Most advertisers open Google Ads Auction Insights once, see a table of percentages, feel vaguely informed, and never open it again. That's a waste, because Auction Insights is the only place Google will ever tell you who you're actually bidding against. Not who you think your competitors are. Who is literally in the same auctions, taking impressions you were eligible for. Read it right and it's an early-warning system: new entrants show up here weeks before you feel them in your CPCs, and a competitor quietly pulling back shows up as free market share nobody else noticed. This guide covers what each metric means in plain language, how a media buyer actually reads the table, and what it can't tell you.
What Auction Insights is and where to find it
Auction Insights is a report inside Google Ads that compares your performance against other advertisers who participated in the same auctions as you. Google builds it from actual auction data, so every name in the table is a confirmed head-to-head competitor for the scope you selected.
You can pull it at three levels, and the level changes the story:
- Campaign level: the broad view. Good for the monthly sweep.
- Ad group level: tighter theme, tighter competitor set.
- Keyword level: the sharpest view. Select specific keywords and see exactly who fights you for them. This is where the interesting decisions live.
To find it: open Campaigns (or Ad groups, or Keywords), select the rows you care about, and click Auction insights in the menu above the table. It's available for Search and Shopping campaigns, with a more limited view for Performance Max. One prerequisite: the report only populates where there's enough auction activity, so brand-new campaigns and tiny keywords may show nothing yet, and very low values often display as "< 10%" rather than a precise figure. Google's own documentation on the report lives in the Google Ads Help Center.
Every Auction Insights metric in plain language
Six metrics, no jargon:
| Metric | What it actually means |
|---|---|
| Impression share | Of all the auctions you were eligible to show in, the percentage where you actually got an impression. Your share of the available market. |
| Overlap rate | How often that competitor showed up in the same auctions as you. High overlap means you're fighting them constantly, not occasionally. |
| Position above rate | When you both showed, how often their ad sat above yours. This is the head-to-head "who's winning" number. |
| Top of page rate | How often ads (yours or the competitor's, per row) showed anywhere above the organic results. |
| Absolute top of page rate | How often the ad was the very first result on the page. The most expensive real estate in search. |
| Outranking share | How often you ranked above that competitor, or showed when they didn't show at all. The single cleanest "am I beating them" summary. |
A useful way to hold it in your head: impression share is your size, overlap rate is how often you collide, position above rate is who wins the collision, and outranking share is the season record.
How a media buyer actually reads the table
The metrics are static definitions. The skill is reading changes over time. Here's the practical playbook.
Spot new entrants before they hurt. A name you've never seen appearing with meaningful overlap rate is a new bidder in your auctions. This is often the first observable evidence of a funded competitor's launch, it shows up in Auction Insights before it shows up in your cost-per-click trend. When a new name appears, go look at what they're running: check their ads and landing pages in the Google Ads Transparency Center and start a proper file on them.
Detect a competitor going aggressive. The tell is a rising combination: their impression share climbing, position above rate climbing against you, and absolute top of page rate climbing. That pattern means more budget or higher bids, usually both. If your own impression share is falling at the same time, you're the one being pushed out, and it's better to know from the table than to discover it in a bad month-end report.
Detect a pull-back. The reverse pattern is a gift. A competitor whose impression share is sliding across consecutive periods is cutting budget, shifting channels, or dying. Their abandoned impressions don't disappear, they get redistributed to whoever's still bidding. Buyers who track this catch windows where top-of-page positions get cheaper for a while. If a name vanishes from the report entirely, they've left your auctions altogether.
Find the auctions they abandoned. Run the report at keyword level and compare competitor presence keyword by keyword. When a rival with high overlap on your head terms is absent from a cluster of adjacent keywords, that's a hole in their coverage, either they tested those terms and quit, or never saw them. Both are worth probing with a modest bid before they notice.
Diagnose, don't just observe. Auction Insights tells you who is beating you, not why. Pair it with your own account data: if a competitor's position above rate is high on a keyword where your Quality Score is weak, the fix might be your ad relevance or landing page, not your bid. The table starts the investigation; your account data finishes it.
