NurLead DeskStart the free week

Results · Noise

Most of what rings a contractor's phone is not a homeowner.

We sorted every number that called or texted the main line of a roofing and construction company in northern New Jersey, September 9 – October 14, 2026. Here is where each one ended up.

350
different numbers called or texted the line
290
were robo-dialers, sales pitches or landlines (83%)
1 of 94
numbers that texted the line first was a possible customer

Where the 350 went

149
Internet phone numbers of the kind auto-dialers buy in bulk

Checked against the carrier record the moment they called. From September 16 they got no text at all. Two of the 149 ever wrote anything: a sales pitch, and a caller from California.

94
Sales pitches: business loans, lead sellers, ad agencies

Recognized by what they wrote and left alone. A few got a one-line holding reply on days the desk could not read them yet; none was treated as a lead.

47
Landlines

A text cannot reach a landline, and in practice these were businesses and call centers. No text sent.

52
Got a text and never answered

Most of these came in the first week, before the bulk-number check was switched on.

8
Everyone else

Two homeowners with roof leaks, both qualified and handed to the office. One question about tile work. One existing customer, recognized from the company's records and greeted as a customer. One asked to stop, one asked about Halloween decorations, one was an internal test, one is still open.

Why this matters more than it looks

An owner who gets pinged 350 times in five weeks learns to ignore the pings, and then misses the two that mattered. The desk's job here was not to answer more. It was to answer the right two, fast, and keep the other 348 from ever looking like leads.

The texts were worse than the calls. Of the 94 numbers that opened a conversation by text, 93 were vendors or spam selling something. A few started out sounding like customers, asking whether this was the roofing company, then got to the pitch one message later. They were caught on the pitch, not on the greeting.

We checked the bulk-number rule against the record, because a filter that drops real homeowners is worse than no filter. Among the 149, none turned out to be a homeowner in the company's area.

Where the numbers come from. The desk's ledger keeps one entry per phone number with how it was classified and why; the log keeps every decision. Counts are unique numbers whose first contact fell between September 9 – October 14, 2026. A number is counted once, in the group it ended up in.