Google Ads Auction Insights: What the Data Really Shows

A competitor moved above you. CPC rose. Impression share fell. The obvious response is to open Google Ads Auction Insights and look for the culprit.

That report can show a meaningful visibility change, but it cannot tell you that a competitor raised its bid, increased its budget, or is winning profitably. Our rule is simple: treat Auction Insights as a hypothesis generator, not a command to spend more.

This guide explains what every percentage actually measures, why the rows don't add to 100%, how Search, Shopping, and Performance Max differ, and what to check before changing bids or budget. It also gives you a decision process where doing nothing is a legitimate outcome.

Quick takeaways

• Auction Insights shows relative ad visibility, not competitor profitability.
• Search Partners are excluded, and your own impression share generally needs to reach 10% before the report appears.
• Impression share, overlap, position above, outranking, top rate, and absolute-top rate do not share one denominator.
• Performance Max Auction Insights covers its Search and Shopping participation on the Search Network, not its complete cross-channel delivery.
• A changed competitor percentage matters only when the affected traffic and business results matter.

What Google Ads Auction Insights Can and Cannot Tell You

Google's Auction Insights documentation describes a report that compares your performance with advertisers participating in the same auctions. Depending on campaign type and scope, it shows how frequently advertisers appeared, how often they appeared together, and how prominent they were when shown.

It does not reveal a competitor's:

• Bid or bidding strategy settings
• Daily or monthly budget
• Spend, clicks, conversions, CPA, ROAS, margin, or profit
• Keyword list, match types, search terms, audiences, or targeting
• Quality Score, auction-time quality, or reason for entering an auction

That distinction matters because readers usually open the report after something uncomfortable happened. A practitioner may see CPC rise with no visible change in overlap or impression share. Another may have already checked ads, locations, negatives, bidding, and budget while impressions continue to fall. Auction Insights adds context, but it doesn't decompose CPC or prove causation.

The report also has availability limits. Search Partners are excluded. Google says your own impression share must be at least 10%, and the selected entity must clear an unpublished minimum activity threshold. A missing report can mean low share, insufficient activity, a narrow segment, incomplete data, or unsupported inventory. It does not prove there was no competition.

And don't diagnose yesterday. Google's data freshness guidance says Auction Insights and impression-share metrics are generally calculated daily, but recent Search data can take several days to settle and Shopping data can take longer. Closed, comparable periods beat anxious day-over-day reactions.

The report looks like a scoreboard. The denominators tell a different story.

Choose the Right Scope Before Comparing Percentages

Two percentages can both be correct and still be unsafe to compare. A campaign-level figure, a Shopping report, and a PMax Search view may represent different eligible auctions. That is why one advertiser reported roughly 10% campaign impression share alongside 75% Shopping Auction Insights. The first question isn't “which number is broken?” It is “what exact inventory and entity does each number describe?”

Search supports the deepest report. Shopping does not include position-above, top-of-page, or absolute-top rates. PMax does not turn Auction Insights into a complete view of every surface where the campaign serves. Google's PMax channel reporting guidance can add channel context, but Google also warns against judging a campaign's marginal value from isolated channel averages.

There is another Shopping trap. Product-group tables can show impression-share metrics, but Google does not document product group as an Auction Insights competitor-report level. Shopping Auction Insights can also filter very-low-volume auctions to reduce latency, so small differences versus other impression-share surfaces are possible.

How to Find Auction Insights in Google Ads

The current consolidated route is:

Campaigns → Insights and reports → Auction insights

Google's reporting navigation documentation describes consolidated account, campaign, and ad-group views. The direct route also remains useful: open Campaigns, Ad groups, or Search keywords, select the relevant entities, then choose Auction insights. Google's current navigation map is worth checking when an account is in a different interface test, and the Report Editor glossary lists the available report levels by campaign type.

Before comparing periods, hold these constant where possible:

• Brand versus non-brand
• Search versus Shopping
• Campaign, ad group, or keyword scope
• Device and geography
• Match type or query category
• Date length, weekdays, promotion status, and conversion lag

An account-wide view can orient you. It is usually too blended to diagnose you.

How the Six Auction Insights Metrics Are Calculated

The cleanest way to read Auction Insights is to ask one plain-language question per column. Each answer applies only to the selected scope and reporting period.

Impression Share: The Portion of Eligible Auctions Captured

For your row:

Impression share = impressions received ÷ estimated eligible impressions

Eligibility is estimated from factors such as targeting, approval, bid, and quality. Google's impression-share explanation also makes clear that this is an estimate, not a fixed market total.

