Lily Ads Autopilot
Set the app, objective, budget boundaries, and review rules. Lily monitors the account and makes only the moves allowed by the program you approve.
Optimization goal
Paying-user ROAS
Write boundary
Program guardrails
Visibility
Action-by-action log
Continuous monitoring
Lily watches cost, revenue, paying users, search terms, bids, and campaign pacing on a recurring schedule. It separates routine maintenance from changes that deserve human attention.
Explicit limits
Each Autopilot program is scoped to an app and country set. Budgets, CPA goals, allowed operations, and review behavior remain visible instead of living in an opaque model prompt.
Revenue loop
Connect RevenueCat or AppsFlyer so the optimization loop can distinguish cheap installs from customers who actually pay. Without a revenue source, Lily can still optimize Apple Ads delivery metrics.
How it works
Authorize Lily as a collaborator and select the account and app the program may access.
Set markets, budget limits, CPA or ROAS targets, and the actions Lily may take.
See what changed, why the rule fired, and how the account responded over time.
Questions
No. The self-serve agent prepares work for you to approve in ChatGPT, Claude, or Lily. Autopilot runs the recurring actions explicitly allowed by a program's guardrails.
Program rules include explicit budget and bid boundaries. Actions outside those boundaries are not part of the permitted program and must be reviewed separately.
No, but connecting a revenue source lets Lily optimize toward paying users and ROAS instead of relying only on taps, installs, and cost metrics.