Why Google Ads is not spending your budget: the order of checks

Low spend is often the system giving the right answer

Unspent budget is not lost money. It is a ceiling you set and that your target does not currently need. Google says so openly in the API documentation for the Maximise conversion value strategy: with a high target ROAS, it may not be able to spend the full budget. The help pages also warn that setting a target that is too high may limit the traffic your ads get.

So I would turn the question around. It is rarely about how to get the budget spent, but whether the gap between target ROAS and break-even ROAS still leaves room for volume that brings in contribution margin. If the target is close to break-even, it stays where it is, and the gap to the budget is a figure for financial planning, not an account problem.

A sudden drop and persistent underspend are two different cases

Persistent underspend points to the target or the market. With a sudden drop, on the other hand, something has changed, because the target was the same last week. Lowering it then treats the wrong problem. In this case, first check what has moved:

1. Changes in the account. If the campaign shows "Eligible (Limited)" after a change of bid strategy or a budget increase, Google says to allow 7 to 10 days for the system to adjust. I would not change anything else within that period.
2. Tracking. According to Google, a sudden drop or increase in tag fires for a conversion action used by Smart Bidding can lead to less serving and spend. The quickest test is to reconcile with the orders in your shop system.
3. Feed. Compare the number of disapproved products with the figure before the drop, especially among your top sellers.
4. Price. Has your price or the market price of your most important products moved, for example because a promotion has just ended?

If you work with Performance Max, the diagnosis for a suddenly weak PMax takes it from here.

Feed and price work through the same lever as the target

According to Google, future conversions and their values are predicted, and the bids are derived from them. A simplified calculation helps to put this in context; it is my assessment, not a Google formula: affordable bid equals expected conversion value per click divided by target ROAS.

Suppose a shop has an order value of €80:

Rows two and three lead to the same bid. If the expected conversion rate falls, it acts like a higher target, even though nobody touched it. Lowering the target ROAS at that point means paying for a price or feed error with margin. That is why the rule in the account is: read the target value first, change it last. The calculation only shows the direction; actual bidding happens per auction with many signals.

With persistent underspend, check in this order

Feed and price only concern Shopping and PMax. Pure Search campaigns jump from step 1 to step 4.

1. Target value: is your target holding back delivery?

If the campaign shows "Limited by bid strategy", Google names bid limits or targets such as a target ROAS that is too high as common causes. The impression share columns tell you more. Search lost IS (budget) and Search lost IS (rank) measure how often ads on the Search Network were not shown because of insufficient budget or poor Ad Rank.

A practitioner's reading, which is not a Google statement but which I share: high loss due to rank with zero loss due to budget, under a strict target, shows the limit you set yourself. The money is there, but the affordable bid wins too few auctions. Without a target, this reading does not apply, and the rank loss is more likely due to bids or ad quality.

2. Feed status: are the products that sell missing?

A product with the status "Not approved" cannot be shown on Google, and one marked "Limited" only appears in certain cases, for example only in certain countries. The percentage across the whole range is of little help here. I would set the affected items against revenue over the last few months: 10% of slow sellers blocked does not matter, three blocked bestsellers can leave the campaign running dry. If you ship to several countries, check the limitation per country, including Switzerland.

3. Price position: are you above the benchmark?

Merchant Center shows under Analytics, in the Pricing tab, how your price compares with the benchmark. Google defines the benchmark as the average price that typically leads to more successful ad auctions, impressions, clicks or conversions. Without a valid GTIN, this data is not available, and all you can do is compare your top sellers by hand.

A price above the benchmark does not yet prove the cause. It is, however, the point at which the matter leaves the hands of whoever runs the ads: a price cut costs margin on every order, including those from other channels. Possible responses are adjusting the price, taking the product out of the campaign or consciously accepting lower spend, and that is the retailer's decision. There is more on this under price position and Shopping reach.

4. Demand: is the market simply exhausted?

Demand comes last because it can only be judged properly once the other three brakes have been ruled out. A Search impression share of around 95 to 100% suggests, by the same practitioner's reading, that you are getting almost everything there is. Lowering the target then achieves little, and a higher budget achieves nothing at all. This limit only applies to what the campaign covers. If you target narrowly or advertise only a few products, you have exhausted your own selection, not the market. The commercial answers are a larger range, more countries, a different channel or an honest plan with less spend.

Lowering the target is a margin decision

According to Google, gradually lowering the target ROAS lets the bid strategy enter more auctions and generate more volume. That says nothing about whether it pays off. What counts is the ROAS on the additional spend, not the new average.

Suppose the contribution margin before advertising is 25%, which puts the break-even ROAS at 400%. The campaign spends €1,000 and generates €5,000 in revenue:

Variant A is clearly above break-even on average and still earns less, because the additional €500 only brings in 350%. The rule that follows: the target stays lower if the ROAS on the additional spend is above your break-even. Otherwise it goes back up.

You need patience for this comparison. According to Google, bids react immediately, but reaching the new target takes 1 to 2 conversion cycles. How strongly and how often to intervene is covered under changing target ROAS.

If you deliberately advertise below break-even, for example for new customers with demonstrable follow-on revenue or to clear stock, you are calculating against a different limit. Seasonality also distorts the before-and-after comparison: a rise ahead of Black Week proves little about a lowered target. If you do not have a break-even figure yet, that is the first step.