# Marcus Delgado, Owner at Delgado Heating & Cooling — read of AfterHours, June 7, 2026

> 14 years running an 8-person HVAC shop in the Sacramento Valley. On Jobber, Google Business Profile, and a $90/month answering service that takes messages but won't book anything.

## How I got here

Googled "AI answering service HVAC after hours" on my phone at 7 AM Tuesday. Woke up to three voicemails from the night before and one was a no-call-no-show I probably could have locked down if someone had caught them in time. A Reddit thread mentioned AI voice agents for trades. This page came up in the results. Clicked expecting a product I could sign up for.

## What I clicked first

The pain scenario in the middle section stopped me cold. "A prospect calls at 9 PM and gets voicemail. Emails at midnight. Requests a callback. Visits your website at 6 AM to book a consultation. By the time you return the call at 9 AM, they've already scheduled with someone else." That is verbatim my Tuesday. I read the whole block twice.

## Where I paused

The stats row: "48% more qualified leads captured after hours" and "89% of captured prospects confirm appointment by next morning." I stopped and looked for a source. Nothing. No footnote, no "across X contractors," no survey citation. They just float there. The "40% of inbound leads come after business hours" number I've actually seen before, I think from a ServiceTitan report. But "89% confirm by morning" feels like it came from a brainstorm session, not a dataset.

## What I distrusted

Scrolled past the pricing block expecting a sign-up form and hit this at the bottom: "Honest disclosure: we don't have live customers on this idea yet. We shipped the strategy package; you ship the customer conversations."

I re-read it twice. The top half of this page reads like a SaaS I can start using tonight. "Start 14-Day Free Trial" button appears three times. "Most service businesses see their first after-hours booking within 24 hours of setup." And then I find out it is a product idea being sold to entrepreneurs for $5 to $99. A plumber who found this page through a search would feel tricked, even if the intent wasn't to trick anyone.

Also this: "We've trained our AI on 10,000+ calls from electricians, plumbers, HVAC, roofing." If there are no live customers, where did those 10,000 calls come from? That needs an answer somewhere on the page.

The navigation also has "Skeptic memos (22)" linked right there in the footer, which tells me this is a product factory pitching ideas to builders, not a product pitching to trades owners. The page is trying to serve two completely different readers at the same time and ends up confusing both.

## What would convince me

If this were a real product: one named contractor, one real city, before-and-after numbers that are specific enough to be verifiable. Not "48% more leads." Show me "Garcia Plumbing in Austin was getting 3 after-hours contacts a month, now gets 9, and 6 of them book same-night." One real number beats five round-sounding ones.

The $49/month price is actually credible. I would pay that if it worked. I just need to see it work on a real business before handing over my phone number.

## What I'd ask in an email reply

1. The "Start 14-Day Free Trial" button: is this a live product I can actually connect to my phone line, or is the trial a landing page that does not exist yet? Because I genuinely cannot tell from this page.

2. The 10,000+ training calls: where did those come from if there are no live customers? Were they licensed? Synthetic? That number is doing a lot of trust-building work and it has no backing.

3. Can I hear a sample call? Not a polished demo. A real 90-second recording of the agent handling a price question or a "how soon can you come out" objection. That one thing would move me faster than all the stats combined.

## Verdict: on-the-fence

The pain is described accurately and the price would not be the obstacle. But the page left me genuinely uncertain whether the product exists, and the stats that should have built credibility ended up making me more skeptical because they had no provenance. Fix the two-audience problem and put one real call recording up and this converts a lot better.

---
*Memo by skeptic persona, generated 2026-06-07. Studio breaks own self-grading loop.*