The sequence is the same in almost every account. The brand campaign serves less, somebody asks why, and the answer comes back: demand for the brand has declined. It sounds plausible, it is hard to disprove – and in most cases it is wrong.
There is a single calculation that settles the question in minutes. It needs no external tools, no market research and no debate. It uses two metrics that sit in every Google Ads account, plus one division.
The short version: divide your brand campaign’s impressions by its search impression share. The result is eligible impressions – the number of auctions your campaign was allowed to enter. If that number falls, the cause sits in the account setup. If only the impression share falls, the cause is bidding or budget.
1. Why “demand has dropped” is almost always the wrong answer
Brand demand is slow-moving. It follows awareness, satisfaction and season – it does not collapse by forty percent within a few months while the business carries on as usual. When a brand campaign’s impressions fall far faster than anything else in the market, the market is rarely the explanation.
The second reason is methodological. Anyone who blames demand is making a claim they would have to prove, and in practice the supporting data is never collected. That is exactly why the counter-check from inside the account is worth doing: it is cheaper than any market research and considerably harder to argue with.
2. The metric almost nobody looks at: eligible impressions
Google reports search impression share. It answers one question: of all the auctions this campaign was eligible for, how many did it actually win?
Which leads directly to a calculation that does not appear in any standard report:
Eligible impressions = impressions ÷ impression share
This figure is the campaign’s real addressable market. It does not describe how well you bid; it describes how many auctions the campaign was permitted to take part in at all. And it has one useful property: bids and budgets do not move it. Bid more and you win more auctions – but the number of auctions stays the same.
In the worked example above, impressions fall by almost forty percent. At the same time the impression share rises sharply. Both can only be true together if the addressable market itself has collapsed – by 74 percent. The campaign wins a larger share of auctions than before, but there are barely any left.
3. The decision tree: two patterns, two entirely different causes
Once you have both metrics side by side over time, only two patterns exist – and they lead to completely different actions.
Case A – impressions fall, impression share rises or holds. Eligible impressions have collapsed. The campaign is being excluded from auctions. Cause: the setup.
Case B – impressions fall and impression share falls with them. The addressable market is stable and the campaign is losing auctions. Cause: bid, budget or Ad Rank.
A third case exists in theory – both figures fall and the market really has shrunk. In that case you have to evidence it with external data rather than assert it. Section 7 covers how.
4. Case A: the campaign is being excluded from auctions
Four mechanisms reduce eligible impressions. All four are settings, not market events.
Negative keywords. The most common cause by far, and the least visible. Section 6 covers it in detail.
Fewer or narrower keywords. Switching from broad match to exact, or removing terms, shrinks the basket immediately. That can be the right call – it simply has to be a known one.
Narrower geographic targeting. A removed source market or a reduced radius acts directly on eligibility, not on competition.
Ads that cannot serve. Disapproved ads, empty ad groups or paused elements take whole sections of a campaign out of the running. Google flags this in the campaign status as Eligible (limited) – a line most reports never export.
5. Case B: the campaign is losing the auctions
Here two further columns tell you which of three causes applies: search lost impression share (budget) and search lost impression share (rank).
Budget. If budget loss rises, the daily budget is the constraint. Easy to verify: multiply daily budget by the number of days and compare with actual spend. If utilisation sits well below one hundred percent, budget is not the cause – whatever the column suggests.
Ad Rank. If rank loss rises, the bid is too low or quality is too poor. On brand terms neither should really happen: your own brand is the most relevant possible result.
The bid target. The case that gets overlooked. A target ROAS set too high, or a target CPA set too low, throttles bids until the target looks achievable – even when budget is left over. The tell-tale pattern: high budget, low utilisation, high rank loss. It gets worse when conversion measurement is leaking value: the system then sees a lower return than reality and throttles harder than it should.
Rule of thumb: when an account spends only a fraction of its budget yet still reports high rank loss, the bid target is almost always the constraint – not the money. In that situation, verify that your conversion values are correct before touching the target.
6. The single most common cause: broad match negatives
Negative keyword lists grow over years. They get created to cut waste and are never read again. That is precisely where the risk sits.
A broad match negative blocks every query containing that term, regardless of what else is in it. A negative for golf also removes “hotel with golf course”. A negative on a district, a building name or a sub-brand can, in the worst case, block your strongest brand queries.
Two checks are worth the time.
First: export the list and filter by match type. Exact negatives are usually harmless. Focus on the broad ones – in many accounts they are a few percent of the entries and cause all of the damage.
