Google Ads Auction Insights: How to Read and Use the Report

Auction Insights is one of the few places inside Google Ads where you can see something about your competitors instead of just your own account in isolation. It's also one of the most commonly misread reports, mostly because people open it once, glance at a table of unfamiliar percentages, and close it without understanding that a single snapshot tells you almost nothing on its own.

This covers what the report actually shows (and the specific things it never will, no matter how you filter it), how to read each metric in a way that leads to an actual decision rather than just an observation, and how a change here can be an early warning sign for something worth investigating more closely, including traffic quality.

None of this requires a paid tool to start — everything described here is available inside a standard Google Ads account, and the discipline of checking it on a schedule matters more than any add-on would.

What Auction Insights actually shows you

The report compares your performance against other advertisers competing in the same auctions, at the keyword, ad group, or campaign level, depending on where you run it from. It's aggregated and relative — every number is expressed as a rate or a percentage relative to the auctions you were both eligible for, not as an absolute count.

What it never shows, regardless of how you filter or segment it: competitor bids, competitor budgets, their keyword lists or match types, their Quality Scores, their ad copy or landing pages, or anything about their conversions, CPA, or ROAS. Any tool or guide claiming to extract competitor budget or bid data from Auction Insights is describing something the report structurally cannot provide — it wasn't built to expose that, and no amount of clever filtering changes what data Google includes in it.

Coverage differs by campaign type. Search gives you the full picture — keyword, ad group, or campaign level, segmentable by time and device. Shopping is limited to the ad group and campaign level. Performance Max reports only at the campaign and account level, which is thinner than either, but it's still more visibility into PMax competition than existed before Google added it.

It's also worth being clear about scope: the report only covers Google Search and Shopping auctions, since that's where the auction mechanic it's built on actually applies. It has nothing to do with Display Network placements, YouTube ad competition, or organic search rankings — those are separate systems with their own (largely invisible, or differently-visible) competitive dynamics.

Three ways to open the report

From Campaigns, Ad groups, or Keywords: check the box next to a specific row, then click Auction insights in the toolbar that appears above the table. This is the route for a targeted question — who's fighting you on one specific product line or one high-value keyword. Keywords can only be selected from a single page at a time unless you run the report for all of them at once.

From Insights and reports in the left-hand menu: this opens the account-wide view directly, which you can then narrow with filters. This is where a routine check should start, because it surfaces account-wide patterns before you drill into any one keyword — starting with a drill-down first means you might miss a broader shift that isn't concentrated on the one term you happened to check.

From a saved or scheduled report, if you're checking this regularly enough to want it delivered automatically rather than pulled manually each time — worth setting up if you're tracking more than a couple of competitors across multiple campaigns.

Whichever route you use, give yourself enough date range to matter — a 7-day pull is too short to distinguish a real shift from ordinary daily noise, the same limitation covered in more detail in the trends-over-snapshots section further down. Default to 30 days for a routine check and extend to 90 when you're specifically trying to confirm whether something is a sustained trend.

The six metrics, and what to actually do with each one

Impression share: how often your ad showed, compared to how often it was eligible to show. Falling impression share with rising cost usually means you're losing to competitors on Ad Rank, not on availability. Falling impression share with a flat cost usually points at a budget cap instead — check "impression share lost (budget)" specifically to confirm which one you're looking at before changing anything.

Overlap rate: how often another advertiser showed in the same auctions as you. This is the simplest way to identify who your real paid-search competitors are, which is not always the same list as who you'd assume based on brand recognition — a smaller, less-known competitor with an aggressive bidding strategy can show up with a higher overlap rate than a household-name competitor who's barely bidding on your terms at all.

Position above rate: how often another advertiser's ad appeared above yours when you both showed. A rising position-above rate from one specific competitor, sustained over several weeks, means they're actively investing in beating you on those terms specifically — that's a signal worth acting on, not one to note and move past.

Top of page rate and absolute top of page rate: how often your ad appeared above the organic results at all, and how often it appeared in the very first paid position specifically. A high absolute top of page rate looks good on paper but is worth checking against conversion rate before celebrating it — position one is often the most expensive position in the auction, and if conversion rate there isn't meaningfully better than position two or three, you may be paying a premium for visibility that isn't converting proportionally better.

