0 and 1 — applied to the buyer's input bid. If a buyer inputs ₹300 CPM and the shade factor is 0.65, the actual submitted bid is ₹195. The DSP aims to shade as aggressively as possible (lower effective CPM) while maintaining a win rate that delivers the campaign's required impressions.
Shade factors are not fixed. They update continuously as the algorithm observes auction outcomes. Heavily competed inventory (IPL live sports on JioHotstar) gets less aggressive shading because many buyers are bidding close to the clearing price. Lower-competition inventory (niche FAST channels at non-peak hours) gets more aggressive shading.
Bid shading in DSPs
The major DSPs used for India CTV all implement bid shading, with different levels of buyer control:
- DV360: Bid shading is applied automatically when automated bidding is enabled (Target CPM or Maximise Viewable Impressions strategies). Manual CPM bidding does not automatically shade — the buyer submits their entered CPM. For India CTV campaigns, automated bidding with a Target CPM is the default recommendation; manual bidding is used for fixed-price PMP deals where the clearing price is known.
- The Trade Desk: TTD's Koa AI system applies bid shading across all bidding strategies by default, including manual CPM bids. Buyers can observe the effective CPM paid versus the input bid in reporting to understand the shading discount being achieved.
- Amazon DSP: Bid shading applied automatically. Amazon's system is optimised for its own inventory (Prime Video) and external exchange inventory separately — effective CPMs may differ significantly between these two pools.
India CTV-specific context
India CTV has some characteristics that affect bid shading behaviour:
- Premium sports inventory resists shading: During live cricket (IPL, India international tours), demand concentration pushes clearing prices very close to maximum bids. Bid shading algorithms detect high competition and reduce the shade factor — effective CPMs approach input CPMs during peak sports periods. A buyer who expects large shading discounts during IPL will be disappointed.
- Thin non-prime inventory enables aggressive shading: India CTV has significant unsold inventory at non-prime hours (11pm–7am) and in niche genre/language segments (regional language FAST channels in smaller languages). Bid shading algorithms find historical clearing prices well below input bids and shade aggressively — effective CPMs can be 30–50% below input CPMs on this inventory.
- Floor prices constrain shading on publisher direct deals: JioHotstar and SonyLIV set meaningful floor prices on their premium content — particularly sports. Bid shading cannot shade below floor, so on high-floor inventory, the achievable discount is limited. The practical effect: programmatic open auction CPM on premium India CTV inventory often looks similar to direct IO CPM because floors prevent deep shading.
- Limited historical data for new inventory types: FAST channels and new publisher CTV inventory in India have less auction history than established AVOD inventory. Bid shading algorithms are less accurate on thin-history inventory and may shade aggressively, causing win rate drops, or conservatively, causing overpayment.
When bid shading hurts delivery
Bid shading is not always a benefit. It can reduce delivery when:
- Campaign pacing is falling behind: If an automated bidding strategy is behind on delivery, bid shading may be shading too aggressively on competitive inventory, causing the campaign to lose impressions it would have won at the unshaded bid. Raising the Target CPM or switching to first-price manual bidding temporarily can unlock delivery.
- Floor prices are high relative to input bid: If the publisher floor is ₹200 and the buyer inputs ₹210, there is almost no room for shading. The algorithm submits ₹210 (or very close to it) regardless of historical clearing data.
- New campaign with no history: Bid shading relies on historical data. A new campaign on a new publisher placement has no win/loss history. The algorithm starts conservatively and calibrates over the first few hundred auctions — effective CPM may be inefficient in the first 24–48 hours of a new campaign.
Practical guidance for India CTV buyers
- Use automated bidding (Target CPM in DV360, Koa in TTD) as the default for programmatic CTV — bid shading is built in and continuously optimised.
- For PMP and programmatic guaranteed deals where you have a negotiated fixed price, manual CPM bidding at the deal price is more appropriate — bid shading on a fixed-price deal achieves little benefit.
- Monitor effective CPM vs input CPM in reporting. A large gap (effective CPM 40%+ below input) signals that your input bid is significantly above clearing price — you can lower the input bid without losing delivery.
- During IPL and major cricket, expect bid shading to provide minimal discount on live sports inventory. Plan budget at close to input CPM for sports tentpoles.