Search terms optimization documentation

#What it does

The Search Term Exclusion Tool scans every search term that has triggered your ads and compares its performance against the thresholds you set. When a term crosses a waste threshold (too expensive per conversion, too little return, or no conversions at all), Adnine either logs it as a suggestion or applies the exclusion directly to Google Ads, depending on your chosen mode. The same performance logic powers both modes: suggestion mode keeps all outcomes review-only, while execution mode turns trusted rules into real Google Ads exclusions.

Search term exclusions can be reviewed before action or automated once your rules are trusted.

Why it matters

Reduce wasted spend

Catch irrelevant or over-budget queries before they accumulate cost, and exclude them at campaign or ad group level automatically.

Keep converting terms safe

A conversion threshold protects high-value queries from being excluded even if another metric looks weak on a given day.

Full audit trail

Every flagged term is logged with its metrics, exclusion level, and status so your team can review what happened and why.

Scale across accounts

Rules work globally, per team, or per account. One setup handles the whole portfolio in execution mode.

#Choose a mode

Every rule runs in either suggestion mode or execution mode. The mode controls whether Adnine writes the exclusion to Google Ads or just logs it for your team to review. You can run the same rule in both modes at different stages of trust.

Suggestion mode vs. execution mode

Mode	What Adnine does	Best suited for
Suggestion mode	Flags matching search terms as suggestions in the log and email summary. No change is made to Google Ads.	New rules, threshold calibration, client-approval workflows, or any account where you want a human decision before exclusions happen.
Execution mode	Applies campaign-level or ad group-level exclusions to Google Ads automatically. Logs and email summaries are still generated.	Rules that have been validated through suggestion mode and are ready to run without manual sign-off every cycle.

#How to set up a rule

Rules define the performance criteria that make a search term eligible for exclusion. You can configure a rule globally for all accounts, restrict it to one team, or pin it to a specific account, depending on your workflow.

Set volume guards first Define minimum impressions, clicks, or spend before any efficiency metric is evaluated. This prevents low-volume terms from being excluded on noisy data.
Define efficiency thresholds Set a maximum CPA, minimum ROAS, or both. A term must exceed your waste boundary and meet the volume guard before Adnine flags it.
Add a conversion safeguard Specify a maximum conversions threshold. Any term that has hit or exceeded that number is protected from exclusion even if CPA or ROAS looks poor.
Configure the delay window Set how many days a term must be present in the data before it can be evaluated. This gives newer queries time to accumulate meaningful performance.
Scope the rule Optionally filter by campaign name, ad group name, or keyword match type to target the rule at the specific part of the account where it should apply.
Choose suggestion or execution mode Start in suggestion mode. After reviewing a few rounds of flagged terms, switch to execution mode when you are confident the thresholds are right.

Rule settings reference

All settings are optional except the core efficiency threshold. Combine them to build rules that are precise enough to avoid false positives.

Setting	What it does	When to use it
Minimum impressions	Requires a term to have reached this impression count before eligibility is checked.	Always set this. Prevents a single irrelevant impression from triggering an exclusion.
Minimum clicks or spend	Additional volume gates that must be met alongside impressions.	Use for high-spend accounts or when CPA is the primary threshold.
Maximum CPA	Excludes terms whose cost per conversion exceeds this value.	Use when you have a target cost per acquisition for the campaign.
Minimum ROAS	Excludes terms that fall below this return on ad spend.	Use for revenue-focused e-commerce or lead-value campaigns.
Maximum conversions before protection	Blocks exclusion for any term that has already converted this many times or more.	Always recommended. Prevents removing terms that have historically performed well.
Delay window (days)	How many days a term must exist in the data before it is evaluated.	Set a delay for accounts with delayed conversion attribution, such as 7, 14, or 30 days.
Include paused campaigns	Controls whether performance data from paused campaigns counts toward the threshold.	Enable only if you want paused campaign data to influence current exclusion decisions.
Campaign / ad group / match type filter	Restricts the rule to matching entities only.	Use to isolate rules by campaign type (brand vs. non-brand), match type (broad only), or account section.

