Auction Insights: Who Is Beating You on Google Ads?

You open Google Ads, see impression share: 31%, and your gut says the same thing every owner's gut says: I need a bigger budget. Two-thirds of the auctions are going to someone else, so buy your way back in.

That reflex is a coin flip. Google reports your missing impression share in two separate columns, and they prescribe opposite fixes. One says add money. The other says your money is fine and your ad is losing the auction — in which case more budget just buys you more losing auctions at a higher cost per click.

Here is how to read the report properly, and how to use overlap rate and outranking share to answer the question owners actually care about: did somebody new just move into my market?

What does the auction insights report actually show you?

For Search campaigns, six numbers: impression share, overlap rate, outranking share, position above rate, top of page rate, and absolute top of the page rate. That is Google's own list. Shopping campaigns get three of them. Performance Max gets a report split by Search and Shopping.

You can run it for a single keyword, an ad group, or a whole campaign, and segment by time and device. Two limits matter before you read a single row.

• Google will not show the report when your impression share is under 10%. If auction insights is empty, that is not a bug — it is your first finding. You are barely in the auction.
• It never shows competitor budgets, bids, CPCs, or conversions. Anyone selling you 'competitor spend data' from this report is selling you an inference, not a number.

What each column is actually for:

Is your impression share lost to budget or lost to rank?

Two columns settle it, and Google defines them plainly. Search lost IS (budget) is the percentage of time your ads were not shown on the Search Network due to insufficient budget. Search lost IS (rank) is the percentage of time they were not shown due to poor Ad Rank in the auction. Both definitions are Google's, verbatim.

Add those two columns to your campaign view and read all three together. Impression share tells you what you got. The other two tell you why you did not get the rest — and which lever moves it.

• Lost IS (budget) is the big number → your ads stopped showing because the daily budget ran dry. More money genuinely buys more impressions.
• Lost IS (rank) is the big number → you had money left and still lost. Ad Rank, not spend, is the constraint. More money changes nothing about who wins the auction.
• Both are meaningful → fix rank first. Raising budget while rank is broken scales the loss.

Now the trap that makes owners misdiagnose this, straight from Google's documentation: lost IS (rank) will not be shown on your Ad groups tab if you ran out of budget at any point during the date range you are looking at. So the budget-capped account — the one that most needs to know whether its rank is also bad — is exactly the account that sees a blank where the rank number should be.

The fix is boring. Look at a date range where the campaign was not budget-capped, or read it at campaign level, before you conclude 'it's all budget.' Also note that impression share metrics update within one to two days and are reported per campaign type — they do not aggregate across the account.

When does raising your budget make things worse?

When lost IS (rank) is the larger number. Budget controls how often your ad can show; Ad Rank decides whether it shows and where. Pour money into an account losing on rank and you enter more auctions you are already losing, pay more per click for the ones you scrape into, and pull in the lower-intent tail of your keyword list.

There is a second reason a scary impression-share number is not always a spending problem: the denominator moves. Google's own guidance on improving impression share says that if you increase your regional targets, you may see a decrease in impression share, because there are now more potential impressions available to you.

Translated for a roofer: you added four suburbs to your radius last month, your impression share fell from 62% to 31%, and nothing about your account got worse. You just made the pie bigger. Before you raise spend, check whether someone widened the geography, added match types, or launched a second ad group. Our budget starting-point guide covers what a real service-business floor looks like.

One more thing worth knowing about the denominator. Google defines search impression share as the impressions you received divided by the estimated number of impressions you were eligible to receive. That denominator is a model, not a headcount of every search that happened — which is why a moving impression share can reflect a changed estimate rather than changed performance.

How do you tell if a new competitor entered your market?

Segment the auction insights report by time and compare two periods — Google supports segmenting by time and device. A genuinely new entrant shows up as a domain row that did not exist last month, with an overlap rate climbing while your impression share and outranking share slide.

The pattern reads differently depending on what actually happened:

Two caveats. The report shows domains, not spend — you can see that someone is there, never how much they are paying. And Google notes you may find google.com in your auction insights if you or a competitor use Google-hosted domains, a Google Business Profile, or a Google Sites landing page. That is not Google bidding against you.

