How to read Google Ads Auction Insights
How to read Google Ads Auction Insights column by column, and which bid, budget or copy decision each impression share or overlap pattern justifies.
How to read Google Ads Auction Insights column by column, and which bid, budget or copy decision each impression share or overlap pattern justifies.
Your CPC on the money keyword crept up 30% over a month. Nothing changed in the account. You open Auction Insights, and there it is: a domain you barely know, sitting in a table with six percentage columns. You stare at "outranking share 71%" and feel briefly reassured. Then you notice they show up above you on almost every search where you both appear.
Both numbers are true. Only one of them describes the problem.
Short answer: Google Ads Auction Insights shows who entered the same auctions as your campaign and how often you beat them. It never shows their spend. Read impression share with overlap rate for presence, position above rate for ad rank, and outranking share only beside overlap rate. Then tie each pattern to one bid, budget or copy decision.
The takeaways
Google Ads Auction Insights reports six statistics for Search campaigns, defined on Google's own help page. Impression share is the impressions an advertiser got divided by the impressions they were eligible for. Overlap rate is how often their ad got an impression when yours did. Position above rate is how often their ad sat higher than yours when both showed. Outranking share is how often you ranked higher, or showed when they did not. Top of page rate and absolute top of page rate say how often an ad appeared above the organic results, or as the very first ad.
The detail people skip: the top of page columns describe whoever owns the row, so your row shows your ads and a competitor's row shows theirs. And every row exists because that advertiser collided with your campaign. A rival who bids heavily on keywords you never touch will not appear at all.
Outranking share is the most misread column because it rewards you for auctions your rival never entered. Google counts a win whenever your ad ranked higher or whenever your ad showed and theirs did not. Against a competitor with a 15% overlap rate, most of your "wins" are auctions where they were simply absent.
That is the opening scenario. A 71% outranking share looks like dominance. Pair it with a 20% overlap rate and an 80% position above rate, and the reading flips: you rarely meet this rival, and when you do, they beat you four times out of five.
So I sort the table by overlap rate first. Overlap tells me which rivals share my auctions often enough to matter, and only for those rows do the head-to-head columns mean anything. A rival at 8% overlap is background noise, however scary the other numbers look.
You can tell whether a Google Ads competitor wins on budget or on ad rank by crossing their impression share with their position above rate. High impression share with a low position above rate means they show up often but sit below you. That is a reach play funded by budget. A high position above rate means that when you meet, Google ranks them higher, which comes down to bid, ad quality and expected click-through rate.
Check your own side too. Google Ads splits your lost impression share into two columns: lost to budget and lost to rank. If most of your loss is budget, the rival is filling auctions you ran out of money for, often in the evening. If most of it is rank, you are present but outbid or outclassed on quality.
The classic mistake is raising bids against a rival who only out-funds you in the evening. Higher bids make every morning click dearer and change nothing after 6pm.
Each Auction Insights pattern points at one lever, and naming the lever before touching anything stops a panic bid increase.
Read the trend over several weeks. One week of auction data is a weather report.
Auction Insights cannot tell you what a competitor spends, bids, or scores on quality. Google's help page is explicit that the report shows none of that. It also states that insights are not shown when impression share is under 10%, so smaller rivals drop out of the table entirely. And the report needs a minimum level of activity for the selected time period, so a quiet keyword or a short date range may return nothing.
Nor can it see motive. A rival's impression share jumping from 20% to 40% might be a bigger budget, a new keyword list, or another bidder going dark. The table records the outcome only.
How much money a rival puts into search is an estimation problem with its own limits, covered in the post on how to estimate competitor Google Ads spend.
Auction Insights tells you who is in your auctions. It is silent on what they say to win them: the offer, the angle, the claim that changed the week their overlap jumped.
That second half is what Adscalr covers. It reads the Google Ads Transparency Center daily, next to Meta's and TikTok's ad libraries, and flags ads still running after 30+ days as durable winners. It does not watch search auctions, so Auction Insights stays the tool for who you are up against. For what they run, that is competitor intelligence.
This is the thinking behind Adscalr.
See the product →