The situation
An independent mortgage brokerage in Atlanta with three loan officers was averaging 12-14 applications per month. Inbound leads waited 2-6 hours for a first response. Past-client refis were happening — but mostly with competitors, because the team didn’t have a system to watch the database.
What got shipped
The snapshot went live on day 4 after purchase. The team prioritized three workflows from the 5 mortgage automations playbook:
- AI receptionist on first-touch SMS — every inbound web form and Facebook lead-ad inquiry started getting a 30-second AI response that qualified and booked.
- Calculator funnel on the homepage — the Affordability and FHA calculators replaced a generic third-party calculator the firm had been linking out to. Email-capture on both was wired into the pre-qual flow.
- Rate-drop alerts on past clients — the LO team imported their past-3-year closed-loan database (about 480 contacts) and turned on the rate-drop workflow.
Illustrative outcomes
By day 90:
- Monthly applications stabilized at 38-44 (up from 12-14).
- 1,840 calculator scenarios run on the site monthly, with an estimated email-capture conversion of ~6%.
- First refinance triggers fired on past clients within 14 days, producing 7 refi closings in the first 60 days.
- Realtor partner referrals grew from ~3/month to ~12/month after the co-branded flyer workflow went live.
What worked
The team’s read: AI receptionist on first-touch was the highest single-lever change. The lead-response time drop (from hours to seconds) compounded with calculator-driven inbound volume to produce the application growth.
The rate-drop workflow was a near-term revenue surprise. The team hadn’t been systematically watching past clients, and the first wave of refis represented loans that would have gone to competitors otherwise.
What we’d do differently
If we ran this engagement again, we’d start with the rate-drop workflow on day 1 of the snapshot install — it produces the fastest first-month revenue and creates internal credibility for the rest of the rollout. The team was initially focused on inbound-lead volume, which is correct longer-term but slower-paying than mining your own database.
Caveat
This is an illustrative scenario. Actual results depend on local market conditions, paid-traffic spend, team execution, and dozens of other variables. Mortgage origination is a regulated business and outcomes vary widely — your mileage will differ.
“We were doing 12-14 applications a month with three LOs scrambling. After the snapshot we crossed 40, and the team has time for actual conversations now instead of chasing replies.”