Outranking share: how often your ad ranked higher than a specific competitor's, or showed when theirs didn't show at all. This is the most useful single metric for tracking one specific rival over time, since it condenses the relationship into one number you can watch trend up or down week over week.

One connection worth making explicitly: impression share and outranking share can move in different directions for the same competitor, and when they do, it's informative rather than contradictory. Rising outranking share against a specific rival while your overall impression share holds steady means you're beating that one competitor more often specifically, even while the broader competitive field around you stays roughly the same — a genuinely good sign, and one that a glance at impression share alone wouldn't have shown you.

The April 2025 double-serving update changed how to read this

Google updated how it handles cases where the same advertiser could serve multiple ads in a single auction (through Performance Max and Search overlapping, for instance). The practical effect: your own impression share, or a competitor's, can shift meaningfully without either of you actually changing a bid, budget, or targeting setting — the shift can come purely from how the auction itself now allocates and counts those overlapping placements.

Concretely: if you noticed your impression share drop by several points in a reporting period and went looking for what you changed, and found nothing, this update is a real possible explanation, not evidence that the report itself is broken or unreliable. It means a short-term shift, especially one appearing right around a Google-side policy or auction change, deserves a beat of skepticism before you assume a competitor did something new — check whether the timing lines up with a platform-level change before reacting to it as if it were competitive activity.

Read trends, not snapshots

A single pull of this report, for one week, tells you almost nothing reliable on its own. The value is entirely in the movement over time. Pull the report segmented by week or month and track it that way, not as a one-time check.

A competitor who's "always" been in the report at a stable overlap rate is background noise — expected, not alarming. A competitor whose overlap rate and position-above rate are both climbing steadily over four to eight weeks is a genuine, timed threat, and worth a specific response: reviewing your Quality Score and bids on the contested keywords before they erode your position further, rather than after.

A brand-new domain appearing in the report for the first time, with a modest but present overlap rate, is a new entrant worth watching rather than reacting to immediately. Give it two to four weeks of data before deciding whether they're a serious, sustained competitor or a short-lived test campaign that will disappear on its own.

A worked example of reading the report

Here's how the metrics combine into an actual read, using illustrative numbers rather than a specific real account. Say you pull the report for your top three keywords over the last eight weeks and see this pattern for one competitor: overlap rate moving from 22% to 41%, position-above rate moving from 15% to 38%, and your own impression share on those same keywords sliding from 61% to 47% over the same period.

Read individually, each number is just a data point. Read together, over eight weeks, this is a competitor who wasn't seriously contesting these keywords two months ago and now is — both showing up more often (overlap) and winning more of those head-to-head matchups when they do (position-above), which is directly costing you impression share on terms that used to be comfortably yours.

The response this pattern calls for is specific: check your Quality Score on exactly these keywords first, since a Quality Score erosion on your end would make you more vulnerable to exactly this kind of competitive pressure and is often cheaper to fix than out-bidding someone. If Quality Score is stable and healthy, the more likely explanation is a genuine bid increase on their side, and a considered bid adjustment on the specific contested terms — not an account-wide increase — is the more targeted response.

What this example is not: a single week where that competitor's overlap rate happened to be 41% instead of 22%, with no trend behind it. That's noise, not signal, and reacting to a one-week blip the same way you'd react to an eight-week trend is one of the more common mistakes people make with this report.

A decision framework for what you find

Rising overlap and position-above rate from one competitor: this is active escalation. Check your Quality Score and ad relevance on the contested terms first — that's usually the cheaper fix compared to simply outbidding them, and it also improves your cost-per-click on every other keyword in the same ad group.

High impression share lost to rank: this points at Ad Rank, which is a combination of bid, Quality Score, and expected ad impact — not budget. Fix relevance and landing page experience before raising bids, since a higher bid on a low-relevance ad is a more expensive way to solve a problem that better ad copy or a better-matched landing page might solve for less.

High impression share lost to budget on a campaign that's converting well: this is directly recoverable volume, not a competitive problem at all. Reallocate budget from a weaker-performing campaign, since the demand and conversion rate are already proven here — you're just capped from capturing more of it.

