Budgets and Campaign Launches

Setting an appropriate launch budget is about giving Meta's delivery system enough data to optimize, not about spending as little as possible to minimize risk. Each ad set needs to generate roughly 50 optimization events (e.g., leads) within a week to reliably exit the "learning phase" and reach stable, efficient performance — a budget too small to hit that threshold often looks like underperformance when it's actually just an underfed algorithm.

A practical rule of thumb when launching a new local-service campaign: budget for at least several multiples of your expected cost-per-lead per ad set per day, so the system has real room to find and show ads to converting users rather than being throttled by an artificially tight daily cap.

Launch day itself should be treated as the start of an observation period, not a "set and forget" moment. Check in daily (not multiple times a day, which risks making premature changes before the algorithm has stabilized) for the first week, watching for obvious red flags — a disapproved ad, an audience that's far too narrow, a landing destination that's broken — while resisting the urge to make major changes before you have a meaningful sample of data to judge from.