How to structure a Meta ads account when you're running multiple trades before scaling past $200 a day

Most multi-trade companies running Meta ads make the same structural mistake: one campaign, one budget, three services competing against each other for the same ad delivery. The algorithm doesn't know whether to show your window ad to a homeowner who searched "foggy glass replacement" or your roofing ad to someone who just had hail damage. So it guesses, splits delivery unevenly, and you end up with one trade eating most of the spend while the other two barely run. When the leads don't come in, the easy conclusion is that Meta doesn't work for your market. Usually the account structure is what broke it.

Before you spend past $200 a day, the structure needs to be right. Past that threshold, whatever's broken gets expensive fast.

Why one campaign for multiple trades is a data problem, not just an organization problem

Meta's delivery system learns which audiences convert. To do that, it needs signal: form fills, calls, booked appointments, whatever conversion event you're tracking. It needs roughly 50 conversion events per ad set per week to exit the learning phase and start optimizing reliably. That's not a soft guideline, it's how the algorithm works.

If you're running $200 a day total across three services in one campaign, you're spending roughly $67 per service. At a reasonable cost per lead for home services, that might get you five to ten leads per trade per week. That's not enough signal for any single ad set to learn. The campaign stays in learning indefinitely, delivery stays unpredictable, and you never get real data on which service actually converts at a profitable cost.

The fix isn't a bigger budget. It's separating the campaigns so each service builds its own signal pool independently.

The structure that works before you hit scale

One campaign per trade. Each campaign runs its own budget, its own audience, its own creative. That means three campaigns for windows, roofing, and painting, not one campaign with three ad sets fighting over the same spend.

Inside each campaign, keep it simple at first: one ad set, three to four ad creative variations. You're not testing twenty things at once. You're giving the algorithm enough creative variety to find what works within a consistent audience, while keeping the data clean enough to read.

On targeting: resist the urge to over-define the audience at this stage. Before $200 a day, broad targeting with strong creative almost always outperforms tightly defined interest stacks. Meta's system is good at finding people likely to convert if you give it enough room to look and a clear conversion signal to optimize toward. Painting and windows typically skew toward homeowners in specific income brackets and property ages, so location radius and demographic layers make sense. Roofing responds to event-based signals like recent weather patterns and can justify broader geographic targeting. These aren't the same audience, which is another reason they need separate campaigns.

One campaign budget, one audience logic, one performance story per trade. That's the structure.

What to track per trade before you decide anything is working or broken

Cost per lead is the starting number, but it's not the ending number. A $15 lead that turns into a phone call with someone who wants two windows replaced next month is worth something. A $9 lead that's a renter asking for a quote they'll never act on is worth nothing. At low daily spend, you need to track what's happening after the form fill, not just whether the form got filled.

The Safe Step case study is a useful benchmark: 247 leads, $11 cost per lead, $2,800 total spend in a concrete resurfacing trade. That's what a tight, well-structured campaign can produce for a niche home service. The number that matters is $11 CPL against a job ticket that likely runs several hundred dollars minimum. The math works comfortably. But you only know your math works if you're tracking the full path: impression, click, form fill, contact, appointment, job closed.

Most contractors who got burned by a prior agency never saw past the lead number. The agency reported leads, the contractor got calls that went nowhere, and the disconnect was never explained. Before scaling, build the habit of knowing your lead-to-appointment rate per trade. Painting leads that don't convert to estimates might mean the creative is attracting tire-kickers. Roofing leads that convert well on appointments but close slowly might mean the seasonality of your market needs to be factored into your budget decisions. Each trade tells a different story, and one combined campaign hides all of them.

What "ready to scale" actually looks like per trade

Scaling past $200 a day shouldn't happen until at least one trade has exited the learning phase and is producing leads at a cost that clears your margin. That means roughly 50 conversions tracked in a given ad set in the past seven days, a stable CPL over at least two weeks, and a lead-to-appointment rate you've confirmed manually by talking to the leads themselves.

When one trade hits that threshold, you scale that campaign. Not all three simultaneously. The campaigns that haven't stabilized yet stay at maintenance spend while you feed budget to what's proven out.

This is where most multi-trade operators get impatient. They see one service working, bump the total account budget, and let the algorithm decide how to redistribute. The algorithm doesn't know your margin per trade. It doesn't know your close rate per service. It optimizes for the conversion event you defined, which is usually a form fill, and it will route spend wherever form fills are cheapest regardless of whether those leads actually turn into jobs for you. Budget allocation by trade needs to stay a human decision, not a default.

If you're managing more than one trade, you're managing more than one business inside the same ad account. Treat it that way.

What to look for if you're evaluating someone to run this for you

If you're not running these campaigns yourself and you're considering bringing someone in after a previous agency that didn't deliver, the account structure question is a direct filter. Ask them: "How will you separate my campaigns by trade, and how will you track performance per service independently?" A competent answer is specific. A bad answer talks about "holistic strategy" or "optimizing the overall account."

The other question worth asking: "Will I have access to my own Meta Ads Manager?" If campaigns run inside your account, you can see every number whenever you want. If they run inside the agency's account, you lose the data when you leave and have no way to verify anything while you're in. That distinction closed or lost more than one deal in ASN's own sales calls, because contractors who'd been burned before knew exactly what it felt like to pay for a black box.

If you want to see how ASN structures multi-trade accounts specifically and what that looks like for your services, the right next step is to reach out at americanservicenetwork.com. No setup fee, no contract, and the campaigns run inside your account so you own the data either way.