Impression share is not click share, spend share, conversion share, revenue share, customer share, or category market share. Its denominator can change when targeting, bids, query matching, quality, competitors, or Google's systems change. Small movements don't automatically deserve action.

Competitor impression share deserves extra caution. Google's public Help wording and the non-public schema language do not describe its denominator identically. The safe interpretation is: a competitor's impression share is a contextual estimate inside the eligible auction universe represented by your report. It is not the competitor's account-wide share. Google even notes that an advertiser showing 100% in your report might see 50% in its own because the two eligible auction sets differ.

Overlap Rate: How Often You and a Competitor Both Appeared

Overlap rate = joint appearances ÷ your appearances

If another domain has 60% overlap, it appeared in six of every ten instances in which you received an impression. That doesn't prove you use the same keywords, match types, bids, or targeting. Broad match, close variants, Dynamic Search Ads, AI-led expansion, and PMax can create overlap without identical keyword lists.

There is no official universal threshold for a “high” overlap rate. Use each advertiser's historical baseline in a clean segment. A paid-search operator asking whether the report is simply good or bad is asking for a decision. A generic 60% benchmark cannot provide one.

Position-Above Rate: How Often a Competitor Ranked Higher

Position-above rate = joint appearances with competitor above you ÷ all joint appearances

A 68% rate means the competitor appeared higher in 68% of the cases where both advertisers received an impression. It does not reveal why.

The difference could reflect bid, auction-time quality, thresholds, expected asset impact, device, location, time, query context, landing-page relevance, or different eligibility. Google's actual CPC definition and manual CPC explanation are useful reminders that visible position does not expose a rival's maximum bid or exact auction-clearing price.

Outranking Share: When You Ranked Higher or Showed Alone

Outranking share combines two situations:

1. Your ad appeared above the other advertiser.
2. Your ad appeared and the other advertiser did not.

That makes it a composite visibility metric. It is not the opposite of position-above rate. Your outranking share can rise simply because the competitor appeared less often, even if your relative order during joint appearances did not improve.

Google's prose definition, simplified example, and non-public schema do not align cleanly enough to justify reconstructing raw pairwise auction counts. Use the displayed percentage as directional context. Don't manufacture auction counts that the interface doesn't provide.

Top-of-Page Rate: How Often an Ad Appeared Above Organic Results

Top-of-page rate = top impressions ÷ impressions received

This is a prominence rate. If an advertiser received 100 impressions and 80 were among top ads, its top rate is 80%.

Google's current dedicated page appears to contain a formula-label inconsistency, but its definition and worked examples support the received-impressions denominator. The top and absolute-top metrics documentation is still the canonical place to understand these columns.

One counterintuitive point is especially useful: increasing bids can sometimes lower top rate. A higher bid may make the ad eligible for more competitive auctions where it appears in a lower location. That is why top rate is not a direct instruction to bid more.

Absolute-Top Rate: How Often an Ad Was the First Top Result

Absolute-top rate = first-position top impressions ÷ impressions received

An advertiser can have a 90% absolute-top rate and very low impression share. That may reflect narrow participation, selective queries, a brand-heavy mix, strong auction-time quality, high bids, or several of those factors. The report cannot tell you which explanation is correct.

Now the percentages that looked impossible start to make sense. They belong to different universes.

Why Rate and Share Percentages Do Not Add Up

Consider one advertiser that was eligible for 1,000 impressions:

• It received 300 impressions, so its impression share was 30%.
• It appeared among top ads 180 times, so its top-of-page rate was 60%.
• It was the first top ad 60 times, so its absolute-top rate was 20%.

Now imagine a competitor that received only 50 impressions, with 40 among top ads. Its top rate is 80%, even though its total visibility may be much smaller.

Advertiser X is not automatically more dominant. It is more prominent conditional on appearing. Advertiser Y captured much more eligible visibility.

This also explains why several advertisers can have high top or absolute-top rates at once. Each rate is divided by that advertiser's own received impressions. A practitioner put the confusion plainly: “We can't all be on the Absolute Top Of Page a combined 300% of the time”. The percentages are not slices of one pie, so they are not supposed to sum to 100%.

The denominator model resolves the math. It still doesn't tell you what to do.

How to Interpret Auction Insights Metric Combinations

Single columns invite confident stories. Combinations force better questions.

The high-share/low-top pattern often unsettles advertisers. One practitioner described being the top advertiser by impression share while ranking last on top-of-page rate. That can happen when you participate broadly but often lower on the page, while another advertiser enters fewer auctions more selectively.

