First-Price Auction

A first-price auction is a sealed-bid mechanism where the highest bidder wins the impression and pays their own bid. The publisher (or exchange) collects the full bid amount, minus any platform fees, and the advertiser’s cost is exactly what they submitted.

In the OpenRTB protocol, first-price auctions are signaled with auction type at=1. This flag tells the bidder that the clearing price is their own bid, not the second-highest. The transition from second-price (at=2) to first-price happened rapidly around 2017–2019 as publishers and exchanges sought to eliminate the revenue leakage caused by bid shading on the buy side.

Key characteristics

• Winner pays own bid – no discount to second price.
• Bid shading emerged as a necessary optimization: advertisers submit bids lower than their true valuation to avoid overpaying, but risk losing the auction.
• Floor prices still apply – if the winning bid is below the floor, the impression may go unsold or be re-auctioned.
• Transparency is higher for publishers (they see the exact clearing price), but advertisers must manage their own bid strategy more carefully.

So what? First-price auctions simplify the publisher’s revenue math but force advertisers to invest in bid shading algorithms and frequent calibration. The platform’s role is to provide the auction type signal (at=1) and enforce floors consistently.

Clearing price = winning bid (subject to floor and fees)

If the winning bid is below the floor price, the auction does not clear (no sale). If the winning bid is above the floor, the advertiser pays their bid, and the publisher receives the bid minus any platform fee (e.g., 10–20% ad tech tax).

Caveats

1. Floors override: If the highest bid is $2.00 but the floor is $2.50, the impression is not sold (or a second-price floor rule may apply depending on exchange configuration).
2. Platform fees: The bid amount is gross; the net revenue to the publisher is bid × (1 – fee rate).
3. Bid shading: Advertisers may submit a shaded bid (e.g., $2.10 instead of $3.00) to avoid overpaying. The clearing price is still the shaded bid, not the true valuation.
4. Multiple price floors: Some exchanges apply different floors for different deal types (open auction vs. private marketplace). The highest applicable floor is used.

When you see a first-price auction metric (e.g., clearing price or winning bid), interpret it as the actual cost paid by the advertiser, not a theoretical second-price discount. A high clearing price may indicate strong competition, but it could also mean the advertiser is overpaying due to poor bid shading.

What to pair it with

• Win rate: If win rate is high and clearing price is high, the bid strategy may be too aggressive. If win rate is low and clearing price is low, the bid may be too conservative.
• Bid-to-clearing ratio: Compare the submitted bid to the clearing price. A ratio near 1.0 suggests no shading or very tight competition; a ratio well above 1.0 indicates the advertiser left money on the table.
• Floor price: If clearing price consistently equals the floor, the auction is likely floor-constrained, not competitive.

Common dashboard misread: A publisher sees a high eCPM and assumes strong demand. In a first-price auction, that eCPM is the exact bid – but if the floor is set too low, the publisher may be leaving revenue on the table because the advertiser shaded down to just above the floor.

For advertisers (buy side)

• Bid shading algorithm: The most direct lever. A well-calibrated shader reduces the bid while maintaining a high win rate. Over-shading drops win rate; under-shading inflates cost.
• Frequency and recency of bid requests: More data on the same user allows better valuation and shading. Low data density leads to conservative (high) bids.
• Competitive density: More bidders in the auction push clearing prices up. Advertisers can respond by segmenting audiences or using deal IDs to bypass open auction.

For publishers (sell side)

• Floor price: Raising the floor increases the minimum clearing price but may reduce fill rate. Lowering the floor increases fill but may lower eCPM.
• Number of demand sources: More bidders increase competition, which raises clearing prices. But too many sources can increase latency.
• Auction type configuration: Some exchanges allow hybrid models (first-price with second-price floor rules). Check your SSP’s documentation.

Tradeoffs

• Fill vs. price: A high floor guarantees a minimum price but may leave inventory unsold. A low floor fills more impressions but at lower revenue. The optimal floor depends on demand elasticity.
• Bid shading aggressiveness: Aggressive shading saves money per win but reduces win rate. The advertiser must balance cost per acquisition (CPA) against reach.
• Latency vs. competition: Adding more bidders increases competition (good for price) but adds latency (bad for user experience and viewability).

First-price auctions reward strategic bidding and punish naive honesty. Advertisers must invest in bid shading; publishers must set floors carefully.

• Advertisers: Always shade bids; calibrate using win-rate data; segment by floor and exchange.
• Publishers: Use floors as anchors, not ceilings; monitor fill rate and bid density.
• Platforms: Signal auction type correctly (at=1) and enforce floors transparently.