How the Google Ads Auction Works · Complete Deep Dive

Every time a user performs a Google search, Google runs an independent auction in real time. The auction takes milliseconds from query to ad serving. The process:

1. Query received. Google receives the search query along with context: user's location, device, time, search history, audience memberships, and the exact query string.
2. Eligible ads identified. Google identifies all ads whose keywords match the query (based on the advertiser's keyword match type settings and bid). Ineligible ads are filtered: paused ads, ads that have exhausted their daily budget, ads with disapproved content, ads targeting different geographies or languages.
3. Ad Rank calculated for each eligible ad. Google calculates an Ad Rank score for each eligible ad using the six-factor formula.
4. Threshold check. Ads that do not meet minimum Ad Rank thresholds are excluded from the auction results — they do not show even if no better alternative exists for that position.
5. Ads ranked and positions assigned. Eligible ads above threshold are ranked by Ad Rank. The highest Ad Rank wins position 1, next wins position 2, and so on.
6. Actual CPC calculated. The price each advertiser pays is calculated using the second-price mechanism (described below) — not the maximum bid.

Google Ads uses a modified second-price (Vickrey) auction model. In a pure second-price auction, the winner pays the minimum amount needed to beat the second-highest bidder — not their own maximum bid. Google's version is "modified" because it incorporates Quality Score into both ranking and pricing.

Why second-price auctions are used

Second-price auctions have a well-established economic property: truthful bidding is the dominant strategy. In a first-price auction (pay what you bid), advertisers are incentivised to bid below their true value to avoid overpaying. In a second-price auction, the optimal strategy is to bid exactly what a click is worth to you — overbidding wastes money on the rare case you pay more, underbidding costs you clicks that would have been profitable. This truthful bidding property produces more efficient market outcomes and reduces strategic gaming.

Ad Rank is recalculated independently for every auction. The same advertiser's Ad Rank for the same keyword can vary between auctions based on contextual signals. The six factors Google uses:

Note that the Quality Score you see in the Google Ads interface (1–10) is a diagnostic aggregate. The actual auction uses a more granular real-time calculation — the reported Quality Score is an approximation for guidance purposes, not the exact value used in each auction.

Ad Rank thresholds are minimum quality requirements that ads must meet to show in specific positions. These thresholds exist to prevent low-quality, irrelevant ads from showing even when no better alternative is available — Google would rather show no ad than a poor-quality one that damages user experience.

Thresholds vary by: position (top-of-page positions have higher thresholds than bottom-of-page), query (more competitive queries have higher thresholds), user signals (context that suggests a higher-quality result is expected), and historical performance (advertisers with consistently poor performance face higher thresholds).

The practical implication: a high bid alone cannot guarantee a top position or even any position if Quality Score is too low. An advertiser with a Quality Score of 2 may not achieve position 1 at any bid level for a competitive query — the threshold prevents it.

The actual CPC you pay when someone clicks your ad is:

Actual CPC = (Ad Rank of the advertiser below you ÷ Your Quality Score) + $0.01

This formula has a critical implication: a higher Quality Score directly reduces your actual CPC. An advertiser with Quality Score 10 paying the minimum to beat an advertiser with Ad Rank 20 pays significantly less than an advertiser with Quality Score 4 achieving the same position.

Worked example

Advertiser A wins position 1 at $1.51 despite having the lowest maximum bid — because their Quality Score of 10 produces the highest Ad Rank. Advertiser B, despite the higher bid, pays $2.41 for position 2 due to their lower Quality Score. This illustrates why Quality Score optimisation is one of the highest-ROI activities in Google Ads.

The same keyword auction produces different results for different users in the same millisecond based on contextual signals. Google factors into each auction:

• Device. Mobile, desktop, and tablet users receive different auction results. Advertisers can set device bid adjustments (e.g. +20% for mobile) that modify the effective bid for specific device types.
• Location. Geographic targeting and location bid adjustments change effective bids. An advertiser targeting "restaurant" with +50% bid modifier for users within 1km of their restaurant location competes more aggressively for nearby users.
• Time of day and day of week. Bid modifiers for time allow advertisers to bid more aggressively during business hours or on days with higher conversion rates.
• Audience membership. Users in Remarketing Lists for Search Ads (RLSA), Customer Match lists, or in-market audiences can receive modified bids — typically higher for users more likely to convert.
• Search history. Google's contextual signals include the user's recent search behaviour — a user who searched for "running shoes" three times in the past day signals higher purchase intent.

The Auction Insights report in Google Ads shows how your performance in the auction compares to other advertisers competing for the same keywords. Available at campaign, ad group, and keyword level. Key metrics:

Authentic Sources