What auction insights shows, and what it cannot
Google's description of how the auction works explains why competitors matter at all. Each auction orders the eligible ads by Ad RankThe value Google Ads calculates at each auction to decide if and where an ad shows., calculated from the bid, the auction-time quality of the ad and its landing page, the Ad Rank thresholds an ad must clear to show, the competitiveness of the auction, the context of the search and the expected impact of assets and formats. The price of a click is what is needed to beat the advertiser ranked below you, which is often less than your maximum bid; the Ad Rank entry sets out the inputs in more detail. When a new advertiser enters your auctions, or an existing one bids more or improves its ads, the Ad Rank needed to hold your place rises, and so does the price. The rising CPC diagnosis starts from that mechanism; auction insights is where you see who is pushing.
The report gives, for each domain that appeared in the same auctions as your ads, a small set of measures. Their exact definitions sit in the column descriptions in the account, and it is worth reading them there once, because the counting rules matter when two figures disagree.
| Measure | What it tells you | What it does not tell you |
|---|---|---|
| Impression share (theirs and yours) | How often each advertiser showed in the auctions your ads took part in | How often they showed in auctions you never entered |
| Overlap rate | How often a competitor's ad appeared in the same auction as yours | Whether they were above or below you |
| Position-above rate | When you both showed, how often their ad ranked above yours | Why: bid, quality or context |
| Outranking share | Your side of the comparison: how often your ad finished ahead of theirs | Their results or their costs |
| Top-of-page rate | How often each advertiser's ad showed at the top of the page | The value of being there for your terms |
Three limits shape every reading. The report covers only the auctions you took part in, so a competitor dominating searches your keywords never matched is invisible here. It contains rates, never bids, budgets, keywords, conversions or spend. And competitors appear as domains, which identify advertisers but not the campaigns behind them. Anything you conclude about a rival's strategy is inference; anything you conclude about your own position is evidence.
Auction insights also differs from share of voice. Your impression share is a coverage gauge for bidding and budgets inside one platform, while Share of VoiceYour share of market visibility. measures a brand's weight in a whole category; the impression share versus share of voice comparison explains why a high figure on one says little about the other, and the competitive intelligence guide shows where paid search visibility fits among the other signals a market watch collects. A local business whose paid rivals are also the firms beside it in the map results can set this report next to their profiles and reviews, as the tutorial on benchmarking nearby competitors does.
Before you start
- Standard or read-only access to the account, with the campaigns that carry most of the spend identified. Reading the report needs no editing rights.
- Your own impression share figures for the same campaigns, including the share lost to budget and the share lost to rank. The impression share entry explains how the three add up to the whole of the demand the campaign was eligible for.
- Cost, cost per click and conversions for the same campaigns and periods, so that movements in the report can be judged against what they cost.
- A note of your own recent changes: budgets, bid strategies and targets, keywords, negatives and ads. The account's change history holds them, as the account management hub describes.
- Last month's reading, if there is one. The value of the report grows with each month on record.
The steps
Choose the campaigns or keywords to read
Pick what the reading is for. The search campaigns hub sets out how keywords, ads and bids fit together; here the campaigns that carry most of the spend come first, then any campaign whose cost per click or volume has moved. Read brand and non-brand terms separately, because the competitors, the stakes and the right response differ between them. Where one campaign mixes several jobs, narrow the selection to the keywords that do one job, such as repair terms or replacement terms, so that a competitor active in only one of them is not averaged away. An affiliate partner's domain can also appear in the brand auctions, and the guide to setting brand bidding rules for affiliates uses this reading as monitoring evidence.
Choose comparison periods
Compare like with like: this month against last month, and against the same month a year earlier if the business is seasonal. Avoid periods that contain your own large changes, such as a new bid strategy or a restructure, unless explaining that change is the point of the reading. Keep the period length the same each month, otherwise movements in the rates partly reflect the window rather than the market.
Record your own impression share and lost share
Before looking at anyone else, write down your impression share and the shares lost to budget and to rank for the selection and both periods. If the share lost to budget dominates, the campaign is capped by its budget and competitors are a secondary question; the budgets and pacing guide owns that decision. If the share lost to rank has grown, someone is winning auctions you used to win, and the rest of the reading is about who and why.
List competitors by impression share
Sort the competitors by their impression share in the current period and set last period's figure beside each. Mark three kinds of change: a name that was not there before, a familiar name whose share rose sharply, and a name that disappeared. The competitive intelligence guide treats a new or fast-rising name as an early warning of increased bidding on your terms; this is where you see it first.
Read overlap
For each competitor that matters, read the overlap rate: how often their ad appeared in the same auctions as yours. A high overlap means you are competing for the same searches and every change in their behaviour reaches your prices. A low overlap with a high impression share means they are strong in auctions you rarely share, often a sign that your keywords and theirs cover the market differently, which is a coverage question rather than a price one.
