The First 30 Days of a Meta Ads Campaign, Week by Week

A business owner launching a first Meta Ads campaign checks the phone every two hours for three days, gets discouraged, then receives two leads the same day and starts hoping again. That emotional swing is normal, it's simply the rhythm of a campaign that's just starting. The real problem isn't feeling it, it's making decisions at the wrong moment, reacting to three days of data as if they told the whole month's story.

Here is, concretely, what unfolds over the first 30 days of a Meta Ads lead generation campaign for a high-ticket service business: what gets set up, what can be read week by week, and above all what can reasonably be concluded, or not, at day 30. This is a French market case throughout, since it's where Solvya operates, but the sequence applies to any local high-ticket service.

Week 0: setup, before the first euro is spent

Before the campaign runs, several things need to be in place. Rushing this stage costs more later than taking two extra days upfront.

• The pixel or Conversions API needs to send the right events. A form that doesn't correctly pass the conversion event skews Meta's learning from day one, without anything visibly wrong showing up in the dashboard.
• The lead destination (native instant form or landing page) should be decided before launch, not tested mid-course in the first week. This choice has different implications for volume and quality, covered in Meta Instant Form vs Landing Page.
• The lead handling process must exist before the first lead arrives: who calls, within what timeframe, with what script, where the record gets logged. A lead that comes in on a Friday evening with nobody to call back before Monday is a half-lost lead.
• The daily budget is set at a level the account can realistically sustain through the learning phase, neither so small that it drags the phase out indefinitely, nor so high it can't be maintained for three weeks straight.

This week produces no leads. It conditions the quality of everything that follows.

Week 1: the campaign runs, the algorithm doesn't know anything yet

From activation, the campaign enters what Meta calls the learning phase: the system tests combinations of audiences, placements, and creatives to identify what produces the most conversion events at the best cost. We cover this mechanism in Meta Ads Learning Phase Explained Simply.

In practice, this first week often looks like this:

• Cost per result swings sharply from one day to the next, sometimes tripling.
• The first leads trickle in, in limited numbers, sometimes with a delay of several days after launch.
• Some creatives clearly stand out in engagement, without that saying anything yet about their conversion quality into actual clients.

The most common mistake this week is cutting a creative or an ad set after 48 hours because the cost per result looks high. At this stage, the algorithm hasn't gathered enough data to stabilize; intervening too early resets its learning phase, which extends the instability period instead of shortening it.

What can be done this week: verify tracking works, confirm leads are flowing correctly into the handling process, make sure nothing is technically broken. What shouldn't be done: judging profitability.

Week 2: gradually exiting the learning phase, first sorting

Toward the end of week two, most ad sets that have received enough conversion events exit the learning phase. Cost per result starts to stabilize, even if day-to-day variation remains.

This is when the first sorting of leads becomes meaningful. Not all leads received are equal, and this week allows a first distinction, based on calls already made, between:

• requests that match the target profile (budget, area, project type);
• requests driven by curiosity without real intent;
• requests that suggest a form setting issue (misunderstood question, inconsistent answer).

This sorting isn't just a quality exercise, it's optimization data. If the form or campaign only sends a conversion event back to Meta when a lead is genuinely qualified (via the Conversions API, for instance), the algorithm learns the right target profile faster. We explain this mechanism in Meta Conversions API for a Local Business.

At this point, two or three creatives usually stand out in lead volume. It's still too early to declare a definitive winner, but it's a reasonable time to start scaling down the budget of creatives producing almost no results, without cutting the whole campaign.

Week 3: the pace settles, the first real decisions

The third week starts to resemble the campaign's cruising rhythm. Lead volume becomes more predictable day to day, even if still subject to normal advertising fluctuations (day of the week, news events, seasonality in the industry).

This is usually the week where several concrete decisions get made:

• Cutting or sharply reducing creatives that have produced no usable result, provided they've had enough budget and airtime to be fairly judged.
• Reallocating budget toward combinations producing qualified requests, not just numerous ones. An ad set with a higher cost per lead but a higher qualification rate can be more profitable than a cheaper one that mostly fills the spreadsheet with curious browsers.
• Adjusting the form if week 2's sorting revealed a question causing confusion or letting through too many off-target profiles.

This is also when callback speed becomes its own topic if volume exceeds the usual handling capacity. A lead called back outside a short window loses much of its chance of picking up, a point developed in Why Your Facebook Leads Don't Answer the Phone.

Week 4: consolidate, don't conclude definitively

The last week of the first month is for consolidating the decisions made in week 3, not for making new drastic ones. A campaign that found its rhythm between weeks 2 and 3 needs stability for those optimizations to take full effect; every change to budget or creatives resets part of the learning process.

In practice, this week usually means:

• letting the working combinations run without daily interference;
• continuing systematic lead sorting to refine understanding of which profile converts best into appointments, then into clients;
• preparing, without launching yet, a second wave of creatives to test a new angle on what has worked best, as described in Meta Ads Creative Testing: The Method to Test and Scale.

What can be concluded at day 30, and what can't

At 30 days, there's usually enough data to draw limited but useful operational conclusions:

• which creatives and audiences produce the most qualified requests, not just the most leads;
• whether the daily budget is correctly sized to durably exit the learning phase;
• whether the internal lead handling process is keeping pace, or has become the real bottleneck.

What generally can't be concluded at 30 days, for a high-ticket service business: the actual return on investment of the campaign. The decision cycle for a client on a multi-thousand-euro project often extends well beyond the first month, from the initial request through the appointment, the quote, and the signature. Judging a campaign's profitability solely on its first month is like judging an entire sales funnel by its first step.

The right reading at day 30 isn't "it's working" or "it's not working," but a more nuanced status check: is the demand-generation mechanism functioning, is it producing a steady flow of profiles matching the target, and is the internal organization able to convert that flow into appointments? These three questions, not a premature ROI figure, determine whether the second month should look like the first or diverge from it.

At Solvya, this staged reading is part of how we support a campaign launch: looking at what the first 30 days actually say, without turning a month of setup into a final verdict.