How to Read the Auction Insights Report in Google Ads

If you want to know how to read the auction insights report in Google Ads, here’s the honest version: it’s the only report where Google tells you who’s actually showing up in the same auctions as you and how often you’re winning — and almost every metric in it gets misread. Impression share tells you how often you appeared when you were eligible, overlap rate tells you who keeps landing next to you, and outranking share tells you who’s above whom. Read it like a weather report — a snapshot of conditions, checked over time — not a battle plan, and it becomes one of the most genuinely useful reports in your whole account.

Okay, let’s be honest about something first: auction insights is the report people open when they’re feeling competitive and close when they’re feeling confused. I get it. It looks like a scoreboard. It is not a scoreboard. It’s closer to a weather report — it tells you what conditions look like out there, who’s in the sky with you, and whether the pressure is rising or falling. It does not tell you what to do about it. That part is yours. And that’s exactly why learning how to read the auction insights report properly is worth thirty minutes of your life: once you know what each metric actually means (and what it absolutely does not mean), you stop panic-bidding every time a competitor’s name climbs a row, and you start making calm, grown-up decisions with real information.

What is the auction insights report, exactly?

The auction insights report shows you which other advertisers participated in the same ad auctions you did, and how your performance compared on a handful of visibility metrics. You can view it scoped to a whole campaign, a single ad group, or even an individual keyword — and that scoping matters more than people realize, because a blended account-level read mixes your brand campaigns, your money keywords, and your experiments into one misleading soup.

Here’s the part nobody tells you: the report is privacy-shaped. Google only shows advertisers who met certain activity thresholds in your auctions, and the whole thing is based on a sample of auctions where you participated — not a census of every auction in your market. If an advertiser barely overlapped with you, they may not appear at all. If you barely ran, your own numbers get thin and wobbly. So the first rule of reading auction insights is accepting what it is: a partial, honest picture of your corner of the sky. Not the whole atmosphere.

And one thing it will never show you: search terms. You see who shared your auctions, never which queries triggered the overlap. Keep that limitation in your back pocket — it explains a lot of the weirdness we’ll get to.

Where do you find auction insights in Google Ads?

As of this writing, you’ll find it inside your campaign or ad group views — look for an “Auction insights” option in the reporting area, or select specific campaigns, ad groups, or keywords and open auction insights for just that selection. Google reshuffles its interface regularly (I promise it’s not just you), so if the menu has moved since I wrote this, search “auction insights” in the Google Ads search bar at the top of the interface and it’ll take you straight there. The report itself has looked essentially the same for years, even as the navigation around it plays musical chairs.

One practical tip before you open it: decide what you’re reading first. “Auction insights for my top three non-brand campaigns, last 90 days” is a read. “Auction insights for everything, all time” is a mood.

How to read the auction insights report metric by metric

This is the spine of the whole thing. Five metrics (well, six-ish), each with a classic misreading that I’d love to retire forever. Let’s take them one at a time.

Impression share: how often you showed up when you could have

Impression share is the percentage of auctions you were eligible for in which your ad actually appeared. If your impression share is 40%, you showed up in roughly four out of ten auctions where you could have competed.

The classic misreading: “My impression share is low, so I’m losing.” No, sweetheart. Low impression share with a good cost per acquisition might mean you have headroom — profitable demand you haven’t bought yet. Or it might mean you’re deliberately focused: tight budgets, tight targeting, cherry-picking the auctions you want. A boutique advertiser with 25% impression share and healthy margins is doing better than a bleeding one at 90%.

The honest read: impression share is a context metric, not a grade. And the real gift hiding next to it is the diagnostic split — in your campaign columns (not the auction insights table itself), Google breaks out impression share lost to budget versus impression share lost to rank. That split tells you which lever the gap points at:

Same symptom, two completely different treatments. That split is the single most actionable thing adjacent to this report, so check it before you touch anything else.

Overlap rate: who keeps showing up next to you

Overlap rate is how often another advertiser’s ad appeared in the same auction as yours. If Competitor A has a 60% overlap rate, they showed up alongside you in six out of ten of your sampled auctions.

The classic misreading: “High overlap means they’re my direct competitor.” Not necessarily — and this is my favorite correction in the whole report. Auction-mates are not always business rivals. I call it the broad-match stranger problem: if you’re bidding broadly, you share auctions with insurance companies, job boards, coupon sites, and giant retailers who sell ten thousand things, one of which vaguely resembles yours. A fictional example to make it concrete: imagine “Maple & Main,” a made-up local furniture store, opens auction insights and finds a national marketplace at 70% overlap. That marketplace is not plotting against Maple & Main. Their enormous broad keyword net simply drags them into the same sky. Panicking about them would be like a sailboat declaring rivalry with a cargo ship.

The honest read: high overlap tells you who shares your auctions, which is genuinely useful for understanding the terrain — especially when the name is a true rival. But confirm business overlap before you treat anyone as competition. That confirmation work is its own discipline, and it pairs beautifully with a proper PPC competitor analysis built on public data — their actual ads, their landing pages, their positioning — rather than inference and vibes.