PMax creates another version of the same anxiety. A reader asking whether low share means it is “just budgets increase game” or assuming 15% share proves profitable headroom is skipping the most important question: are the missed auctions worth buying?

Only then does an action become defensible.

How to Decide Whether an Auction Insights Change Matters

The report becomes useful when it sits inside a disciplined sequence.

1. Check Whether the Business Outcome Changed

Start with qualified leads, revenue, conversion value, CPA, ROAS, conversion volume, and margin where available. If the outcome did not worsen, a visibility change may deserve monitoring rather than intervention.

This is the point many guides miss. A profitable 20% impression share can be better than an unprofitable 80%.

2. Compare the Same Campaign Scope

Separate brand from non-brand. Keep campaign type, device, geography, date length, and day-of-week composition comparable. Account-level orientation is fine; diagnosis should use the cleanest meaningful segment.

3. Wait for Complete Reporting Periods

Compare closed 28-day windows, matched weekdays, or year-over-year periods for strongly seasonal accounts. Avoid making a budget decision from incomplete recent data.

4. Export a Baseline Before Making Changes

Save the date range, level, campaign type, Search or Shopping view, segment, advertiser domains, all Auction Insights columns, Lost IS, and business metrics. This creates a baseline that survives interface and reporting changes.

5. Describe the Exact Visibility Change

Don't write “competition increased.” Write: “Competitor X's overlap rose from 24% to 41% on mobile non-brand traffic while our CPC increased 8%.” Precision prevents the story from outrunning the evidence.

6. Compare Lost Impression Share With Economics

Google's Lost Impression Share guidance separates missed opportunity caused by budget from missed opportunity caused by rank.

• Lost IS (budget): eligible impressions missed because the campaign lacked budget.
• Lost IS (rank): eligible impressions missed because Ad Rank was insufficient.

Neither metric tells you whether recovering those impressions would be profitable. Pair them with CPC, CTR, conversion rate, CPA, ROAS, conversion value, and margin.

7. Write a Hypothesis You Can Test

Weak: “Competitor A is bidding aggressively.”

Better: “Competitor A's overlap and position-above rates increased on mobile non-brand traffic while our Lost IS rank rose, which is consistent with greater relative rank pressure in that segment.”

The second statement can be checked. The first pretends to know a bid.

8. Verify the Pattern With Other Google Ads Reports

• Use the Search Terms report to inspect the queries your ads actually matched.
• Use the Ads Transparency Center to inspect public ads, advertiser identity, creative, and landing-page patterns.
• Use Ad Preview and Diagnosis instead of repeatedly searching for your own ad and creating impressions.
• Use Change History, campaign diagnostics, and policy status to rule out your own changes or eligibility problems.

Ad Preview is simulated and will not reproduce every live auction condition. It corroborates. It does not turn one query check into ground truth.

9. Make One Change and Set a Stop Rule

Possible responses include a budget reallocation in a profitable constrained segment, a new offer or message, better query-to-ad-to-page alignment, tighter negatives, a device adjustment, a feed improvement, or no immediate change.

Define the success metric and stop condition before launch. Judge incremental conversions, CPA, ROAS, value, or profit. Don't judge the test by whether a competitor's percentage moved in your preferred direction.

That final discipline is what keeps a report from becoming a bidding war.

Auction Insights Example: Frequent vs. Selective Competitors

The figures below are illustrative, not benchmarks.

Your Lost IS is 8% from budget and 40% from rank.

Competitor A is the frequent rival. It appears alongside you often and is commonly higher when both show. Rank is also the larger source of your missed visibility. That supports checking query clusters, device, geography, ad and landing-page alignment, and whether the affected traffic can profitably support a bid or target change. It does not prove A's bid is 68% higher.

Competitor B is the selective rival. It rarely appears with you but is prominent when present. An account-wide bid increase would be badly targeted. Find the narrow segment where B appears and decide whether that segment matters.

Your account has a rank constraint, not automatically a business problem. Lost IS rank is much larger than Lost IS budget, but that only identifies the visibility constraint. If conversions and profitability are healthy, the correct response may still be selective testing or no change.

The report does not provide raw pairwise auction counts, so you cannot calculate a conventional confidence interval from these percentages. The 10% visibility gate is not a statistical-significance guarantee. Low-volume segments need longer periods and more restraint.

The model is stable. The reporting environment around it has changed.

