Prompts for auditing, planning, writing and reporting on Google Ads and Meta Ads. Every one assumes you'll check current platform rules and character limits before anything goes live, and keeps each platform's numbers separate. 10 prompts, free to copy.
Replace each {{VARIABLE}} with your own
details. Works with ChatGPT, Claude and Gemini.
ADS-001
Google Ads audit
Use when: Reviewing an existing Google Ads account.
Variables{{ACCOUNT_DATA}}
Audit this Google Ads account: {{ACCOUNT_DATA}}.
Cover structure, conversion tracking validity, search terms quality,
negative coverage, budget distribution, bidding strategy fit, ad relevance,
and landing page alignment. Give a prioritised fix list. Separate
platform-reported conversions from confirmed commercial outcomes if that
data is available.
Build the Google Ads strategy for {{BUSINESS}}. Objective: {{OBJECTIVE}}.
Budget: {{BUDGET}}. Market: {{MARKET}}.
Cover campaign architecture, keyword approach, conversion tracking
requirements, bidding strategy and the transition point to automated
bidding, and budget allocation. Currency matched to {{MARKET}}. State
whether the budget is sufficient to generate meaningful data in a month —
flag if not.
Build the campaign architecture for these services: {{SERVICES}}, across
{{GEOGRAPHY}}.
Group by intent and commercial value. Separate brand from non-brand.
Separate geography only where value genuinely differs. Keep budget control
at the level that matters commercially.
Use when: Finding and organising keywords for a campaign.
Variables{{SERVICES}}{{MARKET}}
Research keywords for {{SERVICES}} in {{MARKET}}.
Group by intent (commercial, informational, navigational). Note volume
with source and date where available — label as an estimate, not a
guarantee. Flag likely negative-keyword candidates from obvious mismatches.
Write responsive search ad copy for {{SERVICE}} in {{LOCATION}}.
Offer: {{OFFER}}
Landing page message: {{LANDING_PAGE_MESSAGE}}
Match search intent exactly. Reflect the landing page message — no
mismatch. Verify and respect current character limits before finalising.
No "best", "number one" or "guaranteed" without substantiation. Each
headline and description should add something distinct, not repeat the
others reworded.
Use when: Building a pre-launch negative keyword list.
Variables{{SERVICES}}{{EXCLUSIONS}}
Build a pre-launch negative keyword list for {{SERVICES}}.
Known exclusions needed: {{EXCLUSIONS}}
Include standard mismatches (jobs, free, DIY, courses, wrong locations,
competitor brand names where appropriate) plus the specific exclusions
given. Organise by negative match type.
Build the Meta Ads strategy for {{BUSINESS}}. Objective: {{OBJECTIVE}}.
Audience: {{AUDIENCE}}. Budget: {{BUDGET}}.
Cover audience approach (including exclusions), lead form versus landing
page recommendation with reasoning, creative testing plan, and attribution
caveats. Note that Meta's reported results will not match GA4 or a CRM —
these should be reported separately, never merged.
Write Meta ad copy for {{OFFER}}, audience {{AUDIENCE}}, at
{{AWARENESS_LEVEL}} awareness (cold/warm/hot).
Give primary text, headline, and a call-to-action recommendation. Hook
matched to a cold, interruption-based audience if awareness is cold. No
invented statistic or guarantee. Flag if this touches a restricted
advertising category (health, finance).
Use when: Briefing creative direction for a Meta campaign.
Variables{{OFFER}}{{BRAND_CONTEXT}}
Write the Meta creative brief for {{OFFER}}. Brand context: {{BRAND_CONTEXT}}.
Cover format recommendation (video/static/carousel) with reasoning, the
first-three-seconds hook direction, required visual proof elements
(genuine only), and mandatory brand or compliance inclusions.
Use when: Reporting on paid advertising performance.
Variables{{PERIOD}}{{PLATFORM_DATA}}
Build the advertising report for {{PERIOD}} using: {{PLATFORM_DATA}}.
One section per platform, never merged. Show spend, impressions, clicks,
CTR, CPC, conversions, conversion rate, cost per conversion where
available. Separate platform-reported conversions from confirmed
commercial outcomes. Interpret commercially, then give next actions.
Most accounts are either too fragmented to gather data or too consolidated to control. Here's how I structure accounts so they can grow without being rebuilt.