Outranking share: the pecking order

Outranking share is how often your ad ranked higher than theirs in the same auction — or showed when theirs didn’t. It’s the closest thing this report has to a head-to-head record.

The misreading: treating it as a trophy to maximize against every name on the list. The honest read: it’s a pecking-order snapshot for the auctions you shared. Outranking a true rival on your highest-intent keywords matters. Outranking the broad-match cargo ship matters not at all. And remember rank is bought with money and relevance together — a high outranking share achieved purely through bids can be an expensive way to feel good.

Position above rate: when you both showed, who was higher?

Position above rate narrows the question: in auctions where both of you appeared, how often were they above you? It’s the head-to-head within the head-to-head. If a true competitor shows above you most of the time you share a page, that’s a real signal about relative Ad Rank on contested terms — worth noting, worth tracking, and worth exactly zero panic.

Top of page rate and absolute top of page rate: position reality

These two tell you how often an advertiser’s ad showed near the top of the results (top of page) and in the very first position (absolute top). The metric names and definitions here are the kind of thing Google has tweaked before, so glance at the little tooltip definitions in your own account rather than trusting any article — including this one — as gospel.

The honest read: position is not everything. Absolute top placements generally cost more per click, and for plenty of businesses the math on position two or three is simply better. A competitor living at absolute top isn’t automatically winning; they’re paying for altitude, and you can’t see their margins from here. If your conversions are healthy from slightly lower positions, that’s not losing — that’s efficiency.

The metrics decoder table

Metric	What it actually means	The classic misread	The honest read
Impression share	% of eligible auctions where you appeared	“Low share = losing”	Low share + good CPA can mean headroom or deliberate focus; check lost-to-budget vs. lost-to-rank
Overlap rate	How often they appeared alongside you	“High overlap = direct competitor”	Auction-mates aren’t always rivals — broad match invites strangers
Outranking share	How often you ranked above them (or showed when they didn’t)	“Maximize it against everyone”	A pecking-order read; only meaningful against true rivals on terms you care about
Position above rate	When you both showed, how often they were higher	“They’re above me, raise bids”	A relative Ad Rank signal to track over time, not a trigger
Top / absolute top rate	How often an ad showed near or at the very top	“Absolute top = winning”	Altitude costs money; lower positions often convert more profitably

How do you read the auction insights report over time?

Here’s the discipline that separates calm advertisers from twitchy ones: one report is weather, the series is climate. A single snapshot tells you it rained Tuesday. The series tells you whether your region is getting wetter. Every workflow below is built on that idea.

The quarterly competitive read

Once a quarter, pull auction insights for your core campaigns and ask three questions — in this order:

• Who’s new? A name you’ve never seen, now sitting at meaningful overlap. New entrants are the most important weather change there is.
• Who left? A familiar name that faded or vanished. Maybe they cut budget, repositioned, or went out of business — you can’t know which, but the absence itself is information.
• Who’s climbing? A steady name whose impression share or outranking numbers have risen across two or three consecutive reads. That’s a trend, and trends deserve attention in a way snapshots never do.

Save the numbers somewhere — a simple spreadsheet with one row per competitor per quarter is plenty. The report doesn’t keep history for you in any convenient way; your spreadsheet becomes the climate record.

The brand-term check

Run auction insights on your brand campaign separately. The question is simple: who is bidding on your name? Sometimes the answer is nobody, which is lovely. Sometimes it’s a rival camped on your brand terms, which changes your defensive math. This read pairs directly with your broader competitor homework — and if someone is on your name, resist the urge for instant retaliation. Watch for a read or two first; a one-week test by a competitor’s agency is not a declaration of war.

The segment move: dates and devices

The report lets you segment — by time and by device — and this is where the hidden characters show up. Let me give you a fictional, clearly-labeled example: imagine “Harbor Supply Co.,” a made-up outdoor gear shop, whose blended auction insights look stable. Segmented by device, though, a competitor appears with high overlap on mobile only. Segmented by day, another name surges on weekends. Neither pattern is visible in the blended view. The competitor who only fights on mobile, or only on weekends, is telling you something about their strategy — and about where your own defenses are thin.

The campaign-specific read

Blended reads mislead. Your brand campaign will always show gorgeous impression share; your experimental broad campaign will always look like a losing battle. Mixing them averages a fortress with a frontier and tells you nothing about either. Scope auction insights to the campaigns that actually carry your business — the money keywords — and read those on their own. This is the same scoping habit that makes all of your Google Ads reports more honest: decide the question first, then cut the data to match it.

What can’t the auction insights report tell you?

This section matters as much as the metrics, because the inference ceiling is where most bad decisions get made. Auction insights cannot show you:

• Their budgets. You see presence, not spend.
• Their CPCs. You have no idea what they pay per click.
• Their conversion rates. Visibility says nothing about what happens after the click.
• Their profitability. This is the big one. A competitor with towering impression share might be brilliantly efficient — or burning investor money on a strategy that will quietly collapse. From this report, those two look identical.

