When a grant account spends $1,800 instead of $10,000, everyone reaches for the bid strategy. Impression share tells you whether that is even the right problem — it separates "we are losing auctions" from "there are no auctions to lose", and those need opposite fixes.
What it measures
Search impression share is the impressions you received divided by the impressions you were eligible to receive. Eligible means: someone searched a term matching your keywords, in your targeted location, at a time your ads were running.
At 20% impression share, four out of every five eligible searches showed somebody else's ad instead of yours. At 85%, you are showing nearly everywhere you could.
The two columns that matter more
Impression share alone tells you there is a gap. Two companion metrics tell you why, and they point in completely different directions.
| Metric | What it means | The fix |
|---|---|---|
| Lost IS (rank) | Your ad was eligible but did not rank high enough — bid or Quality Score too low | Smart Bidding, better ad relevance |
| Lost IS (budget) | Your daily budget ran out before the day did | Raise the daily budget |
On a grant account, "lost to budget" is unusual — the daily budget is normally set at $329 and nothing gets near it. Almost all lost impression share on grant accounts is lost to rank, which is the $2.00 manual bid cap doing exactly what it does.
Add these columns before you diagnose anything. Campaigns view → Columns → Modify columns → Competitive metrics. Add Search impression share, Search lost IS (rank) and Search lost IS (budget). Most accounts have never had them switched on.
The reading that changes the plan
Look at impression share alongside how much the account spends. Four combinations, four different problems:
- Low IS, low spend. You are losing auctions. This is the classic grant account and the fix is lifting the $2.00 cap with Smart Bidding.
- High IS, low spend. The harder case. You are winning nearly every auction available and there simply are not many. The keyword list is too small, too obscure, or the geography is too narrow. Bidding changes will do nothing.
- Low IS, high spend. Healthy. Plenty of demand and you are capturing part of it. Expand carefully.
- High IS, high spend. The grant is close to maxed. Time to think about paid search beyond the grant.
That second row is the one people misdiagnose most, and it is why this metric is worth the five minutes. An account at 78% impression share spending $1,500 does not have a bidding problem. Switching bid strategy will change nothing, and the team will conclude Smart Bidding "does not work" when the real answer is that the account needs more keywords.
Where to look at it
Read it at three levels, and expect them to disagree:
- Account level for the headline. One number, monthly.
- Campaign level to find which campaign drags the average. Usually one does most of the damage.
- Keyword level for the terms you actually care about. Your most important five or ten keywords deserve individual attention — a 12% impression share on your single highest-intent term is a specific, fixable problem.
What impression share does not tell you
It says nothing about whether the impressions were worth having. An account can achieve 90% impression share on terms nobody valuable searches, and that is not success — it is a very thorough way of reaching the wrong people.
It also has no relationship to click-through rate. Chasing impression share by widening match types will raise impressions and sink CTR, which is how accounts walk into the 5% rule while trying to fix underspend. Read the two together, always.
A reasonable target
There is no policy threshold here — impression share is diagnostic, not a rule. As a working guide: on your core mission terms, 60–80% is healthy for a well-run grant account. Below 40% on a term that matters means you are effectively absent from that search.
What to do about it depends entirely on which of the two lost-IS columns is carrying the number, which is the whole reason to add them.