Low absolute-top-of-page rate specifically on your own brand term: a competitor is bidding on your own brand name and sitting above your ad on it, which is worth a direct brand-defense response — strengthening your bid and ad relevance on your own name specifically, since losing position on your own brand term is a different, more urgent problem than losing position on a generic category term.

What a sudden shift here can mean beyond bidding

A sharp, unexplained drop in your impression share, alongside a budget that's disappearing earlier in the day than it used to, is worth cross-checking against your invalid click rate before assuming it's purely a competitive bidding shift. If a source is manually or repeatedly clicking your ads without any intention to convert, your budget exhausts faster during the day, which mechanically reduces the remaining auctions you're eligible to compete in for the rest of that day — showing up in Auction Insights as lost impression share that looks, at a glance, like a competitor simply outbidding you.

This isn't the most common explanation for a drop, and it shouldn't be the first thing you check — genuine competitive pressure and Ad Rank issues are more common causes, and the double-serving update covered above is a real factor too. But if you've ruled those out and the numbers still don't add up, checking invalid click rate and Ads-vs-Analytics session gaps alongside the Auction Insights drop is worth the extra ten minutes before concluding it's purely a bidding war.

Pairing this with your budget and traffic audit

Auction Insights is most useful read alongside the account-level checks covered in a full audit, not in isolation. Impression share lost to budget, checked here, is the same underlying number worth cross-referencing against your invalid click rate and search-terms waste — if a campaign is losing impression share to budget while also showing an invalid click rate trending upward, the two are plausibly connected: wasted spend earlier in the day is directly what caps the budget available for the rest of the day's auctions.

Treat a sustained Auction Insights shift as one input into a broader monthly review, alongside conversion tracking health and traffic quality, rather than something you investigate in a vacuum disconnected from the rest of the account. A competitor genuinely escalating and a budget quietly leaking to invalid clicks can produce a very similar-looking impression-share chart — the audit habit is what tells you which one you're actually looking at.

Reading Auction Insights for Performance Max specifically

PMax reports at the campaign and account level only — no keyword or ad group breakdown, since PMax doesn't organize around keywords the way Search campaigns do. That means you're reading broader movement rather than pinpointing exactly which search term triggered a shift, which is a real limitation compared to Search reporting, not a bug in how you're pulling the report.

The underlying auction data for a PMax campaign still gets categorized into Search or Shopping insights depending on where the specific auction happened, so when you're reviewing PMax competitive data, check both categorizations rather than assuming one view gives you the complete picture. A competitor who's aggressive on Shopping placements specifically but not on text-ad Search placements will show up differently depending on which view you're looking at.

Because PMax pulls from a wider inventory (Search, Shopping, Display, YouTube, Discover, Gmail, and Maps, depending on your asset groups), a shift in your PMax impression share is more often explained by a change in your own asset group quality or audience signals than by a single identifiable competitor — worth checking your own asset completeness before assuming a competitive shift is entirely the cause when the campaign is PMax rather than Search.

If you're checking this across a Google Ads manager account

Auction Insights has to be pulled per account — there's no single cross-account rollup inside Google Ads itself, even under an MCC. If you're tracking a specific competitor who operates in multiple markets or categories you manage separately, that means checking each relevant account individually and keeping your own record of the trend across them, since Google won't do that aggregation for you.

This is worth doing deliberately rather than incidentally, especially if the same competitor shows up across several of your accounts — a pattern that's only visible if someone is actually comparing notes across accounts rather than treating each one as fully independent.

A note on the 10% impression share floor

If your activity on a given campaign, ad group, or keyword falls below roughly 10% impression share for the period selected, Google withholds Auction Insights data for that selection entirely — you'll see the report come back empty or heavily restricted, not because nothing happened, but because your own participation was too limited to report on reliably. This mostly affects newer or lower-budget campaigns, and it's worth knowing about so a blank report doesn't get mistaken for a technical glitch.

There's no way to lower this threshold or request an exception — it's a fixed data-reliability floor Google applies uniformly, and the practical workaround is simply widening the date range or the selection (moving from one keyword up to the full ad group or campaign) until you're above the floor and the report populates normally.

