How a 100-agent team closed 291 deals from Google PPC leads by letting AI nurture contacts before agents ever touched them.
Jeff Cook's team in Charleston closed 291 transactions last year sourced entirely from Google pay-per-click leads.
Running roughly 100 agents across the Carolinas and around 1,000 total transactions annually, that number is notable. What it points to is more instructive.
The first version didn't work
When the team first layered Ylopo AI into their follow-up process, conversion wasn't improving the way they'd hoped. The problem turned out to be structural: agents were jumping into AI-managed conversations before the leads were ready for them.
It played out the same way repeatedly. Ylopo AI would engage a lead, the conversation would begin to warm, and an agent who spotted activity would step in.
A few texts later, the exchange would go quiet. The agent had done nothing wrong.
They had other leads to work and lives to get back to. But the consistency that AI maintains over days and weeks is something most agents can't replicate, particularly after hours or across a high-volume pipeline.
The team didn't change their script or their call cadence. They made a structural decision: once AI engaged a lead, that lead moved into a hidden pond.
Agents couldn't see it. They couldn't interact with it.
Ylopo AI continued nurturing until the lead indicated genuine readiness. When Jeff and Hunter looked back at their closings, they found the average time from registration to closing ran about 271 days.
Keeping agents out of the conversation during that stretch gave the AI the runway it needed.
That structural change was what separated their first attempt from what eventually became a 291-deal pipeline.
A tiered system built around lead age and agent eligibility
What they put in its place sorted leads into four distinct tiers, each with different access rules and different expectations for the agents working them.
The A pond holds leads from 0 to 5 days and rotates on a daily duty cycle. Agents who want access have to show up and earn it that day.
Twelve hours later they lose A pond access, regardless of how much they worked it. That structure tends to produce a different quality of effort than open, rolling access typically does.
The B pond covers 6 to 60 days and requires agents to hit minimum conversation counts or carry a certain number of pending transactions from the prior week. Access becomes something to sustain, not something that's assumed.
The C pond, from 60 to 365 days, is where new agents begin. No B pond, no A pond access until they've completed training and started producing results.
Newer agents work older leads while developing their skills, and given that average 271-day registration-to-close timeline, those C pond leads are often closer to transacting than agents initially expect.
The D pond, beyond 365 days with no contact, doesn't involve agents at all. Remarketing runs here. Property alerts fire.
The team's inactive AI drip works this group at roughly 1,000 drips per day across a database of around 500,000 contacts. When someone in D pond engages, they come back into A pond and get treated the same as a new registration.
When a lead engages with AI
The moment a D pond lead, or a B or C pond lead, engages with the AI, the system shifts. That engagement triggers a tag, and the lead moves into a holding pond that the agents working the standard ponds cannot see.
The team calls their AI Jordan. Once Jordan has nurtured a lead to the point where that person has indicated they want to speak with someone or set a callback time, a second tag fires: AI needs follow-up.
That tag routes the lead into a dedicated follow-up pond handled by a specialized group of agents who applied and trained for this specific handoff role.
The handoff follows a deliberate sequence. The agent steps into Jordan's voice inside Follow Up Boss and sends a final message as Jordan: a heads-up that a named agent will text in the next five minutes, along with that agent's number.
Then, from that exact number, the agent texts the lead directly. The lead already knows who is reaching out and why, which is a meaningfully different starting point than a cold call from an unfamiliar number.
If the lead goes quiet again after the handoff, the agent can step back into Jordan's voice to re-engage. One system, two personas, running inside the same CRM.
What the Jeff Cook model points to
291 transactions from a single lead source over 12 months is a result worth examining closely. But the more transferable takeaway is where their earlier process was breaking down and what the fix actually required.
For many teams using AI follow-up, the breakdown tends to happen not because the AI is underperforming but because human intervention interrupts the nurture cycle before a lead has reached a decision point. Teams that build workflow rules around that specific failure mode, keeping AI-engaged leads away from agents during active nurture, often report stronger conversion outcomes over time.
It's a less intuitive setup than leaving access open, but Jeff Cook's team found it made a material difference.
The pond structure itself, and the handoff mechanics, are not exclusive to a 100-agent operation. Smaller teams run similar versions with fewer tiers.
What scales down less easily is the operational discipline required to leave the system alone and trust that the nurture period is doing something, even when it doesn't feel like it.
Watch the full session
Jeff Cook and Hunter Cook walked through this entire system live at the 2026 Ylopo Success Summit in Charleston, South Carolina — the exact Follow Up Boss configuration, the pond eligibility rules, and the bridge message agents use when stepping in for Jordan.
Watch the session from the 2026 Ylopo Success Summit
Ylopo's inactive AI drip
The inactive drip Jeff Cook's team runs across 500,000 contacts is a Ylopo feature built specifically for leads that have aged out of active agent follow-up.
It re-engages dormant contacts through AI-driven text sequences and routes anyone who responds back into your active pipeline.
If you're sitting on a large lead database that isn't being actively worked, it's worth seeing what it surfaces.
Aaron Franklin
Head of Growth
Aaron "Kiwi" Franklin is the Head of Growth at Ylopo and a serial technologist and entrepreneur who has over 25 years of experience creating digital solutions for major brands and pioneering companies where technology and real estate meet. His depth of expertise stems from leading development of the first website for Apple to founding a global community of over 1,000 elite athletes.