Paid search impression share is useful because it describes how often an ad received an impression relative to estimated eligible opportunities. It is not a measure of demand, lead quality or profit. A campaign can gain share on weak queries, lose share during a deliberate efficiency decision, or show a healthy percentage while the landing page and sales handoff fail.
1. Define the decision
Decide whether the scorecard will support budget allocation, bid changes, coverage of a strategic query set, launch diagnosis or a post-agency review. Name the campaign type, network, geography, device, date range, conversion definition and business outcome.
Do not compare impression share across campaign types as if the denominators were identical. Google notes that impression-share metrics are reported separately for each campaign type. Keep Search, Shopping, Display and Performance Max in separate sections unless the reporting method explicitly explains the aggregation.
2. Understand what impression share means
Google’s About impression share guidance defines impression share as impressions received divided by estimated eligible impressions. Eligibility depends on targeting, approval, quality and auction conditions, and the estimate can change as the system changes.
Record impression share with impressions, eligible-impression estimate where available, campaign type, date range and data freshness. A missing value or dash can reflect insufficient data, a new keyword or a report delay; it is not automatically zero coverage.
3. Keep lost-by-budget and lost-by-rank separate
The same impression share can have different remedies. Lost share due to budget points to budget or pacing constraints; lost share due to rank points to competitiveness, relevance, landing experience or bidding. Use the impression-share data guidance to review which columns are available at campaign, ad-group or keyword level and which are not.
Do not infer that increasing budget will solve rank loss or that improving ads will create more eligible demand. Record the suspected constraint and the evidence required before changing it.
4. Segment by intent and commercial value
Break the scorecard into branded, non-branded, category, competitor, problem and service-specific queries. Add location, device, match behavior, landing page, product or service, and qualification status. More share on an irrelevant or low-margin query may be negative growth.
Keep query classification versioned. If negative keywords, match types or campaign structure changed, mark the before-and-after boundary. A blended account average can hide a useful high-intent pocket and a wasteful expansion elsewhere.
Preserve the query sample that supports the classification. Search terms can be omitted, delayed or grouped by the platform, and a campaign label may not identify the actual customer language. When the sample is too small to interpret, state that the segment is unresolved instead of filling it with an average from a larger campaign.
5. Connect exposure to qualified outcomes
Measure the path from eligible impressions to impressions, clicks, engaged sessions, valid forms or calls, qualified conversations, opportunities, booked work and margin. Google’s measure-your-results guidance encourages looking beyond a single campaign statistic and considering the effect of changes across paid, organic and combined traffic. Preserve the platform conversion and the business outcome as separate fields.
Use a consistent time zone and attribution window. Reconcile ad click IDs, source parameters, landing page, CRM record, status, revenue and refunds where relevant. If a join is missing, show the exception rather than allocating pipeline to the campaign by proportion.
6. Add landing and sales quality
A campaign can buy more share without producing more pipeline when the ad promise, query, landing page, form, response time or offer is misaligned. Review search term quality, page relevance, mobile behavior, form errors, call dispositions, speed of response and sales acceptance.
Treat impression share as an exposure constraint that helps explain what the campaign could have reached. It cannot explain why a qualified visitor chose not to contact you. Use qualitative evidence and row-level samples to avoid mistaking a media metric for a customer diagnosis.
7. Use an impression-share scorecard
| Layer | Metric | Question | Guardrail | | — | — | — | — | | coverage | impression share | how much eligible exposure was captured? | keep campaign type separate | | constraint | lost by budget/rank | what limits additional exposure? | do not assume remedy | | intent | query and landing segment | was exposure commercially relevant? | show counts and period | | response | clicks and engaged sessions | did the ad earn a useful visit? | inspect device and page | | quality | qualified leads or calls | did the visit meet the business definition? | preserve disposition | | pipeline | accepted opportunity and booked work | did sales accept the demand? | use stable stage rules | | economics | margin or contribution | was additional exposure worthwhile? | include costs and refunds |
Add owner, data source, join key, freshness, confidence and action threshold to every row.
8. Choose the next action cautiously
Increase budget only when high-intent coverage is constrained by budget, marginal demand is acceptable, landing and sales capacity are ready, and the measurement join works. Improve rank when the evidence points to relevance or competitiveness and the page can support the query. Reduce or restructure when the additional share is low-quality or unprofitable.
Avoid fixed “good” impression-share targets copied from another account. Auction density, geography, brand demand, seasonality, product margin and eligibility estimates make the decision context-specific.
9. Run a bounded measurement review
Select one campaign family and a fixed period. Export impression share, lost-by-budget, lost-by-rank, query segments, landing events, qualified outcomes and exceptions. Compare with the unchanged report, document the hypothesis and simulate the proposed budget or bid change before activating it.
Release a change only after the scorecard, definitions, data freshness, capacity and rollback are approved. This article remains a local draft until overlap, platform, privacy, commercial and prepublication review are complete; no bid, budget or campaign setting is changed by the draft.
How did this article land?
Choose one reaction. You can change it anytime.
