Ten Dead Leads Talking Again. One Application In 27 Days. Nothing Gets Forgotten.
Kate Matties is a loan officer at VanDyk Mortgage, a nationwide lender of roughly 700 employees, working the Nashville, Tennessee market. Her book is first-time homebuyers: VA, FHA, and down payment assistance loans. She is a President's Club producer, top ten in the company. The business runs on speed to lead and persistence. She buys Zillow leads at roughly $150 each, about $6,000 a month, and estimates a closed loan is worth $2,500 to $4,000 to her personally. Every lead that goes quiet is paid inventory expiring.
Discipline was not the issue. Her own rule was nine touchpoints per lead in the first week: three calls, three texts, three emails. All of it manual, run from memory, between closings. In her words, leads that did not respond “either go on a drip campaign or they get forgotten about.” The tooling made the rule impossible to keep: her company's CRM seat shipped without text messaging and without a mobile app, and 80% of her work happens on her phone.
- Roughly $6,000 a month on Zillow leads at about $150 each, with around three responding by her own count
- Nine manual touchpoints per lead in week one, then leads “get forgotten about”
- A corporate Total Expert seat with no SMS and no mobile app, for someone doing 80% of her work from a phone
- A Zapier attempt abandoned in an hour and a Twilio A2P registration rejected before the project began
An automated SMS follow-up engine that runs Kate's own messages, in her own voice, against every Zillow lead, without her touching it. A lead lands in her Zillow inbox and the first text goes out within minutes. Kate wrote every message herself, three variants of each; NEPTA built the machine that sends them. AI reads each lead email, extracts the name, number, and source, drops the contact into a pipeline Kate can open from her phone, and starts a timed sequence that tapers on schedule. The whole system lives on an Airtable hub Kate controls outright, with a two-way sync.
- First text within minutes of a Zillow lead arriving, phrased exactly the way Kate would say it
- AI extraction of name, number, and source from every lead email into a pipeline she runs from her phone
- Kate's copy, verbatim: three variants of every message, sent on her behalf
- Reply detection that halts the sequence instantly and forwards the response to her phone with the message it answers
- Compliance handled end to end: A2P campaign approved after her own application was rejected, quiet hours after 8pm Central, opt-outs honored on the first no
Twenty-seven days after the first call, the system was handed off. By that day the campaign, running against leads that had already ignored her calls, texts, and emails, had produced 10 to 12 positive responses, revived three active conversations, and put one completed loan application in her pipeline.
The forgetting step is gone. A lead now exits follow-up in exactly two ways: they answer, or they opt out. Her nine-touchpoint rule did not get faster. The part after it, where paid leads quietly died, no longer exists.
The economics are hers: mortgage applications take 30 to 90 days to close, and a single closed loan pays her $2,500 to $4,000. Her verdict at handoff: “If one of them turns into a closed loan, it's worth it.” By that same call she was already framing next year's budget around the system, saying the follow-up engine may prove more valuable than the Zillow spend itself, and forwarding NEPTA to loan officer colleagues inside VanDyk.
Want results like these?
Book a call and we'll map out an agent for your specific workflow.
BOOK A CALL