Take a Quick Look at Arbor Financial Group
Arbor Financial Group is a mortgage lender with a question every lender asks and few can answer: which advertising campaign actually produced each funded loan (not just each form fill)?
The Challenge: Attribution That Dies at the Form
Like most lead-gen businesses, Arbor’s visibility ended the moment a lead was created. Ad platforms saw form submissions; the loan pipeline lived somewhere else entirely. The campaigns that generate cheap leads and the campaigns that generate funded loans are rarely the same ones. But without connecting the two, budget flows to whoever fills the top of the funnel.
The Build: A Closed Loop From Ad Click to Loan Pipeline
Working with the team at From the Future, I designed a full closed-loop attribution system:
- Capture: website scripts (via Google Tag Manager) capture both first-touch and last-touch campaign data (UTMs, referrer, landing page) plus click IDs for every major ad platform (Google, Meta, LinkedIn, Microsoft, TikTok) and the GA client ID.
- Store: that data rides along in hidden form fields and is written onto the lead record in the CRM at the moment of creation: 22 fields, mapped one-to-one.
- Report back: as the lead moves through the pipeline, the CRM fires webhooks on eight key funnel events, all the way through to funded loan, which flow through automation into the ad platforms as offline conversions and into reporting dashboards.
Because builds like this live or die in the handoff, the whole thing shipped as a developer implementation spec written for an engineering team with zero marketing context: exact field names, sample payloads, an acceptance checklist, and an explicit “what NOT to build” section.
The Result
Google and Meta stop optimizing toward “someone filled out a form” and start optimizing toward “someone funded a loan.” Every funded loan traces back to the campaign, ad, and click that started it, and ad budgets get judged on loans, not leads. The system is designed and now in implementation with Arbor’s development team.