A GoHighLevel CRM, power dialer and buyer pipeline for a multi-state US cash home buyer

The company buys houses for cash from homeowners in difficult situations, then assigns or resells the contract to investor buyers. A remote team works each lead in a set order: cold callers, lead managers, home buying specialists, a disposition manager and the owners. Seller lists, website and ad leads, cold calls and investor buyers all need to land in one CRM, each tagged with the right state.

When we started, cold callers worked in a separate dialer, and the website lead flow had been repurposed as a makeshift dialer. The vendor's AI seller follow-up wasn't set up for each state. The CRM holds one company phone number and email, so a message to a seller in one state risked showing another state's details. The disposition team had several buyers lists for every contracted property and no clean way to call through them inside the CRM.

• AI seller sequences for each state. The CRM's AI Seller Prospector sends SMS and email to imported seller lists, replies to sellers and classifies each reply as hot, warm or cold. When it's unsure, the lead moves to Pending Review, AI messaging stops and the team is notified. We set up a short-term and a long-term sequence per state, with localised copy that names the state brand and sends from that state's own number.
• Fixed contact details instead of placeholders. Because the CRM holds one company phone and email, each state's workflow has its own phone, email and signature typed in, so a seller never sees another state's details. Workflows skip the email steps when skip tracing finds no email address.
• Same-day tasks for every new lead. Every new lead from a website form or a Google or Facebook ad creates a New Lead Followup task for the head lead manager, due the same day.
• Cold calling inside the CRM. We moved cold calling out of a separate dialer and onto GoHighLevel's power dialer, with one cold calling workflow per state, a smart list per state, call scripts on screen, a cold caller role with only the permissions callers need and two cold calling numbers whose callbacks ring the callers.
• A Cold Calling pipeline with six stages.

New Prospect ─► No Contact ─► Cold ─► Warm ─► Transfer ─► Booked Appointment

• A Buyers Outreach pipeline. Several buyers lists per contracted property are merged into one import, so the disposition team can call investors about a specific property. We fixed the import mapping: when to create or update an opportunity, which details belong to the contact and which to the opportunity, and how to handle buyers that are companies.
• Buyer smart lists and a "Buying in area" field. A multi-select field lets the team filter investors by city instead of piling up tags.

Weekly CSV per state ─► tagged import ─► smart list per state ─► power dialer ─► Cold Calling pipeline
Website or ad lead ─► "New Lead Followup" task, due same day ─► head lead manager shares it out
Contracted property ─► merged buyers list ─► Buyers Outreach pipeline ─► disposition calls

• One weekly import, fixed rules. The founder's own AI agent scrapes and skip traces lists and exports a CSV per state. We agreed the tags and field formats once and built them into its instructions, instead of maintaining per-lead webhooks or Zapier.
• The state is on every card. Smart tags show each opportunity's state, because the pipeline view can't show descriptions.
• Callers queue their own batch. They select a stage or a smart list, start the power dialer and dial one call after another. We tested it with sample numbers, walked the founder through it and showed the team in a cold calling training session.
• AI sequences are paused on purpose. They went live, then were paused in mid-September 2026 for two to three months, until the lead team is trained and the lists are tighter.