Google Ads Auction Insights Changes in 2025 and 2026

One Advertiser Can Now Appear at the Top and Bottom

On April 21, 2025, Google documented separate top and bottom ad auctions that can allow the same advertiser to appear in both locations on one results page. Google's multi-location ad guidance says advertisers do not bid against themselves and recommends the Top vs. other segment for performance differences.

Google's test reported approximately 10% more highly relevant bottom ads and 14% more bottom-ad conversions. Those figures describe the serving test, not a guaranteed advertiser outcome. Don't claim Auction Insights denominators doubled. The safe conclusion is that multi-location serving changed impression-counting context and complicates comparisons across the April 2025 boundary.

AI Overview Ads Count as Top Ads but Cannot Be Segmented

Google's AI Overview ad documentation says eligible Search, Shopping, and PMax ads can appear around or inside AI Overviews in supported markets. Ads inside an AI Overview are reported as Top Ads, but there is no reporting segment that isolates their contribution.

A shift in top rate may include AI Overview inventory. Auction Insights cannot tell you how much.

Why Auction Insights Still Requires Manual Exports

At the July 24, 2026 research cutoff, Google Ads API release notes listed v25 as the latest major version. The v25 Metrics schema contains Auction Insights fields marked not publicly available. Google's Insights API overview does not change that practical limitation.

An older API schema diff also reflects restricted field availability, and an API forum discussion shows why scripts are not a dependable ordinary-user workaround. Build your process around UI exports unless Google publicly changes access.

Why Looker Studio No Longer Shows Auction Insights Fields

New Looker Studio sources lost the connector fields on August 24, 2024, and existing sources lost them on September 23, 2024. Industry reporting on the removal and the Google developer forum confirmation support the manual export-and-append workflow.

Practitioners felt the loss. One advertiser wrote that they loved having the data in Looker Studio, while another reported that even a very large company could not obtain restricted access.

How Google Ads Bidding Labels Changed in 2026

Starting in June 2026, Google renamed “Maximize conversions with a target CPA” to Target CPA and “Maximize conversion value with a target ROAS” to Target ROAS. Google's bidding organization notice says the underlying behavior did not change.

Google has also announced a target-strategy behavior change for after August 17, 2026. As of July 27, 2026, it is not live. The target-strategy change announcement should be treated as a future watch item, not an explanation for current results.

These changes make historical context more valuable, not less.

How to Build an Auction Insights Tracking Process

A practical tracker appends each export instead of overwriting it.

Weekly review can make sense for high-volume accounts, launches, promotions, or brand pressure. Monthly review is often enough for steadier accounts. The right reporting cadence is the shortest interval that still provides complete data and enough conversion volume to judge the business effect.

For stakeholder communication, use five lines:

That format is more useful than a “competitor threat score,” which would combine incompatible denominators and ignore profitability.

Official Google Ads Sources to Keep With Your Review

You do not need a sprawling source library, but operators should know where the definitions come from. Alongside the core Auction Insights page, keep these official references available:

• Get impression share and Lost IS data
• Top and absolute-top metrics
• Actual CPC
• Quality Score guidance
• Ad Strength guidance
• Smart Bidding organization changes
• Future target-strategy change

For a broader example of how competitive visibility should be corroborated, Google's IPRoyal international expansion case is useful. The team paired Auction Insights with Google Trends and acquisition economics. The case reports lower acquisition cost in Kenya, lower average order value, major revenue growth, and expansion into more countries, but Auction Insights alone did not cause those results. It helped form a market hypothesis that actual performance then had to validate.

Industry analysis can add perspective, too. A practitioner-focused Auction Insights article reinforces the value of combining the report with Transparency Center checks. LinkedIn discussions describe the report as a “brain-twister” when numbers are read without context and, more usefully, as imperfect data that provides clues.

The history tells you whether a signal persists. It also creates the next job: deploying a measured response without losing control of the test.

How to Turn an Auction Insights Finding Into a Controlled Test

Auction Insights does not tell you which new asset, offer, feed change, landing page, or message will fix a problem. Once the evidence supports a test, execution becomes the bottleneck.

That is where we position AdManage. Teams can use AdManage's Google workflow and structured Google Sheets launch documentation to organize approved inputs, reusable copy, media, names, tracking fields, and launch status. Consistent naming and UTMs do not improve Ad Rank. They make the test easier to deploy and interpret.

For deeper planning, our existing resources can help teams set a creative testing budget, build a consistent ad naming convention, and browse the broader AdManage blog. The public comparison page explains the operator-control philosophy, while pricing describes the fixed-fee model without tying software cost to ad spend.

