Every time you search on Google, a request is sent to Google servers to evaluate advertisers bidding for that search query. All eligible advertisers are then put into an auction where their ad positions are determined based on the Max CPC bids, Ad quality, and other contextual factors.
Ad with the highest Ad Rank is placed at the top position and receives the most clicks. The average CTR of an Ad at the first position is 7.94% whereas the last Ad above organic results has an average CTR of 1.95% - a drop of 75.4% compared to the first position and we're still talking about results on the first page above organic results. As we move beyond the top 4 Ads, the CTR drop is logarithmic.
What is Google Ad Auction?
Google Ads (formerly AdWords) is an auction where advertisers bid on keywords that are relevant to their business in order to compete to show up in Google search results or partner websites.
Unlike a typical auction where the highest bidder ends up winning, a Google Ad Auction ranks eligible advertisers based on their Ad Rank.
Ad Rank = Ad Quality x Max CPC Bid
Ad Rank also takes into account the context of the Ad - location, device and the expected impact of ad formats (text, responsive, call-only, dynamic or shopping) and extensions - sitelink, callout, structured snippet, call, message, location, affiliate location, price, app, and promotion. Existence of one or more of these extensions has a positive impact on Ad Rank as it provides more information to the user to directly engage with such as store location, phone number, and reviews.
Ad quality estimates the expected CTR, Ad relevance, and landing page experience at auction time using a statistical model. This estimate, along with the maximum bid the advertiser is willing to pay, is used to calculate the Ad Rank.
Google filed for a patent in 2005 titled Estimating ad quality from observed user behavior, which outlines how statistical models are used to estimate the quality of unrated ads.
An excerpt from the patent about evaluating the quality of an Ad:
The multiple different first user actions comprise at least one of the following: a duration of an ad selection; a number of selections on other ads before and/or after a given ad selection; a number of selections of search results before and/or after a given ad selection; a number of selections of other types of results before and/or after a given ad selection; a number of document views before and/or after a given ad selection; a number of search queries before and/or after a given ad selection; a number of search queries associated with a user session that show ads; a number of repeat selections on a same given ad; or an indication of whether a given ad selection was the last ad selection for a given query or the last selection in a user session.
Google recalculates Ad Rank each time an ad is eligible to compete at an auction which is why you might notice ad position fluctuations throughout the day.
Note: Auction time Ad quality should not be confused with quality score that advertisers are able to see in their Google Ads account. Google has been very clear that Quality Score is not used at auction time to determine Ad Rank.
Google Ads Vs RTB
A lot of people get muddled between real-time bidding, programmatic, auction time bidding, and smart bidding. So let's get this out of the way.
Real-time bidding is a type of programmatic advertising where demand-side platforms (DSP) and supply-side platforms (SSP) participate in auctions where the cost of impression is calculated in real time. An RTB auction employs second-price model wherein the advertisers with the highest bid pays $0.01 more than the bid of the second highest bidder - similar to how it works with Google Ads.
In mid 2016, AdWords introduced Smart bidding strategy. Just like real-time bidding, smart bidding strategy optimizes bids using machine learning algorithms at auction time, for each impression.
If you're using any of the 4 smart bidding strategies in Google Ads - Target CPA, Target ROAS, Maximize conversions or Enhanced CPC, then Google's ML algorithm is bidding in real time for you.
How Is CPC Calculated?
Your actual cost per click(CPC) is measured based on your Ad Rank, quality score and how much the advertisers below you in Ad Rank is willing to pay for that keyword at auction time. So your CPC is usually much less than the max CPC bid you have set.
Max CPC bids can be breached in two scenarios -
- Enhanced CPC bid strategy is chosen to maximize conversions
- Bid adjustments have been set for - device, location, demographics, topics, placements, remarketing lists or ad schedule.
This is how your CPC is calculated:
What you pay per click = Ad Rank of advertiser below you/Quality Score + $0.01
Let's use an example to put this into perspective -
4 advertisers are bidding for the keyword "mens eyeglasses". Google assigns each of them an ad quality score based on ad relevance, landing page score and expected CTR.
So if you were the only advertiser bidding for a particular keyword in an auction, your CPC bid would not necessarily be $0.01, it would vary based on your quality score.
CPC Calculation for Display Network Auction:
Display network auction uses relative CTR of position to determine the expected incremental clicks an advertiser would receive by being in a more visible position on the website and charges advertisers based on that.
If the relative CTR of top position of a website is determined by Google to be 2.0x and the relative CTR of position 2 is 1.0x, it would mean 50% of the clicks received by being in the top position are incremental clicks. So the Advertiser would be charged more for 50% of those incremental clicks.
Ad Auction With CPA Bidding
Similar to manual bidding where the advertiser sets the Max CPC bid, target CPA bidding allows advertisers to leverage Google's smart bidding strategy, which uses machine learning algorithms to automatically optimize bids to get the most conversions for the target cost per action.