For law firms and med spas where a single case or package pays for a month of ads, we'd check this monthly. For emergency trades, quarterly is enough unless cost per lead moves.

What do overlap rate and outranking share really mean?

Overlap rate is how often another advertiser's ad received an impression when your ad also received one — Google's example: a 60% overlap rate means that in 6 out of every 10 times your ad showed, theirs showed too. Outranking share is how often your ad ranked higher than theirs, or showed when theirs did not. A 20% outranking share means you won 2 out of every 10 of the auctions you entered together.

That bolded clause is the part everyone misreads. Outranking share is not a clean head-to-head win rate — it also counts auctions where the rival simply was not there. A rising outranking share can mean you got stronger, or it can mean they got budget-capped and disappeared. Read it next to overlap rate or you will congratulate yourself for someone else's outage.

Position above rate is the cleaner head-to-head number: how often their ad showed in a higher position than yours when both of you showed at the same time. If that is climbing, you are being outranked, full stop.

And do not treat a competitor's impression share in your report as their real impression share. Google spells this out: the figure you see for a competitor can differ from what they see in their own account, because their eligible auctions only partly overlap with yours. Google's own example — you can show 100% in the impression share column of your report while that same advertiser has 50% impression share across their eligible auctions.

What impression share should a local service business aim for?

Google publishes no target, and any agency quoting you a universal number is guessing. Our position: chase high impression share on the handful of keywords that produce booked jobs, and accept low impression share everywhere else. The last increments of impression share are the most expensive ones you will ever buy.

Impression share is a diagnostic, not a goal. Google says exactly this about its sibling metric, Quality Score: it is not a key performance indicator, should not be optimized or aggregated with the rest of your data, and is not an input in the ad auction. Treat impression share the same way. The KPI is cost per booked job.

• Emergency and now-intent terms (burst pipe, roof leak repair, DUI attorney) — push impression share hard, and watch absolute top of page rate, because position one takes the call.
• Considered, research-y terms (cost of a new roof, best med spa treatments) — low impression share is fine. Winning 90% of a browsing auction is a way to spend $2,500 on nothing.
• If auction insights shows you nothing at all — your impression share is under 10%. Stop optimizing and fix the campaign structure.

This is why we run a 90-day kill switch on every channel. If ninety days of paid search returns impression share and clicks but no qualified calls, the metric was never the problem — the offer, the landing page, or the keyword list was. A leaderboard you win with no revenue attached is a vanity metric with a billing cycle.

What do you change when you're losing on rank, not money?

Four levers, cheapest first — and raising your bid is the last one, not the first. Google defines Ad Rank as a set of values calculated from your bid amount, the quality of your ads and landing page, the Ad Rank thresholds, the competitiveness of the auction, the context of the person's search, and the expected impact of assets and ad formats. Bid is one input among six.

The line worth taping to your monitor is Google's own: even if your competition has higher bids than yours, you can still win a higher position at a lower price by using highly relevant keywords and ads.

• Tighten what you match. Broad match dragging in unrelated searches drives down expected clickthrough rate. Start with a negative keyword sweep.
• Make the ad match the search. Ad relevance is one of the three components of Quality Score, alongside expected CTR and landing page experience. One ad group per real service, not one per city.
• Fix the landing page. Sending 'emergency AC repair' traffic to your homepage is a landing page experience problem, and it is priced into the auction. Here is what a paid-ads landing page should do.
• Then, and only then, bid. Once relevance is real, a bid increase buys position instead of renting it. Check your bid strategy before you move a number by hand.

Work them in that order and the same $2,500 buys more auctions than it did last month. Work them backwards and you will pay a premium for the privilege of losing.

If lost IS (rank) is the number staring back at you and you are not sure which lever is broken, that is precisely what a paid ads teardown is for. We'll read your auction insights, split budget from rank, and tell you which one is actually costing you jobs — no lock-in, no retainer required to find out. Get my free audit.