Read position above and outranking
Next read the position-above rate, how often the competitor ranked above you when both appeared, and your outranking share against them. Rising position-above rates on terms where your bids have not changed mean their Ad Rank is rising relative to yours, through bids, quality or both. Because Ad Rank includes the auction-time quality of the ad and landing page, a rival can overtake you without spending more, which is why the next steps look at quality before money.
Read top-of-page rates
Compare your top-of-page rate with theirs and with last period. Losing the top of the page while keeping your overall impression share means you still enter the auctions but finish lower, which usually costs clicks before it costs impressions. The impression share entry lists a top-of-page share far below total share as the pattern of winning auctions in lower positions; here you can see whose ads took the places.
Set the movement beside your costs and conversions
Now bring in the money. For each selection, put the competitor movements next to your cost per click, conversion rate and cost per conversion. Competitive pressure that raises cost per click on terms that still convert within target is a cost of doing business; the same pressure on terms already near the limit is a decision. Check the change history for your own edits in the same period, so a competitor is not blamed for a target you tightened. The impressions dropped diagnosis sets out that order of checks when volume, rather than price, is what fell.
Decide: defend, concede or improve quality
Rank the affected terms by the value of a conversion and choose one of three responses for each. Defend the terms that pay at the higher price. Concede position on terms that do not, and let them settle lower where they cost less. And on the terms you defend, work on quality before bids: Google documents that Quality ScoreGoogle's 1–10 keyword diagnostic. is a diagnostic rather than an auction input, but its components (expected click-through rate, ad relevance and landing page experience) summarise the same quality the auction assesses, so the Quality Score guide shows where a better ad or page would lower the price of the position. The CPC calculator shows what a given rise in price does to the clicks a budget buys.
Record the reading and repeat monthly
Write one line per selection: the period, your share and lost share, the competitors that moved, the decision and the date to look again. Next month, start from that line. A record turns a report that shows one month into a history that shows when a competitor arrived, how hard they pushed and whether your response held. For brand terms during a brand search holdout test, the same record is the guardrail that shows whether a rival takes the position you vacated.
A simple way to keep the reading honest is to separate what the report shows from what you conclude and what you will do, which is the order of the what, so what, now what framework. The table below is the decision part in compact form.
| What the reading shows | Likely meaning | Response |
|---|---|---|
| New name, high overlap, rising position-above rate, your cost per click up | A new or stronger rival in your core auctions | Defend the terms that pay; work on ad and page quality there |
| Your share lost to rank up, no competitor movement | Your own quality or bids slipped, or thresholds rose | Read Quality Score components and your recent changes |
| Competitor share up, low overlap with you | They are strong in auctions you rarely share | A coverage question for keyword research, not a bidding one |
| Your top-of-page rate down, impression share steady | You still compete but finish lower | Decide whether the top position is worth its price on these terms |
| Share lost to budget dominant | Your budget caps you before competitors matter | Budget or target decision first |
A new competitor, read month by month (illustrative)
A roofing contractor meets a new rival on repair searches
Pantilefox Roofing, an invented roofing contractor, runs two search campaigns: urgent repairs and full roof replacement. Over two months the average cost per click on repair terms rises from £3.10 to £4.05 and the cost per repair lead from £48 to £66, against a maximum of £60 the owners set from the margin on a repair job.
| Repair terms | Month one | Month three |
|---|---|---|
| Pantilefox impression share | 62% | 49% |
| Share lost to budget | 17% | 15% |
| Share lost to rank | 21% | 36% |
| Competitor A impression share | not present | 58% |
| Competitor A overlap rate | not present | 61% |
| Competitor A position-above rate | not present | 64% |
| Pantilefox top-of-page rate | 71% | 58% |
On replacement terms Competitor A's overlap is 9% and Pantilefox's share holds at 54%. The reading is clear: a new rival is concentrating on repairs. Change history shows no edits to the repair campaign in the period, so the rise is not self-inflicted. The owners decide not to chase the top position on every repair term. They defend the emergency terms that bring the most valuable jobs, rewrite the repair ads to state the call-out time and send them to a repair-specific page instead of the homepage, where landing page experience had been rated below average, and let the general repair terms settle lower. A month later, share lost to rank on repair terms is 30%, cost per click is £3.80 and cost per repair lead is £57, back under the limit, while the replacement campaign is unchanged.
Common mistakes
- Reading competitors before your own lost share. If the budget caps the campaign, the competitor list explains little; the share lost to budget comes first.
- Treating the report as a view of the whole market. It covers only the auctions your ads entered.
- Inferring bids or budgets as facts. The report holds rates; a rival's strategy is a hypothesis.
- Averaging brand and non-brand terms. Different competitors, different stakes, different responses.
- Answering every rise with higher bids. Ad Rank includes ad and landing page quality, and a better ad or page lowers the price of a position for good; bid only where the higher cost per click still pays.
- Blaming competitors for your own edits. Check change history for target, budget and keyword changes in the same period before drawing conclusions.
- Reading it once. A single month shows positions; the monthly record shows arrivals, pressure and whether the response worked.