Common mistakes when reading this report

Reacting to a single week's numbers is the most common one, covered above — a report pulled once and acted on immediately is more likely to produce a wrong decision than a genuinely useful one, since normal week-to-week variance can look identical to the start of a real trend until you have enough data points to tell them apart.

Treating every advertiser in the list as an equal threat is another. A report often surfaces five, ten, or more overlapping advertisers, and most of them are irrelevant to your actual business — different sub-market, different intent, or just occasional overlap on one broad keyword you both happen to bid on. Focus tracking effort on two or three competitors who are genuinely relevant to your specific offer, not the full list.

Assuming a rising position-above rate always means a competitor outbid you is a mistake worth naming specifically, because it's frequently wrong. Ad Rank is bid multiplied by Quality Score and expected impact, not bid alone — a competitor can win more auctions against you by improving their Quality Score while keeping their bid flat, which looks identical in this report to them simply spending more. Checking your own Quality Score trend on the same keywords is how you tell the difference, and it changes which fix is actually correct.

Ignoring the report because "we don't have real competitors online" is a mistake too, in the other direction. Auction Insights frequently surfaces competitors advertisers didn't know were bidding on their terms at all — smaller, newer, or less brand-recognized companies that wouldn't come to mind if you were listing competitors from memory rather than from actual auction data.

Can Auction Insights show me a competitor's actual ad copy?

No — it never has and doesn't now. For that, use Google's Ad Preview and Diagnosis tool, which lets you see what's currently live for a given search term without clicking the ad yourself (clicking a competitor's ad to check on it wastes their budget and yours, and tells you nothing Ad Preview and Diagnosis doesn't already show safely).

Why does a competitor I recognize as a major player show a low overlap rate?

They may simply not be bidding aggressively, or at all, on the specific keywords you selected for the report. Overlap rate reflects actual auction participation on the terms you're looking at, not general brand size or market presence — a smaller, more targeted competitor can easily show a higher overlap rate on your core terms than a larger company that's focused its budget elsewhere.

How often should I check Auction Insights?

Monthly for a routine competitive check is reasonable for most accounts. Weekly is worth it specifically while you're actively watching a competitor whose numbers are already trending in a concerning direction, since a weekly cadence catches the inflection point sooner than a monthly one would.

Does a higher position in Auction Insights always mean more conversions?

No, and this is worth checking directly rather than assuming. Position and conversion rate don't always move together — a higher position generally gets more clicks, but if those additional clicks come from a less qualified part of the audience (people scanning results quickly rather than comparison-shopping carefully), conversion rate at the top position can actually be lower than at position two or three for some categories. Check conversion rate by position, where your account gives you that visibility, before assuming top position is automatically the goal.

Why does Auction Insights sometimes show different competitors than I expected?

The report only includes advertisers who were eligible and entered the auction for the specific keywords, ad groups, or campaigns you selected — not every company you'd consider a competitor in the broader market. A company you think of as a direct competitor might not bid on Google Ads at all, or might target a different set of keywords than the ones you pulled the report for, which is enough to keep them off a report scoped narrowly to your selection.

Can I get Auction Insights data for a competitor who doesn't advertise on Google?

No — the report is built entirely from live Google Ads auction participation, so a company that doesn't bid on Google Ads at all, or doesn't bid on any of the keywords you've selected, simply won't appear, regardless of how significant a competitor they are to your business in other channels. For competitors who compete with you primarily through SEO, social, or offline channels rather than paid search, Auction Insights won't tell you anything about them at all — that's a genuinely different kind of competitive research, outside what this report is built to do.

Setting up a recurring check

The single most useful habit to take from this is turning a one-time read into a recurring one. Pick two or three keywords or ad groups that matter most to your business, pull the report for them on the same schedule every time — monthly is a reasonable default — and keep a simple running log of impression share, overlap rate, and position-above rate for your two or three most relevant competitors.

That log, built up over a few months, is worth more than any single detailed read of the report taken once. It's what turns a page of unfamiliar percentages into an early-warning system you can actually act on before a competitive shift has already cost you a quarter's worth of impression share.