Second: read that list against your own offering. Any term describing one of your own products, locations or sub-brands needs checking. Then pull the change history: it shows when each entry was created. If that date lines up with the start of the decline, the question of cause is answered.
7. The external cross-check – and its pitfalls
When the market claim is on the table, an outside view helps: Google Trends for the relative curve, a visibility tool such as Sistrix or Semrush for estimated search volumes. Two independent sources pointing the same way make a solid indication.
Know three pitfalls first.
Brand names are rarely unambiguous. Many brands share a place name or an everyday word, so the volume contains queries that have nothing to do with you. The way out is a basket of unambiguous brand terms – brand plus category, location or product name. They cannot be misread and every tool can evaluate them.
Noise does not create a trend. The non-brand portion of a search term stays broadly constant over the years. It shifts the level, not the direction. Anyone claiming brand demand fell while the total curve rose would have to show that the noise grew disproportionately.
Google Trends is relative. The index sets the peak of the selected period to 100. It is a shape, not a quantity – and it only supports statements about direction.
8. Six steps to an evidenced cause
Step 1 – pick the window. Compare identical periods across two years, not month against previous month. Otherwise seasonality beats every analysis.
Step 2 – add the columns. Impressions, clicks, cost, conversions, conversion value, search impression share, lost impression share (budget), lost impression share (rank), bid strategy. Segment by month.
Step 3 – calculate eligible impressions. Impressions divided by impression share, per month. This column decides the case.
Step 4 – check budget utilisation. Daily budget times days against actual spend. That either confirms or eliminates budget as the cause.
Step 5 – export the change history. Same period, no filters. It is the only source that cannot be tidied up after the fact.
Step 6 – review negative lists and ad status. Broad match negatives, disapproved ads, campaigns flagged as limited.
With access in place this takes about fifteen minutes. What you end up with is not a suspicion but a number.
9. What this means for working with an agency
The calculation changes the conversation. Instead of “why are the numbers falling?” – a question that always has a general answer – you ask one that permits only a specific one: eligible impressions are down X percent. Which setting in the account caused that?
Ask for three things, in writing: the monthly series of the columns above over twelve months, the change history for the same period, and an assessment of which account factors explain the trend. Anyone who knows their account can deliver that in an afternoon.
10. Conclusion: one division replaces one debate
Falling brand impressions are almost never a market problem. They are an eligibility problem – and eligibility is a setting, not fate. The decisive metric appears in no standard report, yet it can be derived from two existing columns in a single line.
Once you have calculated it, you stop debating demand. You start looking for the setting that shrank the basket – and you usually find it in the change history.
11. FAQ: common questions about falling brand impressions
What are eligible impressions in Google Ads?
Eligible impressions are the estimated number of auctions a campaign was allowed to enter. Google does not report them directly, but they can be calculated: impressions divided by search impression share. If this value falls over time, the campaign is being excluded from auctions – through negatives, narrower keywords, tighter geographic targeting or ads that cannot serve.
Why are my brand impressions falling even though the brand is growing?
In most cases it is the account setup rather than the market. First check whether impression share rose over the same period. If it did, eligible impressions have fallen and the cause lies with negatives, keywords, targeting or disapproved ads. If impression share fell too, the issue is bid, budget or Ad Rank.
Can a target ROAS throttle delivery while budget is left over?
Yes. With value-based bidding, the system lowers bids until the stated target looks achievable. If the target is set too high, or the account measures too little conversion value, budget goes unspent and lost impression share due to rank rises. The classic signature is a high daily budget with low utilisation.
How do I spot harmful negative keywords?
Export the list and filter for broad match. Those entries block every query containing the term. Read them against your own offering: terms describing your own products, locations or sub-brands are the critical ones. The change history additionally shows when each entry was created.
Is Google Trends enough to evidence brand demand?
Not on its own. The Trends index is relative and sets the period peak to 100; it shows direction, not volume. The statement becomes robust through a second independent source and by restricting the analysis to unambiguous brand terms rather than a brand name that has other meanings.
What data should I request from my agency?
Three things: the monthly series of impressions, clicks, cost, conversions, conversion value, impression share and both lost impression share columns over twelve months; the complete change history for the same period; and a written assessment of which account factors explain the development.
How long does this check take?
With read access in place, roughly fifteen minutes for the core calculation and one to two hours for the full root-cause analysis including negative lists and change history.