AdManage does not import Auction Insights data, reveal competitor bids or economics, choose a bidding response, or guarantee higher impression share. Current public pages also differ on Google campaign-type availability, so we are deliberately not claiming support for a specific Search, PMax, or app-campaign launch here. Check the current Google workflow page before planning implementation.

For agencies and in-house teams, the clean handoff is:

1. Use Auction Insights and adjacent data to choose a justified hypothesis.
2. Define one test, success metric, and stop rule.
3. Use a structured launch process to deploy only what the team approved.

Speed is valuable after the diagnosis is sound. Before that, speed only helps you make the wrong change faster.

What to Do Before You Change Bids or Budget

Auction Insights gives you a partial, relative view of visibility. Its precision can tempt you into a story that the report cannot support: a competitor raised bids, your budget is too low, or a higher position must be more profitable.

Our take is narrower and more useful. A percentage is an observation. The decision comes from scope, Lost IS, query quality, conversion economics, and a controlled test. If the business outcome remains healthy, monitoring may be the best action. If a valuable segment is genuinely constrained, make one measured change and judge the incremental result.

Start by writing one sentence that names the exact movement, scope, and business effect. If you cannot write that sentence yet, you are not ready to change the account.

You can share or bookmark the canonical Google Ads Auction Insights guide for the next review.

Google Ads Auction Insights FAQs

Why Is Auction Insights Missing From My Account?

Your impression share may be below 10%, the selected entity may have insufficient activity, the segment may be too narrow, recent data may still be processing, or the campaign inventory may be unsupported. Broaden the scope and use a complete date range before concluding there was no competition.

Does Auction Insights Show Competitor Keywords, Bids, or Spend?

No. Auction Insights shows relative visibility inside intersecting auction contexts. A reader looking for competitor spend, CPA, and related economics is asking for data the report does not contain.

Can Other Advertisers See My Domain in Auction Insights?

Your display domain can appear in another advertiser's eligible report when your ads intersect with its auction universe. Their percentages may not mirror yours because scope, eligibility, and filters can differ.

What Counts as a Good Overlap Rate?

Google publishes no universal benchmark. Compare a competitor against its own historical baseline in a clean segment, then check whether the business outcome changed. High overlap without economic harm may need no response.

Should You Aim for 100% Impression Share?

No. Capturing marginal, low-value auctions can make CPA or ROAS worse. Increase budget only when budget is constraining a commercially valuable segment and marginal performance supports more spend.

Why Can Impression Share Be High While Top Rate Is Low?

You may be entering a broad set of eligible auctions but appearing lower on the page in many of them. Another advertiser can participate selectively and show at the top more often without matching your overall visibility.

Does PMax Auction Insights Include YouTube or Display?

No. It covers the Search and Shopping portions of PMax on the Search Network. It is not a full competitor report for YouTube, Display, Discover, Gmail, or every other PMax surface.

Is Auction Insights Available in the Google Ads API?

Not through the standard public Google Ads API at the research cutoff. The metrics exist in the schema but are marked non-public. Manual UI downloads remain the dependable ordinary-user workflow.

Can You Connect Auction Insights to Looker Studio?

Not through the former native connector fields, which were removed in 2024. Export the report, append it to a controlled Sheet or table, then visualize the historical data.

What Does the google.com Row Mean?

Google says the row may represent Google-hosted destinations, a Google Business Profile, Google Sites, or Google Play landing pages. Use Google's documented explanation rather than guessing which Google property appeared.

Does Quality Score Explain Competitor Position?

Not by itself. Google says the displayed 1 to 10 Quality Score is a historical diagnostic and is not an auction input. Its components can reveal improvement areas, but the score does not expose the competitor's auction-time quality.

Does Ad Strength Affect Ad Rank or Auction Wins?

No. Google describes Ad Strength as feedback on assets. It is not used to calculate Ad Rank, Quality Score, auction eligibility, or auction wins.

How Often Should You Check Auction Insights?

Weekly can work for high-volume accounts, launches, promotions, or active brand pressure. Monthly is often better for steadier accounts. One agency operator described the interface as unwieldy across many competitors, which is why a repeatable export and review template matters more than constant checking.

Why Can CPC Rise When Auction Insights Looks Unchanged?

Check query mix, match-type expansion, bidding behavior, demand composition, your own quality and landing-page changes, and advertisers that do not appear consistently in the report. A separate practitioner example of rising CPC shows why a flat visible competitor set does not close the diagnosis.