So treat every conclusion you draw about a competitor’s strategy as an estimate, and label it that way in your own notes. “Competitor B appears to be pushing hard on mobile (estimate, based on device-segmented overlap)” is an honest sentence. “Competitor B is outspending us” is fan fiction. The report shows outcomes in shared auctions; everything about why is inference, and inference deserves a label.

What should you actually do with what you read?

A weather report earns its keep when it changes what you wear, not when it makes you yell at clouds. Here are the responses that make sense — and the ones that don’t.

Responses that make sense

• Losing to budget? Have the budget-and-focus conversation. Either fund the campaign to match its opportunity, or deliberately narrow it so your money concentrates on the auctions that convert best. Both are legitimate; drifting between them isn’t.
• Losing to rank? Work the relevance levers: tighter ad-to-keyword alignment, better landing pages, stronger expected click-through. Rank bought with relevance is cheaper and more durable than rank bought with bids alone.
• A new aggressive entrant? Watch their series before reacting. Two or three reads. New advertisers often launch loud and settle — or disappear. The no-panic rule exists because the most expensive mistake in this report is an instant bidding war against someone who was going to leave anyway.

Responses that don’t

• Outbidding wars as ego. Raising bids to sit above a specific name, on principle, transfers your margin to Google and proves nothing.
• Copying competitor positioning blind. You can’t see their results, so you might be cloning a failure. Study their public ads and pages ethically — what they say, who they target — and then decide what your customers need. Imitation without evidence is just expensive flattery.

The no-panic response menu

Tape this next to your monitor for the next time a competitor’s row makes your stomach drop:

• See something alarming → check the date range. Is this one period or a trend across several?
• Trend confirmed → check the split. Budget problem or rank problem? Different levers.
• Rank problem → improve relevance first, bids second.
• New entrant → log them, watch two more reads, then decide.
• Still worried → go look at their actual public ads and landing pages before changing anything in your account.

What mistakes should you avoid when reading auction insights?

Impression-share worship. Chasing 100% impression share usually means overpaying — buying every marginal auction, including the ones that were never going to convert. Total coverage is a vanity milestone, not a business goal.

Treating auction-mates as a market map. The names in this report are who shared your auctions, filtered by thresholds and shaped by your own keyword choices. Your real competitive landscape is bigger, and some of your truest rivals may not appear here at all.

Reacting to one period’s noise. Weather, not climate. A weird week is just a weird week until the next read says otherwise.

Assuming the report shows everything. It’s a sample of your auctions, trimmed for privacy. Absence of a competitor is not evidence of their absence.

Your quarterly auction insights checklist

• Pull auction insights for your core non-brand campaigns (90-day range).
• Pull it separately for brand campaigns — who’s on your name?
• Log each meaningful competitor’s impression share, overlap, and outranking share in your tracking sheet.
• Compare to last quarter: who’s new, who left, who’s climbing?
• Segment by device and by time for your top campaign — any hidden specialists?
• Check lost-to-budget vs. lost-to-rank on campaigns where your own share dipped.
• Write one sentence of “so what” per finding — labeled estimate or confirmed fact.
• Decide on at most one or two responses. Calm beats busy.

That’s the whole practice. Thirty minutes a quarter, a humble spreadsheet, and the discipline to read weather as weather. Do that for a year and you’ll know your competitive climate better than advertisers who check this report daily and panic weekly.

And one last honest note: while you’re watching the paid auction skies, remember that your competitors can’t bid against you in your own channels. The search auction is rented ground — your social presence, your content, your community are owned ground, and they compound while CPCs only rise.

FAQ: how to read the auction insights report

What does the auction insights report show in Google Ads?

It shows which other advertisers appeared in the same ad auctions as you — scoped to a campaign, ad group, or keyword — and compares you on visibility metrics like impression share, overlap rate, outranking share, position above rate, and top-of-page rates. It’s based on a sample of your auctions and only includes advertisers above activity thresholds, so it’s a partial picture, not a census.

Is a low impression share always bad?

No. Low impression share with a healthy cost per acquisition can mean untapped headroom or deliberate focus. The useful diagnostic is the split between impression share lost to budget and lost to rank — budget losses point at funding or narrowing decisions, while rank losses point at relevance and ad quality work.

Does a high overlap rate mean someone is my direct competitor?

Not necessarily. Overlap only means you shared auctions, and broad match targeting pulls in advertisers whose businesses barely resemble yours. Confirm the business overlap by looking at their actual public ads and landing pages before treating anyone in the report as a true rival.

Can auction insights tell me what competitors are spending or earning?

No. It reveals nothing about their budgets, CPCs, conversion rates, or profitability. A competitor with high impression share might be brilliantly efficient or losing money at scale — the report cannot distinguish the two, so label every strategic conclusion you draw from it as an estimate.

How often should I check the auction insights report?

A deliberate quarterly read works well for most accounts, with an extra look if performance shifts suddenly. One report is weather and the series is climate: track who’s new, who left, and who’s climbing across quarters in a simple spreadsheet, and avoid reacting to any single period’s noise.