ResilientNiche Facebook ad: “I Lost a $40K Job Before I Ever Got to Compete.”

About this ad

Meta Ad Library ID

1393561846269932

Platforms

Facebook, Instagram, Audience Network, Messenger and Threads

Relaunches

2

EU reach

298

Niche

AI & Software

How we count

Ad text

I lost a $40,000 contract to a competitor I've beaten on quality for nine straight years. I found out how the client picked him three weeks later, sitting in my truck outside his office, and I still think about it every day. His name is Danny. We've bid against each other for almost a decade. I win more than I lose. Better crews, better materials, and honestly, better work, which I don't say to be arrogant, I say it because I have the photos and the reviews to back it up. Three weeks ago I lost a commercial buildout to him. Forty grand. Not the biggest job I've ever bid, but the kind that pays two guys' salaries for a quarter, and the kind I really wanted going into a slow season. I called the client afterward, which I don't usually do, because I wanted to know what I'd missed. Was it price. Was it timeline. Was it something in my proposal. She said something I did not expect. She said she'd asked ChatGPT which commercial contractor she should use in our area for a buildout like hers, and it recommended Danny's company by name, mentioned his licensing and a specific project he'd done that was similar to hers, and she called him that same afternoon. She never got to my proposal. She never needed to. The decision was basically made before I ever got in the room. I sat in my truck in his parking lot after that call for probably fifteen minutes. Not because I was scoping his place out, I want to be honest, I think I just needed somewhere to sit with it. Nine years of beating this guy on the work. And he beat me on a channel I didn't know existed, using information I didn't know AI even had access to. I went home and did the thing I should have done years earlier. Asked my own phone the same question she probably asked. "Best commercial contractor for Dallas office buildout." Three names came back. Danny's was one of them. Mine wasn't. I run a good business. Eighteen employees, a fifteen-year track record, more five-star reviews on Google than anyone else in my category in a fifty mile radius. None of that showed up. The AI didn't say I was bad. It just didn't say I existed. I spent about a month being angry about this in a completely unproductive way. Then I started trying to fix it, and every road I went down was a dead end. My web guy, who I'd paid for years, told me he'd "look into AI SEO" and never sent me anything concrete. I bought a visibility tool that scored me a 41 out of 100 and would not tell me what a 41 even meant, let alone how to move it. I got a call from an agency that wanted $2,400 a month and a full website rebuild, which would have been my fourth rebuild in six years, each one promised by someone as the thing that would finally fix it. None of it explained the actual mechanism. Why did the AI say Danny's name and not mine? Nobody could answer that. They could only sell me something. The answer came from a guy who isn't in this industry at all. Malik. He's a software engineer, did engineering work for companies like Spotify and Netflix, and when everyone else was guessing at how these AI models decide who to recommend, he actually went and read how the systems work. What he found changed how I think about my entire website. These models aren't handing your buyer ten options to scroll through anymore. They're picking one page to quote out loud, as an answer, the way a person would recommend a friend. That means the game isn't "rank for a keyword" anymore. It's "be the one page the AI would rather quote than anything else it found on this exact question." Most contractor websites, mine included, are built broad. Services page, about page, generic descriptions that could belong to any contractor in the country. Nothing narrow enough, nothing specific enough, for a model to have a real reason to pick us over Danny or anyone else. Malik proved this on his own site before selling it to anyone. A gluten-free baking blog he started from zero, no audience, no ad spend. It's been recommended by Microsoft Copilot more than 340,000 times since, with over 160 separate pages quoted by name, straight from his own Bing Webmaster data. He built the method into something called the AI Recommendation Playbook. No code, no developer, and no rebuild. You find the exact question your buyer is asking an AI right before they hire someone in your field, and you build one page around it that only you could have written, because it's built on your actual process and your actual numbers. I wrote my first page around the exact kind of buildout I lost to Danny. Took me most of one evening after the crew left, sitting in the office with the space heater on because I'm too cheap to fix the furnace timing. It's $27. One time. I'm not going to tell you I asked ChatGPT a week later and heard my name instead of Danny's. That's not how this works and I don't trust anyone who tells you it is. What I'll tell you is I finally understand the mechanism I lost that $40,000 job to, and I have a page live that's built specifically to be the one an AI would rather quote. There's a 60 day guarantee, no questions asked, directly from Malik. I still see Danny at the supply house sometimes. We're civil. I don't tell him about any of this. But I'm done losing jobs to a channel I don't understand. 👉 https://resilientniche.com/ai-citation-playbook-2

Where the ad sends people

resilientniche.com

I Lost a $40K Job Before I Ever Got to Compete.

The system I used to get one small site named more than 340,000 times by Microsoft Copilot, with no audience and no ads. $27. 60-day refund.