Why niche-matched proof closes more contractors than ROAS screenshots, and how to tell the difference before you hire

A 22x ROAS screenshot looks impressive until you ask what trade it came from. If the agency can't tell you the industry, the ad spend, the cost per lead, or whether the business was anything like yours, that number means nothing. It's decoration, not proof.

This is the gap most contractors miss when they're evaluating an agency for the second or third time. They've already been burned, so they ask for proof. The agency produces a screenshot. The number is big. The contractor signs. Then month one looks exactly like the last time.

The problem isn't that agencies lie about ROAS. The problem is that ROAS, on its own, tells you almost nothing about whether an agency can generate leads in your specific trade, at your price point, in your service area.

What ROAS actually measures (and what it doesn't)

Return on ad spend is a ratio: revenue attributed to ads divided by the amount spent on ads. A 10x ROAS means for every dollar spent, ten came back in attributed revenue. That's a real metric and it matters, but it only tells part of the story.

It doesn't tell you the trade. A 22x ROAS running LED lighting wholesale campaigns looks nothing like what it would take to generate roofing leads or auto detailing bookings. The audiences are different, the intent signals are different, the ad formats that convert are different.

It doesn't tell you the cost per lead. High ROAS can coexist with wildly variable CPL. An agency might achieve 10x ROAS by generating a small number of high-ticket conversions, or a large number of low-quality form fills that closed at a low rate. Both could produce the same ROAS number.

It doesn't tell you whether the results translated to actual booked jobs. Attribution in Meta's ecosystem counts a conversion when someone submits a form or takes an action the pixel is tracking. Whether that person actually showed up, paid, or was even reachable is a separate question the screenshot doesn't answer.

One real case from ASN's client work: Safe Step, a rubber resurfacing business, ran a campaign that produced 247 leads at $11 per lead on $2,800 in total ad spend. That's a number you can work backwards from. If a job is worth $500 and you close one in eight leads, your cost per acquired job is $88. That math holds or it doesn't based on your margins, and you can actually evaluate it. A ROAS ratio doesn't give you the same lever to pull.

Why generic proof actively backfires with burned contractors

On a real ASN sales call, a prospect named Sam Far was evaluating the service. The rep showed an LED lighting case study. Sam runs bottled water delivery. His response was direct: the case study wasn't relevant to his business, and showing it felt like the same pattern he'd seen from every other agency. It didn't just fail to convince him, it confirmed his suspicion that the agency was pitching a generic service dressed up as something specific.

That's the exact failure mode. A burned contractor isn't a neutral evaluator. They've already paid money, gotten nothing, and had to walk away from a contract they couldn't escape. When they're sitting across from another agency, they're not asking "will this work?" They're asking "how will this go wrong?" Generic proof answers that second question the wrong way.

Niche-matched proof does something different. When the epoxy flooring prospect Alifya saw the Safe Step case study, rubber resurfacing was close enough to her trade that the numbers felt applicable. She moved from skepticism to "I'll review the proposal." The numbers didn't change. The trade proximity did.

The mechanism here isn't complicated. A contractor who paints residential homes and hears about a 247-lead campaign for a surface restoration business can run the mental comparison: similar job size, similar customer type, similar ad format logic. A contractor who hears about a lighting wholesaler's ROAS cannot make that comparison, so they dismiss it entirely, and they're right to.

What to actually ask before you hire

If you're evaluating a Meta ads agency and you've been burned before, the useful questions are more specific than "do you have results?"

Ask for a case study from a business in your trade or an adjacent trade. Not a cousin business, not a "we've worked with service companies like yours" generalization. The closer the match, the more predictive the result.

Ask for the cost per lead, not just ROAS. CPL is a number you can stress-test against your own close rate and average job value. If an agency generated roofing leads at $18 CPL and you close one in five, your cost per job is $90. You know whether that works for you. ROAS doesn't give you that.

Ask what the ad spend was. A $11 CPL sounds different at $500 in spend versus $28,000 in spend. Budget scale changes audience saturation, ad fatigue, and targeting efficiency. Knowing the spend context tells you whether the result is replicable at your budget.

Ask what happened to the leads after they submitted. Did the client follow up? What was the contact rate? Did the leads actually turn into booked jobs or just form submissions? This is where a lot of agency results fall apart, because Meta can generate form fills at scale while the client's follow-up process loses half of them within the first hour.

Finally, ask what the agency will do if the results don't show up in month one. The answer to that question tells you more about the agency's confidence in their own work than the case studies do. An agency that locks you into a three-month contract before a single lead appears is structurally betting that you can't leave when the proof doesn't arrive. An agency that runs month-to-month is betting that the proof will.

What to do before your next agency decision

If you're not ready to hire again, that's a rational response to getting burned. But if you're going to evaluate, evaluate on the things that are actually predictive: trade proximity in the proof, real CPL numbers you can stress-test, and a contract structure that lets you leave if the results don't show up.

ASN runs Meta ads for home service contractors with no setup fee, no contract, and a 30-lead guarantee. If you want to see whether we have a case study close to your trade before committing to anything, the contact page is the right place to start.