What is impression share?

The usual metrics of an account count what happened: impressions, clicks, cost, conversions. None of them says anything about what did not happen. An advertiser can run a campaign with good numbers and still be absent from most of the searches that matter, without any spend report giving it away.

Impression share covers exactly that gap. It compares what you achieved with what you could have achieved and expresses the result as a percentage. A value of forty per cent means that in six out of ten opportunities your ad was never seen.

The decisive word is “eligible”. The denominator is not every search in the world related to your product, but an estimate of the auctions your ad could have taken part in given its approval status, your targeting settings and its quality. If your campaign only targets Barcelona, searches made from Seville do not count. Google also leaves out auctions in which you would have needed a disproportionate bid increase, in the order of a thousand per cent, because including them would turn the percentage into a figure of no practical use.

The ratio is simple in form: impressions obtained divided by estimated eligible impressions. What comes next is the interesting part, because Google breaks the missing portion into two reasons. Impression share lost to budget captures the time your ads stopped serving because the campaign ran out of money, and it is only available at campaign level. Impression share lost to rank captures the auctions in which the ad was not competitive enough. The three figures together approach one hundred per cent, so a glance at which one dominates is enough to know where to start.

There is a second family of columns that answers a different question. Top impression share measures what proportion of eligible impressions you achieved in the ad zone above the organic results. Absolute top impression share measures the very first slot of that zone. Each has its own loss reasons for budget and for rank. Appearing often is not the same as appearing high, and these columns are what separate the two.

The data is reported by campaign, ad group, product group and keyword, and takes one to two days to settle. The campaign types that report it are Search, Shopping, Performance Max, Hotels and travel campaigns with Things to do ads.

It answers a question the rest of the metrics dodge: how much available market you are letting slip. Two campaigns with the same cost per acquisition can sit in opposite situations, one brushing its ceiling and the other capturing a small fraction of the demand it could reach.

Breaking down the loss turns that observation into a concrete decision. If most of it is lost to budget, the lever is money or the split between campaigns, and it is worth checking whether spend is being consumed in the least profitable hours of the day. If it is lost to rank, adding budget will change nothing, because the ad is not even competing well in the auctions you already pay for.

It is also the usual route to size an advertising market before scaling investment. From the current share and the volume obtained you can approximate how many impressions are left on the table, always with the caution that the denominator is a Google estimate and not a closed count. There is no universally good value: what suits you depends on whether you want maximum coverage in your main category or selective presence in secondary queries.

• Always read the share alongside its two loss reasons. The percentage on its own warns that something is missing, but does not say what to do about it.
• Analyse it by campaign type and by segment, never summed into an account total, because each type is calculated separately and mixing them produces a meaningless figure.
• Compare the general share with the top share when visibility is the objective. The difference between the two shows how much presence is stuck in the lower positions of the page.
• Cross the evolution of the percentage with the changes you made in the account, including targeting changes, because narrowing the targeting raises the share without anything having improved.
• Before reading a drop as your own failure, check whether search volume in the category changed or a new competitor entered the auction.
• Set the objective according to the role of each campaign and the margin of the product, and write it down before opening the report.

• Reading impression share as market share. It measures auctions your settings made you eligible for, not the total demand in your sector.
• Raising the budget when the loss is to rank. The extra money goes unspent, because the problem lies in quality, in the bid or in relevance inside the auction.
• Interpreting the dash in the column as a zero. Google hides the figure when traffic volume is low or the keywords are new, and that absence does not mean there were no impressions.
• Celebrating a rise that actually comes from having narrowed the targeting. A smaller denominator produces the same visual effect as better performance.
• Copying someone else's targets. Google publishes no benchmark values, and what is reasonable for a well-known brand is not reasonable for a generic acquisition campaign.