Definition
Impression share is the number of impressions an ad actually received divided by the number it was eligible to receive, expressed as a percentage. Google estimates the denominator from the auctions the campaign qualified for, using targeting, approval status, bids and quality. It is the metric that answers how much of the available demand a store is currently reaching, and it is reported alongside the two reasons for missing the rest: budget and rank.
Formula
Impression share = Impressions received ÷ Estimated eligible impressions × 100
- Impressions received
- Times the ad was shown in the period for the campaign, ad group or product group being measured
- Estimated eligible impressions
- Google's estimate of the auctions the ad could have entered, based on targeting, approvals, bids and quality signals
- Lost impression share
- The remainder, split into share lost to budget and share lost to rank, which together with impression share adds to 100 percent
One Shopping campaign, read with its lost share
- Impressions received
- 182,000
- Estimated eligible impressions
- 520,000
- Impression share
- 182,000 ÷ 520,000 = 35 percent
- Share lost to budget
- 41 percent
- Share lost to rank
- 24 percent
- What it points at
- The campaign is capped for most of the day, so raising the budget reaches more of the demand than raising bids would
- The same campaign after the cap is lifted
- Lost budget share falls, lost rank share rises, and bids become the next constraint
Illustrative figures. The split is what makes the metric actionable: a low impression share caused by budget and a low impression share caused by rank call for opposite responses.
Why it matters
Impression share matters because it is the clearest measure of headroom a store has. A campaign hitting a strong return on a small share of the available auctions is a growth opportunity that no efficiency metric will reveal, since ROAS looks the same at 20 percent share as at 80. Read with the lost share columns, it also tells you which lever to pull: budget when the campaign runs out of money before the day does, bids, feed quality or relevance when it is being outranked. On brand terms it doubles as a defence report, showing how much of the demand for the store's own name is being captured.
Where it goes wrong
- Treating 100 percent as the goal: the last stretch of any auction is the most expensive, and buying it usually costs more than the incremental orders return
- Comparing impression share across campaign types as if it meant the same thing, when a brand campaign in a narrow auction and a broad prospecting campaign face very different denominators
- Reading it without absolute volume: a high share of a tiny eligible pool is a small opportunity, and the metric alone does not say how big the market is
- Forgetting that the denominator is an estimate that moves: a competitor entering or leaving the auction changes your reported share with no change on your side
- Assuming the columns are always available, since Google suppresses these figures where data is thin and reports them differently across campaign types
Questions about impression share
What is the difference between impression share lost to budget and lost to rank?
Lost to budget means the campaign was eligible but stopped serving because the daily budget ran out or was being paced. Lost to rank means the ad entered auctions and did not place highly enough to show, which comes down to bid, quality and relevance. Budget loss is fixed with money; rank loss is fixed with bids, feed quality, landing page relevance or targeting.
Which campaign types report impression share?
Search and Standard Shopping campaigns report impression share and the lost share breakdown as standard columns, with absolute top share and click share available alongside for Shopping. Performance Max reports on a different basis, so stores that run their catalogue entirely through it have fewer of these columns and lean more on click share and on their own headroom estimates.
How high should impression share be on brand terms?
High enough that a shopper searching for your name reliably finds you, and no higher than the incremental value justifies. Much of that traffic would reach the store through organic results anyway, so the honest test is a measured one: hold the brand campaign out in a set of regions and see how much of the revenue survives before deciding what the share is worth.