#Read the dashboard

The dashboard gives you a quick health check on your exclusion activity: how many terms were actioned, where the spend savings came from, and when the next automated run is scheduled.

Dashboard signals and what they mean

Signal	What it tells you	How to use it
Search terms excluded in the last 7 days	Recency of rule activity.	A sudden spike may mean a rule is too aggressive; zero may mean thresholds are too conservative.
Total spend saved by exclusions	Estimated waste removed, calculated from the excluded terms' historical spend.	Use as a headline efficiency metric for client or management reporting.
Top campaigns and ad groups affected	Where most flagged terms are concentrated.	Helps you decide whether to tighten scope filters on busy rules or create separate rules per campaign.
Scheduled run frequency and next run time	When the next automated evaluation is due.	Confirm the schedule matches your reporting cadence and review cycles.
Overall log of excluded terms	The full audit trail with per-term metrics.	Scan for patterns: recurring terms, specific match types, or campaigns that need structural fixes rather than more exclusions.

#Review terms in Search Term Exclusion Explore

Search Term Exclusion Explore is the detailed review table for every term your rules have flagged. Use it before approving automation and as an ongoing audit surface to confirm your exclusions are doing the right work.

Fields available in Explore

Field	What to look at
Search term	Read the actual user query. Check whether it is genuinely off-topic, too broad, or just underperforming in this window.
Impressions, clicks, spend, conversions	Confirm there is enough volume to make the exclusion judgment meaningful.
CPA and ROAS	Compare against your account target. A term above your CPA cap with zero conversions is a clear exclusion candidate.
Matched keyword and match type	Understand why this query triggered your ad. Broad or phrase match often explains unexpected search terms.
Account, campaign, and ad group	Decide whether the problem is isolated to one ad group or indicates a wider match type issue across a campaign.
Exclusion level	See whether Adnine applied the exclusion at campaign or ad group level.
Status	Distinguishes current suggestions from already-executed exclusions.

Filter by status to separate suggestions from executed exclusions. Sort by spend or CPA to prioritize the highest-impact terms for manual review. Use match type and ad group filters to isolate patterns that may point to a structural fix rather than another exclusion. If you repeatedly see the same type of query being flagged from a broad-match keyword, the better fix may be to tighten the match type or add a negative keyword at the keyword level rather than excluding each term individually.

#When to add a negative vs. exclude a search term

The Search Term Exclusion Tool handles automated per-term exclusions based on performance data. But some situations call for a different approach. Use this table to decide the right action when a problematic query appears. For best results, pair Search Terms Optimization with the Keyword Additions tool: exclusions remove waste while keyword additions grow coverage with high-performing terms. The two tools work in opposite directions on the same data source.

Exclusion vs. negative keyword decision guide

Situation	Recommended action	Why
A specific query consistently exceeds your CPA threshold across multiple cycles	Let execution mode exclude it automatically	Performance-based exclusion at campaign or ad group level is exactly what the tool is built for.
An entire category of irrelevant queries keeps appearing (e.g. brand names you don't carry, job seekers, unrelated industries)	Add a negative keyword list at campaign or account level	A structural negative prevents the entire category from ever matching, which is more efficient than excluding individual terms one by one.
A term is converting well but has a high CPA this week	Do nothing. The conversion safeguard in your rule should protect it.	Short-term CPA spikes on converting terms are often noise. The maximum-conversions threshold blocks exclusion for terms with proven history.
A broad-match keyword is triggering dozens of irrelevant queries	Switch the keyword to phrase or exact match, or add targeted negatives	Excluding individual terms from a badly-matched keyword is a losing race. Fix the root cause at the keyword level.
A term had low spend in the past but is accelerating in spend with no conversions	Rule will flag it once volume thresholds are met	Set a minimum spend threshold in your rule so the term is caught early but only after enough data is present.

