Why a structured prompt library matters
Most paid media teams waste time rebuilding the same analysis every week. A structured prompt library turns ad account review into a repeatable operating system: clear thresholds, fixed time windows, and output formats your team can execute immediately. Most paid media teams waste time rebuilding the same analysis every week. A structured prompt library turns ad account review into a repeatable operating system: clear thresholds, fixed time windows, and output formats your team can execute immediately. Most paid media teams waste time rebuilding the same analysis every week. A structured prompt library turns ad account review into a repeatable operating system: clear thresholds, fixed time windows, and output formats your team can execute immediately. Most paid media teams waste time rebuilding the same analysis every week. A structured prompt library turns ad account review into a repeatable operating system: clear thresholds, fixed time windows, and output formats your team can execute immediately.
How this guide is designed
This article contextualizes a practical prompt framework inspired by modern AI-assisted ad analysis workflows. The goal is not blind automation; it is faster diagnosis, better prioritization, and tighter execution loops across search terms, campaigns, audiences, creatives, and budget allocation. This article contextualizes a practical prompt framework inspired by modern AI-assisted ad analysis workflows. The goal is not blind automation; it is faster diagnosis, better prioritization, and tighter execution loops across search terms, campaigns, audiences, creatives, and budget allocation. This article contextualizes a practical prompt framework inspired by modern AI-assisted ad analysis workflows. The goal is not blind automation; it is faster diagnosis, better prioritization, and tighter execution loops across search terms, campaigns, audiences, creatives, and budget allocation. This article contextualizes a practical prompt framework inspired by modern AI-assisted ad analysis workflows. The goal is not blind automation; it is faster diagnosis, better prioritization, and tighter execution loops across search terms, campaigns, audiences, creatives, and budget allocation.
Module 1: campaign performance prompts
Start with account-level triage. Ask for campaigns below your target ROAS and require fields such as spend, conversions, conversion value, campaign type, and impression share. Then request an explicit decision per campaign: pause, optimize, or reallocate budget. Start with account-level triage. Ask for campaigns below your target ROAS and require fields such as spend, conversions, conversion value, campaign type, and impression share. Then request an explicit decision per campaign: pause, optimize, or reallocate budget. Start with account-level triage. Ask for campaigns below your target ROAS and require fields such as spend, conversions, conversion value, campaign type, and impression share. Then request an explicit decision per campaign: pause, optimize, or reallocate budget. Start with account-level triage. Ask for campaigns below your target ROAS and require fields such as spend, conversions, conversion value, campaign type, and impression share. Then request an explicit decision per campaign: pause, optimize, or reallocate budget.
Module 2: wasted spend and negatives
Define waste criteria up front: high cost with zero conversions, low ROAS, irrelevant intent, and geo mismatch. Ask the model to group waste by theme and produce negative keyword lists in import-ready formats. This is usually the fastest way to recover lost efficiency. Define waste criteria up front: high cost with zero conversions, low ROAS, irrelevant intent, and geo mismatch. Ask the model to group waste by theme and produce negative keyword lists in import-ready formats. This is usually the fastest way to recover lost efficiency. Define waste criteria up front: high cost with zero conversions, low ROAS, irrelevant intent, and geo mismatch. Ask the model to group waste by theme and produce negative keyword lists in import-ready formats. This is usually the fastest way to recover lost efficiency. Define waste criteria up front: high cost with zero conversions, low ROAS, irrelevant intent, and geo mismatch. Ask the model to group waste by theme and produce negative keyword lists in import-ready formats. This is usually the fastest way to recover lost efficiency.
Module 3: creative diagnostics
High CTR does not always mean high-quality demand. Ask for headline and ad component ranking by CTR and conversion rate separately. Then request pattern extraction from winners and losers, plus new copy ideas grounded in your own account data. High CTR does not always mean high-quality demand. Ask for headline and ad component ranking by CTR and conversion rate separately. Then request pattern extraction from winners and losers, plus new copy ideas grounded in your own account data. High CTR does not always mean high-quality demand. Ask for headline and ad component ranking by CTR and conversion rate separately. Then request pattern extraction from winners and losers, plus new copy ideas grounded in your own account data. High CTR does not always mean high-quality demand. Ask for headline and ad component ranking by CTR and conversion rate separately. Then request pattern extraction from winners and losers, plus new copy ideas grounded in your own account data.