The limits nobody mentions
Auction Insights is honest data with hard edges, and pretending otherwise leads to bad conclusions:
- It only shows advertisers in your auctions. If you're not bidding on a keyword, you learn nothing about who's competing there. The report can't map a market you haven't entered, for that you need outside-in competitor keyword research.
- No spend, no bids, no budgets. Percentages only. A rising impression share implies more aggression, but the report will never say what anyone pays. Any tool claiming to know a competitor's exact Google spend from auction data is guessing, the number isn't published.
- No creative, no landing pages. You see that a rival outranks you, not what their ad says. Pair the report with the Transparency Center for the qualitative half.
- Aggregated and delayed, not real-time. Fine for weekly and monthly reads, useless for day-trading your bids.
- Low-volume rows get masked. "< 10%" is a floor, not a measurement.
None of this makes the report weak. It makes it one instrument among several, the same way competitor keyword research covers the auctions you're not in yet.
A monthly reading cadence that takes 20 minutes
Consistency beats depth here. Once a month:
- Pull Auction Insights at campaign level, same date range each time (last 30 days), and export it.
- Diff against last month's export: new names in, old names out.
- For your top 3 competitors, note the trend in impression share, position above rate, and absolute top of page rate. Rising, flat, or falling is all you need.
- Drill to keyword level on your money keywords, check who's present and who's missing.
- Write three lines: who's new, who's aggressive, who's fading. Decisions come from three months of these notes, not from one snapshot.
The exported diffs matter because the interface shows you a window, not a memory. Like most competitive data, the value compounds only if someone keeps the history.
Reading your auctions from chat with AdWhispr
Everything above is doable in the Google Ads interface. It's also exactly the kind of repetitive read-and-diff work that agents are good at now. AdWhispr's MCP server connects your Google Ads account to Claude, ChatGPT, Cursor, or Claude Code: get_account_performance surfaces your account's numbers with per-keyword diagnostics, so "which keywords am I losing position on, and is it a bid problem or a quality problem?" becomes a question you ask in a sentence rather than a pivot table you build. research_competitor_keywords covers the report's biggest blind spot by showing what rivals bid on outside your current auctions. And when the analysis says act, the same server launches real Search and Performance Max campaigns from the chat. The full setup is in our Google Ads MCP guide, and if you live in Claude, start with how to run ads from Claude.
FAQ
Where is Auction Insights in Google Ads?
Open Campaigns, Ad groups, or Keywords in your Google Ads account, select the rows you want to analyze, and click "Auction insights" above the statistics table. It's available for Search and Shopping campaigns, with a limited version for Performance Max.
What is a good impression share?
There's no universal number, it depends on your budget, your goals, and how contested the auction is. The useful reading is directional: falling share on keywords you care about means you're being outbid or out-scored, and rising share means you're consolidating. Chasing a high share on every keyword is a recipe for overpaying on terms that don't matter.
Does Auction Insights show competitor spend or bids?
No. It shows only comparative percentages: impression share, overlap rate, position above rate, top of page rates, and outranking share. No spend, bid, budget, or clicks data for any competitor is exposed, and no third-party tool can read those numbers out of the auction either.
Why don't I see a competitor I know is advertising?
Two common reasons: they're not participating in the same auctions as you (different keywords, geos, or schedules), or there isn't enough overlapping activity to meet the reporting threshold. Auction Insights only reflects advertisers who actually collided with you in auctions during the selected date range.
Auction Insights is one of the few places in advertising where the competitive data is real rather than modeled. The work is in reading it regularly and pairing it with what it can't see. AdWhispr handles both sides: per-keyword diagnostics on your own account, competitor keyword and ad-library intelligence on everyone else's, and campaign launch when you're ready to act, all from one chat. Connect at https://adwhispr.com/api/mcp or run npx adwhispr-mcp-server config, the free plan needs no credit card.
Stop skimming the percentages, start reading the auction: diagnose your account with AdWhispr.