Smart bidding algorithm also has access to exclusive bid adjustments that advertisers cannot manipulate manually. These include:
- Ad Characteristics - Optimize bids to show the best version of the Ad
- Site Behavior - Optimize bids based on users behavior on your website - session duration, pages viewed, products viewed, etc
- Product Attributes - Optimize bids for Google Shopping campaigns by product attributes - price, brand, category, stock status, etc
- Actual Search Query - Optimize bids based on keyword that is more likely to convert
- Operating System - Optimize bids for people using an operating system
- Browser - Optimize bids based on the users' browser
At auction time, the smart bidding algorithm uses these real-time signals (in addition to the ones mentioned earlier) to automatically optimize bids each time.
Although advertisers can set the maximum cost per click for smart bidding strategies, Google recommends against this since it limits the algorithms ability to get the most conversions.
Improving Ad Rank
Ad Rank takes a number of factors into consideration, one of which is Ad quality. Google wants to make sure users only see the most relevant Ads, even if you set a very high Max CPC bid and your Ad is not relevant to the query, it won't show up in search results.
The most straightforward way to increase Ad Rank would be to increase your Max CPC; this would work ONLY if your current maximum cost per click is lower than competing advertisers. If you're already outbidding your competitors, it would have no effect on your Ad Rank. An advertiser that is paying much lower than you would still rank higher if their Ads are more relevant.
Google Ads looks at 3 things to calculate auction time Ad Quality
Ad Relevance - This is a measure of how closely your Ad is related to the keyword you are targeting. Google's algorithms are exceptionally good is understanding relevance and semantics between keywords - It is able to identify that "eyeglasses" and "spectacles" are closely related. So you can have the word eyeglasses in your Ad copy and target the keyword spectacle and it would not negatively affect your ad quality.
On the other hand, if your Ad copy and targeted keyword are not within the same domain - eyeglasses and mirrors, then your Ad quality would take a hit and it would be less likely to win the auction.
To ensure your Ad copy, keywords and landing page are specific to the topic, create topical ad groups with specific keywords. Ex - Setup individual ad groups for sunglasses, eyeglasses, and lenses with relevant keywords and Ads and have users land on those category or product pages.
Google Ads has recently launched Ad relevance recommendations which suggests keywords to add to your ad copy to increase relevance and boost ad quality.
Expected CTR - Google derives the expected click through rate of the ad in auction based on the performance of previous ads. While doing so, it ignores the effect of ad position since ads in higher positions tend to have higher click through rates. So the performance of your ads in auction is not determined in isolation since Google is using existing data from your account to predict click through rates of your ads in auction.
Landing Page Experience - Users expect a fluid, navigable and informative experience that satisfies their search intent once they land on a page after clicking an Ad. Google rewards advertisers who provide a good user experience by ranking those Ads higher as a consequence of having a good landing page score.
Optimize content on the landing page by providing all the core information about the business or service at the top of the page, above the fold. Interactive content such as images and videos have been found to have a positive impact on user engagement and increase dwell time.
To ensure your Ad Rank remains unaffected by ongoing changes to your website, put checks in place to monitor and redirect landing pages that have been moved or changed. Google will notify you and disable any ads associated with landing page its not able to crawl. This will also severely affect your Ad Rank and might take days to weeks to return to normal after applying the fix.
Two new metrics were added to Google Ads recently - Mobile Speed Score and Mobile Frriendly Click Rate. Earlier these were built into the algorithm but now Google is making it available so advertisers can improve landing page experience. Newly created landing pages can be tested at Test My Site to determine if it needs to be optimized before launching a campaign.
What Is Ad Rank Threshold?
In order for Ads to be eligible to show at a particular position, it must meed the Ad Rank Threshold. Let's say Google has defined the Ad Rank Threshold for Ads above organic results on the first page as 60. Any ads below that threshold would be shown at the bottom of the first page.
These thresholds are defined to ensure users only see the highest quality Ads at the top of the page.
Now that we are done with the explanations, let's run through an example to understand how it all ties in together.
Here are four advertisers participating in a Google ad auction -
|Advertiser||Ad Quality||Max CPC||Ad Rank||Position||Quality Score|
*Using arbitrary values
Google analyzes their ad copy and landing pages to estimate relevance, then uses machine learning algorithm to determine estimated click through rate based on the performance of previous ads.
Eyebuydirect with the highest Ad Rank will show up at the top position and their cost per click will be calculated by dividing the Ad Rank of the advertiser below them by their quality score + $0.01.
100/9 + $0.01 = $11.12 (Ad Rank of Advertiser next in position/QS + $0.01)
Similarly, Warby Parker being in position 3 would pay $10.51.
Advertiser with an optimized landing page, relevant keyword rich ad and perfect quality score controls the auction.