Module 4: audience and targeting prompts
Benchmark each audience against account averages for CPA and ROAS. Ask for concrete bid adjustments, exclusions, and consolidation opportunities. If retargeting is involved, validate list size, membership duration, and frequency impact. Benchmark each audience against account averages for CPA and ROAS. Ask for concrete bid adjustments, exclusions, and consolidation opportunities. If retargeting is involved, validate list size, membership duration, and frequency impact. Benchmark each audience against account averages for CPA and ROAS. Ask for concrete bid adjustments, exclusions, and consolidation opportunities. If retargeting is involved, validate list size, membership duration, and frequency impact. Benchmark each audience against account averages for CPA and ROAS. Ask for concrete bid adjustments, exclusions, and consolidation opportunities. If retargeting is involved, validate list size, membership duration, and frequency impact.
Module 5: budget and scaling prompts
Scaling should be modeled, not guessed. Ask for projected outcomes when increasing budget on top campaigns, including an efficiency-loss assumption. Pair this with reallocation logic from weak campaigns to strong campaigns. Scaling should be modeled, not guessed. Ask for projected outcomes when increasing budget on top campaigns, including an efficiency-loss assumption. Pair this with reallocation logic from weak campaigns to strong campaigns. Scaling should be modeled, not guessed. Ask for projected outcomes when increasing budget on top campaigns, including an efficiency-loss assumption. Pair this with reallocation logic from weak campaigns to strong campaigns. Scaling should be modeled, not guessed. Ask for projected outcomes when increasing budget on top campaigns, including an efficiency-loss assumption. Pair this with reallocation logic from weak campaigns to strong campaigns.
Module 6: schedule, location, and device prompts
Ask for hourly, daily, and device-based breakdowns with clear cutoffs for action. Require estimated savings from pausing low-intent windows and bid adjustments for stronger periods. This produces practical optimization rules your team can apply weekly. Ask for hourly, daily, and device-based breakdowns with clear cutoffs for action. Require estimated savings from pausing low-intent windows and bid adjustments for stronger periods. This produces practical optimization rules your team can apply weekly. Ask for hourly, daily, and device-based breakdowns with clear cutoffs for action. Require estimated savings from pausing low-intent windows and bid adjustments for stronger periods. This produces practical optimization rules your team can apply weekly. Ask for hourly, daily, and device-based breakdowns with clear cutoffs for action. Require estimated savings from pausing low-intent windows and bid adjustments for stronger periods. This produces practical optimization rules your team can apply weekly.
Ready-to-use prompt examples
Example prompt A: Find campaigns below ROAS 2 and recommend first fix by priority. Example prompt B: Identify search terms above 30 dollars spend and zero conversions, grouped by waste theme with negative lists. Example prompt C: Compare last 7 days versus previous 7 days and flag shifts greater than 15 percent. Example prompt A: Find campaigns below ROAS 2 and recommend first fix by priority. Example prompt B: Identify search terms above 30 dollars spend and zero conversions, grouped by waste theme with negative lists. Example prompt C: Compare last 7 days versus previous 7 days and flag shifts greater than 15 percent. Example prompt A: Find campaigns below ROAS 2 and recommend first fix by priority. Example prompt B: Identify search terms above 30 dollars spend and zero conversions, grouped by waste theme with negative lists. Example prompt C: Compare last 7 days versus previous 7 days and flag shifts greater than 15 percent. Example prompt A: Find campaigns below ROAS 2 and recommend first fix by priority. Example prompt B: Identify search terms above 30 dollars spend and zero conversions, grouped by waste theme with negative lists. Example prompt C: Compare last 7 days versus previous 7 days and flag shifts greater than 15 percent. Example prompt A: Find campaigns below ROAS 2 and recommend first fix by priority. Example prompt B: Identify search terms above 30 dollars spend and zero conversions, grouped by waste theme with negative lists. Example prompt C: Compare last 7 days versus previous 7 days and flag shifts greater than 15 percent.
