How Much Should a Search Term Spend Before I Block It?

A search term has been spending for weeks without a single conversion, and your cursor is on the exclude button. Block it now and you might cut a query that was close to converting. Leave it and the waste compounds, with only gut feel telling you which mistake you are making.

There is no universal dollar amount a search term should spend before you block it. Block irrelevant terms on sight, whatever they have spent. For relevant terms the line is relative: let each spend roughly one to two times your ad group's typical cost per conversion without converting, confirm it has never converted, then exclude it.

Why does a fixed dollar amount fail?

The same spend means different things in different accounts. Suppose an ad group usually converts at $25. A term that has spent $60 with nothing to show has had more than two fair chances and missed both. Now suppose a B2B ad group where a qualified lead costs $400. That same $60 is a handful of clicks, nowhere near enough to judge anything.

A flat threshold is too strict for the expensive account or too lenient for the cheap one. Across a portfolio of client accounts, it is usually both at once.

Click counts mislead in the same way. If your ad group converts at a few percent, most runs of three clicks will produce nothing. Three clicks with no conversion is the expected outcome, not a verdict on the term. The real question is whether the term has had enough chances that its silence means something. The natural unit of a "chance" is your cost per conversion.

Conversion lag adds a third distortion. Spend from the past few days may simply not have converted yet, especially where buyers take weeks to decide. Judge a term on spend that is old enough to have produced a conversion if it was ever going to.

Should irrelevant terms get any budget at all?

No. Relevance is decided before spend ever enters the picture. Consider a job seeker, someone hunting for a free version, a user looking for a competitor's login page, or an existing customer searching for support documentation. None of them is going to buy from this ad. Waiting for them to cross a threshold just pays for proof you already had. Exclude them the moment you see them, even at a single click.

The spend threshold exists for the grey zone, where terms could plausibly convert. "How to" queries can come from someone doing it themselves or from a buyer researching before they purchase. A competitor's brand name can overlap with a generic word for what you sell. These terms earn the spend test, because their intent cannot be read from the words alone.

Where you genuinely cannot tell, leave the term running rather than reaching for a broad negative. One irrelevant term costs you a few wasted clicks. A badly scoped negative can quietly block every converting variant that shares a word with it. How tightly you scope the exclusion matters as much as when you add it, and choosing the right negative keyword match type decides how much collateral damage a block does. Sorting terms into these buckets is the first half of the wider job of working out which search terms waste your budget. The threshold handles the second half.

How do you set the line for relevant terms?

Start from the ad group's typical cost per conversion, not the account average. Ad groups serving different products or intents convert at very different costs. A threshold borrowed from the account average will misjudge most of them.

Then apply three bands:

• Below one CPA with no conversions: the term has told you nothing yet. Leave it alone.
• Between one and two CPAs: put it on a watch list. Check its click-through rate and whether the landing page matches the query.
• Past two CPAs with zero conversions, and no conversion in its history: this is a strong candidate for a negative.

Adjust the bands for your economics. If your conversion rate is low, the bands should widen, because each conversion takes more clicks to arrive. If one conversion is worth far more than your average, such as a large contract or a high-value order, you can afford to let a promising term run longer before writing it off.

The data window follows from the threshold rather than the calendar. It has to be long enough for a term to reach two CPAs of spend. In a high-volume account that happens quickly. In a low-volume one it can take months, and judging from last week's report means judging on noise. Sort the search terms report by cost, then look down the conversions column for terms with real spend and nothing beside it.

Before you block anything, check the term's full history, not just the current window. A query that converted last quarter and has gone quiet may be seasonal. Excluding it now means it will not be there when demand returns.

When should a non-converter stay live?

Some terms cross the line and should still not be blocked, because the term is not the problem.

When everything relevant is failing at once. If several high-intent terms have all spent past threshold with zero conversions, suspect your conversion tracking or your landing page before you suspect the queries. Blocking them treats a measurement fault as a targeting fault.

When intent is strong but the page is weak. A term that exactly describes what you sell, gets clicks and never converts is usually pointing at the offer, the price or the page. Fix the funnel first.

When a softer move exists. A relevant term that converts too expensively can often be isolated as its own exact-match keyword with a lower bid. You keep the occasional conversion without funding the volume.

When the waste is in queries you cannot see. The search terms report does not show every query that triggered your ads. If your visible terms convert well but the campaign average lags behind them, the losses sit in the hidden queries. No negative on a visible term will reach those. Tightening match types will.

How often should you check against the line?

A threshold is only as good as the gap between reviews. The rule that follows is simple: review often enough that no term can spend past your threshold between checks. A fast-spending account reviewed fortnightly can let a bad term burn through several CPAs before anyone looks. A slow account reviewed daily just produces noise, and with it the temptation to block on too little data.

The work itself is repetitive arithmetic: compare each term's spend to its ad group's cost per conversion and flag the ones past the line. That is why it slips when the week gets busy. SproutMe Companion answers questions grounded in live spend and performance data. "Which search terms have spent more than twice their ad group's CPA without converting?" becomes something you ask on any morning, not a report someone builds on Friday.

Conclusion

The amount a search term should spend before you block it is not a dollar figure. It is a multiple of what a conversion normally costs in that ad group. Irrelevant terms get nothing and go immediately. Relevant terms get one to two CPAs to prove themselves, judged on spend old enough to have converted and checked against their full history. Before any of it, confirm the problem is not your tracking, your page or the queries the report hides.

See how SproutMe Execute adjusts bids and budgets continuously from performance data, within spend limits a human sets.