#Logs and email summaries

After every run, Adnine records a log entry and can send a summary email. Together, these keep your team informed about what happened without requiring anyone to log in and check manually. Email gives your team a run-level summary they can act on quickly; the log provides term-level evidence for deeper review, client reporting, or audits.

What is captured after each run

Record type	What it contains
Email summary	Count of suggested or excluded terms, top campaigns affected, spend saved in execution mode, and a downloadable report link with all term-level metrics.
Log entry	Per-term record with date of identification, impressions, spend, conversions, CPA, ROAS, exclusion level (campaign or ad group), and current status.
Status tracking	Each term is marked as suggested or excluded so review-only and automated outcomes stay separate in the same log view.

#Pitfalls and limits

If paused campaigns include high-CPA terms from a period they were active, including paused campaign data in your rule evaluation may flag terms that are no longer running. Use the paused campaign toggle carefully to avoid distorting your thresholds with stale data.

#Best practices

Always start in suggestion mode. Review at least two to three run cycles before switching to execution mode.
Set minimum spend, click, and impression thresholds on every rule. Never evaluate CPA or ROAS on fewer than 20 to 30 clicks.
Add a maximum-conversions safeguard to protect terms with proven conversion history.
Use the delay window to account for attribution lag, especially in accounts with long sales cycles or offline conversions.
Review Search Term Exclusion Explore weekly. Sort by spend or CPA to find the highest-impact terms first.
When the same category of query keeps appearing, fix the keyword match type or add a structural negative instead of relying on per-term exclusions.
Use email summaries for team and client transparency; use the log for term-level audits and quality control.
Check the dashboard for spend-saved trends. A sudden increase may signal a match type problem that needs a structural fix.
Treat exclusions as a last resort for structural issues: the tool is most powerful alongside good keyword hygiene. Structural negatives, match type discipline, and the Keyword Additions tool handle the foundation.

#Frequently asked questions

What is the difference between suggestion mode and execution mode?

In suggestion mode, Adnine flags search terms that meet your waste criteria but makes no changes to Google Ads. The suggested terms appear in the log and email summary for your team to review. In execution mode, Adnine applies campaign-level or ad group-level exclusions to Google Ads automatically. Logs and emails still run in both modes.

Can I undo an exclusion that was applied in execution mode?

Yes. The fastest way is the in-app Undo Run, which reverses the most recent run. You can also remove an exclusion manually: check the log to find the term and its exclusion level (campaign or ad group), then delete the negative keyword entry in the corresponding Google Ads entity. Adnine does not reverse exclusions automatically over time.

How do I protect a search term that converts well but has a high CPA this week?

Set a maximum-conversions threshold in your rule. Any term that has converted at or above that number is protected from exclusion regardless of its current CPA or ROAS. This prevents short-term efficiency dips from removing historically strong queries.

Why is a search term I added as a negative still appearing in the Explore log?

The Explore log shows all terms flagged by your performance rules, including ones you have already manually excluded in Google Ads. The log is based on performance data, not on the current negative keyword list. Filter by status to distinguish current suggestions from previously actioned terms.

How often does the tool run?

The run frequency is shown on the dashboard alongside the next scheduled run time. The schedule is configured as part of your Adnine setup. Check the dashboard to confirm the cadence matches your reporting and review cycle.

Can I apply different rules to different campaigns or ad groups?

Yes. Use campaign name, ad group name, or match type filters in the rule settings to scope each rule to the part of the account where it should apply. You can run separate rules with different thresholds for brand campaigns, non-brand campaigns, or specific product lines.

What happens if I have no conversions in my account?

If your account does not track conversions, use spend-based rules: set a maximum spend threshold and rely on minimum impressions or clicks as the volume guard. CPA and ROAS rules require conversion data to function correctly.

Search terms optimization