The lead business has a trust problem. We’d like to not have one.

Why this exists

If you originate mortgages, you have been sold bad data. Numbers that never existed. “Exclusive” records three other people were already calling. Files scraped from somewhere nobody wants to describe. And a compliance story that falls apart the first time you ask a specific question about it.

The economics push that way. Leads are sold on a promise that can’t be checked before purchase, so the vendor with the boldest claim wins the order. It’s an information problem, and it’s why the category’s reputation is what it is.

Vonta is built the other way round: publish the data dictionary, publish the provenance, publish where our responsibility ends and yours begins, and let buyers verify what they can before they spend. Everything on this site is browsable without an account — pricing, minimums, field lists, the selector. Nothing is gated behind a form.

Where the AI comes in

Most buyers over-index on one thing. They buy only exclusive because it feels safest, then run out of budget by the 14th. Or they buy the cheapest bulk pull and drown three seats in records nobody will ever work.

The lead selector exists to size that mix properly. It takes what you can originate, where you’re licensed, what you can spend, how many seats you have and how fast you dial, and returns a specific plan with the reasoning and the caveats attached. A deterministic scorer guarantees the plan fits your budget and clears every minimum; the model’s job is to improve it and argue for it in plain language.

It recommends segments. It never profiles the homeowners inside them. How the model is constrained.

Who we sell to

Licensed mortgage loan originators, brokers and lenders — verified through NMLS Consumer Access before the first delivery, with an active National DNC Registry subscription of their own. We turn people away over this. It costs us orders and it is not negotiable.