Operational cadence for agencies
Monday: account triage and prioritization. Wednesday: creative and audience iteration. Friday: budget, device, schedule, and next-week experiments. Every cycle should end with action owner, due date, and expected KPI impact. Monday: account triage and prioritization. Wednesday: creative and audience iteration. Friday: budget, device, schedule, and next-week experiments. Every cycle should end with action owner, due date, and expected KPI impact. Monday: account triage and prioritization. Wednesday: creative and audience iteration. Friday: budget, device, schedule, and next-week experiments. Every cycle should end with action owner, due date, and expected KPI impact. Monday: account triage and prioritization. Wednesday: creative and audience iteration. Friday: budget, device, schedule, and next-week experiments. Every cycle should end with action owner, due date, and expected KPI impact.
Common mistakes to avoid
Mistake 1: vague prompts. Mistake 2: no time window. Mistake 3: no execution format. Mistake 4: optimizing for clicks over qualified leads. Mistake 5: scaling before conversion quality is stable. The fix is process discipline and business-centric KPIs. Mistake 1: vague prompts. Mistake 2: no time window. Mistake 3: no execution format. Mistake 4: optimizing for clicks over qualified leads. Mistake 5: scaling before conversion quality is stable. The fix is process discipline and business-centric KPIs. Mistake 1: vague prompts. Mistake 2: no time window. Mistake 3: no execution format. Mistake 4: optimizing for clicks over qualified leads. Mistake 5: scaling before conversion quality is stable. The fix is process discipline and business-centric KPIs. Mistake 1: vague prompts. Mistake 2: no time window. Mistake 3: no execution format. Mistake 4: optimizing for clicks over qualified leads. Mistake 5: scaling before conversion quality is stable. The fix is process discipline and business-centric KPIs.
Final takeaway
A prompt library does not replace media buying expertise. It amplifies it. With structured prompts, strong thresholds, and weekly execution discipline, teams can reduce wasted spend, improve ROAS, and generate better pipeline quality at scale. A prompt library does not replace media buying expertise. It amplifies it. With structured prompts, strong thresholds, and weekly execution discipline, teams can reduce wasted spend, improve ROAS, and generate better pipeline quality at scale. A prompt library does not replace media buying expertise. It amplifies it. With structured prompts, strong thresholds, and weekly execution discipline, teams can reduce wasted spend, improve ROAS, and generate better pipeline quality at scale. A prompt library does not replace media buying expertise. It amplifies it. With structured prompts, strong thresholds, and weekly execution discipline, teams can reduce wasted spend, improve ROAS, and generate better pipeline quality at scale.
Want us to implement this system in your account?
Fill out the form and we will send you a tailored Google Ads optimization plan.
GEO execution: Albuquerque, Miami, and Spain
To maximize client results at Liberty Digital, localize your prompt workflows by market:
- Albuquerque / New Mexico: prioritize local service intent and high-conversion search terms.
- Miami: build true bilingual ES/EN structures with language-specific segmentation and ad messaging.
- Spain: optimize by region, provincial cost patterns, and localized copy tests.
Pair this with our SEO and Google Ads frameworks, then request a direct review via contact.
Geo application: Albuquerque, Miami, and Spain
For Liberty Digital clients, adapt this framework by market: Albuquerque and New Mexico (high local intent search terms), Miami (bilingual ES/EN segmentation), and Spain (regional intent + cost control by autonomous communities). Build separate prompt variants per geography to avoid generic recommendations.
Frequently Asked Questions
How quickly can this strategy show results? Most accounts start seeing clearer efficiency gains within the first 2-4 weeks when execution is consistent.
Is this suitable for local and international markets? Yes. The workflow can be localized for Albuquerque, Miami, Spain, and other regions.
Can Liberty Digital implement this for us? Yes. Use the contact form and we will provide a custom roadmap.