Google Ads diagnostic · SEM
Search Lost IS (budget) and rank describe two different failures
Read the columns as a reconstruction of delivery, not as two generic reasons to spend more.
- Budget loss is a campaign-level capacity signalThe campaign could have served in more estimated eligible opportunities, but its budget and pacing constrained delivery.
- Rank loss happens inside auction-time competitionBid matters, but so do ad and landing-page quality, thresholds, competition, query context and the expected impact of assets.
- The denominator is estimated and can moveA bid, quality or targeting change can alter the auctions Google considers you competitive enough to enter, so lost share is not fixed inventory.
- Change one constraint and require a business resultA lower loss percentage is useful only when qualified conversions, value or profit improve at an acceptable marginal cost.
Search Lost IS (budget) and Search Lost IS (rank) look like adjacent percentages in Google Ads, but they describe missed impressions at different points in delivery. The budget column is about campaign capacity. The rank column is about auction-time competitiveness. Treating both as a single instruction to “raise spend” is how an account turns a diagnostic report into an expensive guessing game.
The distinction matters because impression share is not market share and not a count of guaranteed future clicks. Google defines Search impression share as impressions received divided by the estimated impressions an ad was eligible to receive. That eligibility estimate uses targeting, approval, bids and quality, and it can change when those inputs change. The report is therefore a reconstruction of a moving auction environment, not a warehouse of impressions waiting to be purchased.
Auctions in which the ad actually produced an impression.
A changing set shaped by targeting, approval, bids, quality and auction conditions.
A percentage can improve because you captured more opportunities, because the eligible set changed, or because both happened.
Start with the denominator, not the recommendation
Google estimates eligible impressions from the auctions over the day and can include auctions where the ad did not show but was considered competitive enough to have shown at a higher bid. It may exclude auctions where the gap was extreme. A campaign’s Search IS can therefore change even when search demand is stable, because its own settings, quality signals or bids changed the set of plausible opportunities.
This is why multiplying Search lost IS by current conversions creates false precision. The missed auctions may occur at different hours, locations, devices, queries and levels of competition from the auctions already won. Recovering more impressions can change CPC, conversion rate and mix. The column identifies where delivery was constrained; it does not certify the economics of removing that constraint.
Budget loss happens before the auction can spend another impression
Search lost IS (budget) is the percentage of estimated eligible Search impressions missed because the campaign had insufficient budget. Google exposes it at campaign level, not at ad-group or keyword level. That reporting boundary is not cosmetic: a shared campaign budget and its pacing logic decide which ad groups and queries receive the available capacity.
A high value means the campaign could have participated more often under its current delivery conditions. It does not automatically mean the next dollar is profitable. Before increasing budget, separate a campaign that is capped while meeting its business target from one whose target or query mix is already weak. A campaign can be budget-constrained and still be a poor scaling candidate.
Budget question
Would another unit of campaign capacity buy qualified incremental demand at an acceptable marginal CPA or ROAS?
- Check daily pacing and high-value hours.
- Separate profitable demand from broad low-value expansion.
- Review whether a restrictive bid target is also limiting entry.
Rank question
Why did this ad fail eligibility or lose position when an estimated opportunity existed?
- Inspect bids or automated-bidding targets.
- Improve relevance, landing-page usefulness and assets.
- Account for thresholds, competition and query context.
The phrase “limited by budget” also deserves context. Google’s bidding system may reduce the frequency of appearances to avoid exhausting the daily budget too early. With automated bidding, an aggressive Target ROAS or low Target CPA can simultaneously reduce the auctions the system is willing to enter. Adding money without checking the target can leave the underlying constraint untouched or expand into a different marginal auction set.
Rank loss is not a synonym for a low bid
Search lost IS (rank) estimates how often the ad did not show because its Ad Rank was too low. Ad Rank is recalculated for every auction and Google describes it as a set of values, not one permanent score. Bid is one input. Auction-time ad quality, landing-page experience, thresholds, competition, the context of the search and the expected impact of assets also contribute.
That makes “raise bids” only one possible intervention. A higher bid may win more auctions, but it can also expose the campaign to more expensive competition. Improving the match between query, ad and landing page can increase auction strength without accepting the same cost increase. For automated bidding, the practical lever may be the target rather than a keyword bid: a lower Target ROAS or higher Target CPA gives the system permission to compete more broadly.
There is also a reporting trap. Google says Lost IS (rank) will not be shown on the ad-group tab if the campaign ran out of budget at any point in the selected period. Missing rank data at that level is therefore not proof that rank was healthy. Move to the campaign view, choose a clean date range and resolve budget interference before diagnosing ad-group rank.
