Why $25/day to Instant Forms stalls out, and what actually fixes it

Plenty of roofing contractors have tried exactly this setup: a small daily budget, a handful of creatives, Meta's built-in Instant Forms. It feels like a real test. The ads go live, a few leads come in the first week, and then things slow down or stop entirely. The contractor concludes that Meta ads don't work for roofing, or that their market is too competitive, or that they need a bigger budget. Most of the time, none of those are the actual problem.

The problem is structural. The setup itself is built to stall.

Why $25/day can't do what you're asking it to do

Meta's ad algorithm needs data to optimize. Specifically, it needs conversion events, form submissions in this case, to learn which users are actually worth showing your ad to. The general threshold Meta's own system requires before it exits the "learning phase" is 50 conversion events within a 7-day window per ad set.

At $25/day with a roofing audience in a competitive metro like Houston, you might pay $40 to $80 per Instant Form submission if the targeting and creative aren't dialed in yet. That means, in a best case, you're generating somewhere between 2 and 4 leads per day. To hit 50 conversions in a week, you'd need the algorithm to be almost perfectly calibrated before it has enough data to calibrate itself. It's a circular problem. The budget is too small to exit learning, and the algorithm can't optimize until it exits learning.

Four creatives in a single campaign splits that budget further. If you're running four ads inside one ad set, the spend per creative drops to a level where even the best-performing one can't gather enough data fast enough. The algorithm rotates spend across them, nothing gets enough impressions to produce signal, and results look flat. The temptation is to turn off the "bad" creatives, but you often can't tell which ones are actually bad because none of them ran long enough at adequate spend to know.

What Instant Forms do and don't do

Instant Forms are fast to build and they reduce friction at submission. A user taps, Meta pre-fills their name and phone number, they hit submit. That sounds good. The problem is it sounds good for the user and the form, not necessarily for the contractor.

Pre-filled forms lower the barrier so much that people submit without really thinking about whether they want a roofing quote. They were scrolling, the ad appeared, they tapped twice, and now they're a "lead." In roofing specifically, where the job requires a physical inspection, a real commitment from the homeowner, and often insurance coordination, a frictionless form attracts a high volume of low-intent contacts. The contractor calls them back and gets voicemails, wrong numbers, or people who barely remember submitting anything.

This is exactly the complaint that came up on a real sales call with a painting contractor named Kurt Welch: he was getting "people looking for jobs instead of people looking for to-do work." The lead volume looked real. The leads weren't. Instant Forms optimized for submissions, not for buyers.

The fix isn't to abandon Meta entirely. It's to change what the form is optimizing for, and how follow-up is structured around it.

The follow-up problem that compounds everything

Even when an Instant Form generates a genuinely interested lead, that lead's intent has a short shelf life. They submitted the form while they were thinking about their roof. Ten minutes later they're thinking about something else. An hour later, if nobody has contacted them, the moment is essentially gone. They may still pick up when you call, but the warmth is different.

Most contractors, especially ones running their own campaigns without support, aren't structured to respond within the first few minutes. They're on a job site, or driving, or managing a crew. By the time they check their leads and call back, it's been two or three hours. That's not negligence, it's just reality. But from the lead's perspective, it feels like the contractor isn't interested, and the conversion rate drops accordingly.

This is the follow-up speed problem that Meta ads surface very quickly. It's not unique to roofing. It shows up in every trade where the owner is also the one doing the work. The solution ASN built into its model is Remi, an AI assistant that responds to a new lead within seconds, holds a real text conversation, handles basic objections, and books a call or site visit onto the contractor's calendar. The contractor shows up to the appointment, not to the cold-call sprint.

What a setup that can actually produce results looks like

The structural fixes are specific. On budget: the minimum to generate enough conversion data to teach Meta's algorithm is closer to $50 to $75 per day, especially in a competitive market like Houston roofing. That's the floor to exit learning in a reasonable timeframe, not the ideal spend.

On creative: four creatives isn't too many, but they need to run with enough budget per creative to actually learn. Testing more creatives at lower individual spend produces noise, not signal. Fewer creatives with real spend behind each one tells you something.

On the form itself: a landing page with a short application-style form, rather than a pre-filled Instant Form, filters lead intent before the contractor ever picks up the phone. Yes, fewer people submit. That's the point. The ones who do submit are meaningfully more likely to want the service.

On follow-up: the contractor who responds within five minutes of a form submission will close a materially higher percentage than one who responds in three hours. This is not an opinion. It is the direct implication of every sales conversation where lead quality complaints turned out to be lead response complaints in disguise.

The Safe Step rubber resurfacing campaign ASN ran produced 247 leads at $11 cost-per-lead on $2,800 in total spend. That result came from the combination of these elements working together, not from any one of them in isolation. The ad generates the lead. The follow-up converts it.

What to actually do next

If you've run a Meta campaign that stalled, the question to ask your next agency isn't "do you guarantee leads." It's "what's your minimum viable daily budget for my market, how do you structure your ad sets to exit the learning phase, and what happens to a lead in the first five minutes after they submit."

If they can answer all three specifically, you're having a different conversation than the ones that burned you before. If they can't, the campaign structure will stall for the same reasons the last one did, regardless of what the contract promises.

ASN works with contractors who've been down this road before. No setup fee, no contract, and the follow-up problem is handled before the first lead comes in. If you want to see how the model works for your trade and market, the contact page is the right place to start.