Impression share (IS) is the fraction of available auctions you actually showed up for: impressions received, divided by the impressions Google estimates you were eligible to receive. Eligibility is Google's estimate — based on your targeting, approval status, bids and quality — which makes the metric a modelled denominator under a measured numerator. It is as close as paid search gets to market share, and like market share it is a number whose meaning depends entirely on what you were trying to buy.
How impression share works
The reporting family has structure worth learning once:
- Search impression share — the headline figure, available per campaign, ad group or keyword.
- Top and absolute-top impression share — the same idea restricted to premium real estate: appearing above the organic results, and appearing first of all.
- Lost to budget — auctions missed because budget was exhausted or pacing throttled entry.
- Lost to rank — auctions missed because your Ad Rank — bid and quality, roughly — did not clear the bar.
The two "lost" metrics are the diagnosis, and they are different diseases with different cures. Lost to budget is a money problem: the account wanted in and could not afford to enter — the levers are budget and pacing, or a deliberately smaller map. Lost to rank is a competitiveness problem: you could afford the auction and did not qualify strongly enough — the levers are bids, targets, and the quality trinity of ads, relevance and landing pages. Spending more cures the first and merely subsidises the second.
Small print worth knowing: very low shares are floored in reporting — Google shows "below ten percent" rather than a precise tiny figure — so arithmetic on the extremes is approximate by design.
When chasing impression share is wrong
IS is a coverage metric, not a profit metric, and maximising coverage is only sometimes the mission. On brand terms, near-total share is usually right — absence from your own name hands cheap intent to rivals. On prospecting, the last stretch of coverage is the most expensive stretch: the auctions you lose to rank are, by construction, the ones the system priced you out of, and buying into them raises average costs. An efficiency-targeted campaign will rationally cede share — that is the target working, not failing. Chase impression share when visibility is the strategy; read it as context when profit is.
The weekly pull that makes it an early warning
Impression share earns its keep as an early-warning channel, because its movements often precede the metrics everyone stares at. A weekly conversational pull — "impression share, lost-to-budget and lost-to-rank by campaign, against the previous period; flag material shifts" — catches stories while they are cheap. Rising lost-to-budget says demand outgrew the budget. Rising lost-to-rank with steady bids says the auction toughened, and auction insights will usually name the domain doing it.
A falling share is also the natural first stop in any "why did performance drop?" investigation — one early check that replaces a day of speculation with a sentence of evidence. The metric was always available. The habit of actually watching it is what a connected assistant makes cheap.