Loss-column decision matrix
What to test before changing spend
| Observed pattern | Primary hypothesis | First controlled change | Business guardrail |
|---|---|---|---|
| High budget loss, acceptable efficiency | Profitable campaign capacity is exhausted | Increase budget modestly without changing targeting or target | Marginal CPA, ROAS or contribution |
| High budget loss, weak efficiency | Capacity exists, but demand quality is poor | Fix query mix, measurement or value signals first | Qualified conversion rate and value |
| High rank loss, budget available | Auction competitiveness is insufficient | Adjust one target/bid or improve one quality surface | Incremental CPC and conversion value |
| Both losses elevated | Budget masks a separate rank constraint | Create a clean budget window, then inspect rank | Pacing plus marginal economics |
Use a one-change loss test instead of a blanket optimisation
The cleanest experiment changes one delivery constraint while keeping the rest of the campaign stable for an appropriate observation window. Google notes that impression-share data can take one to two days to update, while automated bidding changes may need one or two conversion cycles before performance settles. A same-day before-and-after verdict mixes reporting delay with learning.
Require three readings from the same intervention
The diagnostic column should move in the expected direction, the other loss column should remain interpretable, and the business outcome must justify the new auction set.
Budget, bid target, ad/landing relevance or a clearly defined targeting restriction.
Use the same campaign level, network and comparable date window.
Qualified conversions, revenue, contribution or another business-owned result.
For a budget test, raise the budget modestly while holding bidding targets, targeting and creative stable. The expected diagnostic result is lower Search lost IS (budget) and higher impressions. The required business result is acceptable marginal efficiency. If spend increases while value does not, the column was accurate about capacity but irrelevant to profitable growth.
For a rank test, change one auction lever. That may be a tCPA/tROAS target, bid, ad-to-query alignment, landing-page experience or asset coverage. Do not change all of them together. If rank loss falls, you need to know which intervention earned the additional auctions. The experiment discipline is similar to the control logic in our AI Max within-campaign experiment guide, but the unit here is a delivery constraint rather than a campaign variant.
Do not confuse impression share with top-of-page placement
Search impression share asks whether the ad received an impression among estimated eligible opportunities. Search top impression share and Search absolute top impression share ask whether it captured eligible opportunities in prominent locations. Impression rate metrics answer a different question again: of the impressions already received, what percentage appeared at the top or absolute top.
Google explicitly warns that top impression rate can sometimes fall when bids increase because higher bids open more competitive auctions in worse positions. This is another version of the moving-denominator problem. A campaign may gain impressions and conversions while a prominence rate declines. Choose the metric that matches the decision instead of treating every percentage as a score to maximise.
Lost impression share tells you where estimated delivery was constrained. Only conversion and value evidence can tell you whether removing the constraint was worth buying.
A practical reporting view for the next review
At campaign level, add Search impr. share, Search lost IS (budget), Search lost IS (rank), impressions, cost and the outcome metric used to control investment. Segment the diagnosis only after checking whether each column is available at that reporting level. Google limits the budget-loss metric to campaign reporting, and custom-column exports cap some values: below 10% can appear as 9.99% or blank, while above 90% can be represented as 90.01%.
Then annotate every material change. A new budget, target, broad-match expansion, negative list, location setting, asset set or landing page can alter eligibility and auction strength. Without the annotation, a falling loss percentage may be credited to the wrong intervention. This is especially important after target-based bidding changes; our target-change audit explains why budget and target adjustments need separate ownership.
Questions about Search Lost IS budget and rank
Should Search impression share plus both loss columns equal 100%?
At a consistent Search campaign scope, the three values describe captured eligible impressions and losses attributed to budget or rank, so they are commonly used as an approximate partition. Rounding, unavailable values and reporting thresholds can prevent a clean displayed total.
Why is Search lost IS (rank) blank at ad-group level?
Google does not show ad-group Lost IS (rank) when the campaign ran out of budget at any point in the selected period. Check the campaign view and use a clean date range before concluding that rank is healthy.
Does high Search lost IS (budget) mean I should increase budget?
It means budget constrained estimated eligible delivery. Increase budget only when the campaign meets its business target and a controlled test shows acceptable marginal conversions or value.
Can I fix Search lost IS (rank) without raising bids?
Yes. Ad Rank also incorporates auction-time quality, landing-page usefulness, thresholds, search context and expected asset impact. With automated bidding, changing the target may be more relevant than a keyword bid.
Official Google Ads sources
Google Ads materials checked 30 August 2026
- About impression share — denominator, eligibility and campaign-type behavior.
- Get impression share data — column definitions, availability and reporting delay.
- About Ad Rank — auction-time inputs, eligibility and position.
- About top and absolute top metrics — placement share versus impression rate.
- Fix low traffic in Search campaigns — serving diagnosis and rank loss.
- About columns in custom columns — export fields and displayed value caps.
