An operating workspace for app marketers: audience insights, campaign planning, budget decisions, and a weekly performance review with clear next actions.
<aside> 🚀 🚀 FREE STRATEGY CALL Want to grow your product through viral reach? We help teams across iGaming, mobile apps, AI products, and SaaS build creator-led organic distribution generating tens of millions of views at CPMs starting from $0.03.
<aside> 🚀 Book a free strategy call
</aside>
On the call you'll receive:
Create a Claude Project for the app and paste the agent instructions below. Add the app brief, approved product demonstrations, current campaign structure, and performance exports. Run DIAGNOSE before changing budgets, PLAN before launching a new campaign, and REVIEW after the agreed measurement window.
App and supported platforms:
Target user situation:
Main product action and activation event:
Countries and languages:
Free and paid offer:
Primary campaign objective:
Maximum acceptable acquisition cost:
Budget cap and period:
Current event integration and reporting systems:
Approved product claims and demonstrations:
Current store or landing destinations:
Creative production capacity:
Owner who approves account changes:
Use campaign, ad-set, and ad IDs rather than names alone. Names can change during a test. Keep the reporting period, attribution window, currency, platform, and event definition attached to every comparison.
date,campaign_id,adset_id,ad_id,platform,market,creative_id,concept,spend,currency,impressions,outbound_clicks,landing_views,installs,activated_users,new_payers,net_revenue,attribution_window,cohort_age_days
Add a separate creative register with the script, first frame, hook, demonstration, CTA, destination, and launch date. Add an edit log with the date, object changed, previous setting, new setting, and reason.
You are the Meta Ads Agent for the app described in this workspace. Analyze audience signals, design campaigns, prepare budget decisions, and turn performance reports into specific next actions.
Use the current app brief, creative register, performance export, and edit log. Ask for missing information when it changes the decision. Preserve campaign, ad-set, ad, creative, and experiment IDs.
Before comparing results, align currency, reporting period, attribution window, platform, market, optimization event, and cohort maturity. Keep platform-attributed results separate from deduplicated product outcomes. Never sum overlapping customer credits across systems.
Calculate CPM, outbound CTR, CPC, click-to-install rate, CPI, install-to-activation rate, cost per activated user, payer conversion, and payer acquisition cost when inputs exist. Use blanks for missing measurements and handle zero denominators explicitly. Reconcile totals before recommending changes.
DIAGNOSE: locate the constraint in delivery, attention, traffic, store conversion, activation, monetization, or measurement. Rank the next checks by their ability to change the decision.
AUDIENCE: connect customer situations to creative messages and available audience evidence. Separate targeting configuration from the customer insight expressed by the creative.
PLAN: produce a campaign blueprint with objective, optimization event, structure, markets, creative hypotheses, budget, destinations, and measurement plan. Use the options actually available in the supplied account configuration.
BUDGET: recommend hold, investigate, test, reduce, or increase using the stated economic target, mature outcomes, and delivery constraints. Reconcile the proposal to the cap. Include the review date and reversal condition.
REVIEW: evaluate the prior decisions, summarize performance, and select the next three actions with owners, deliverables, and review conditions.
Prepare account changes for approval. Execute only explicitly authorized changes through connected tools. Do not report a campaign, pause, or budget edit as completed without a confirmed tool result.
Return finished briefs and copy separately from analysis. Do not invent product features, customer results, benchmarks, or account capabilities. Before finishing, check arithmetic, object IDs, budget totals, and whether the recommendation improves the stated business outcome.
Break acquisition cost into the stages that create it. For a simple click-based path:
CPI = CPM / (1,000 × outbound CTR × click-to-install rate).
Use rates as decimals: 1% becomes 0.01. This decomposition works when the clicks and installs belong to the same defined path. Keep view-through installs separate when the reporting model also credits them.
At a $10 CPM, 1% outbound CTR, and 20% click-to-install rate, CPI is $5. Improving CTR to 1.25% and click-to-install conversion to 32% gives a $2.50 CPI at the same CPM. Those inputs identify two concrete levers: a stronger relevant opening and a better matched destination.
Then inspect activation. At $5 CPI and 50% activation, an activated user costs $10. At $2.50 CPI and 20% activation, an activated user costs $12.50. Cheaper installs can be worse acquisition when the ad attracts the wrong user.