Your Funnel Is Lying to You

The marketing funnel is one of the most deeply embedded mental models in our industry. Awareness. Consideration. Decision. It's clean. It's logical. It fits perfectly on a slide deck. And it has almost nothing to do with how borrowers actually choose a mortgage lender in 2026.

I'm not saying the funnel is completely useless. It describes a general directionality — people do move from not knowing you to knowing you to choosing you. But the funnel implies a sequence that doesn't exist. It implies that awareness must precede consideration, that consideration must precede decision, and that each stage is distinct and measurable. None of that is true anymore.

How Borrowers Actually Choose a Lender

We tracked borrower behavior across our digital properties for six months and mapped out what the actual journey looked like. The results were humbling. The average borrower who closed a loan had 15-20 touchpoints across 4-6 channels over a period that ranged from 72 hours to 18 months. There was no consistent sequence. Some borrowers went from a Google search directly to an application — skipping "awareness" entirely. Others spent months reading content before ever engaging with a human.

• 38% of closed borrowers first interacted with us through a rate comparison, not branded content. They were in "decision" mode before they were ever in "awareness" mode.
• Borrowers routinely moved backward. Someone who started an application would abandon it, spend two weeks reading reviews and educational content, then return and complete it.
• AI search is compressing the journey. Borrowers using AI assistants to research mortgages arrive at our site with far more knowledge — and far more specific questions — than those who come through traditional search.
• Referral borrowers skip the funnel almost entirely. A recommendation from a real estate agent or friend collapses what might be weeks of consideration into a single phone call.

The Measurement Problem

Here's where the funnel really fails: it forces you to attribute outcomes to stages that may not have mattered. Your attribution model says a borrower came through a paid search ad (awareness), visited your rate page three times (consideration), and then applied (decision). Clean story. Except that borrower's real estate agent had already told them to call you. The paid ad was just how they found your phone number.

When we audited our attribution data against post-close surveys, the stated reason for choosing us matched the attributed channel less than 40% of the time. Nearly half the time, the thing our analytics said "caused" the conversion wasn't even in the borrower's top three reasons for choosing us.

A Better Framework: Signals, Not Stages

Instead of tracking where borrowers are in a funnel, we started tracking buying signals — behaviors that indicate intent regardless of where they appear in a theoretical sequence.

1. Research signals. Rate page visits, calculator usage, content downloads. These indicate active consideration but don't tell you how close someone is to a decision.
2. Engagement signals. Return visits, email opens, chat interactions, social engagement. These indicate sustained interest and are the strongest predictors of eventual conversion.
3. Intent signals. Application starts, pre-approval requests, direct contact. These indicate readiness to act and should trigger immediate, personalized outreach.
4. Advocacy signals. Review submissions, referral link usage, social sharing. These indicate borrowers who have already decided and are now influencing others — a stage the traditional funnel completely ignores.

The difference is subtle but transformative. Signals are observable and actionable in real time. Funnel stages are retrospective categories we impose on data after the fact. You can respond to a signal. You can only report on a stage.

What to Measure Instead

If the funnel is lying, what should you actually track? Here's what we shifted to.

• Signal velocity: How quickly is a borrower accumulating buying signals? A borrower who hits three intent signals in 48 hours is fundamentally different from one who accumulates them over six months, even if the total count is the same.
• Channel influence, not channel attribution: Instead of crediting one channel with a conversion, measure how each channel contributed to signal acceleration. This is messier but more honest.
• Time-to-engagement: How quickly do borrowers move from first signal to first human interaction? This matters more than which stage they're "in."
• Post-close signal generation: Are closed borrowers generating advocacy signals? This is your most undervalued marketing channel and the funnel has no way to account for it.

Letting Go of the Linear Story

The funnel persists because it tells a clean story. Executives like clean stories. Board decks need clean stories. But optimizing for a clean story means you're optimizing for your reporting, not your borrowers. The messy, nonlinear, channel-hopping reality of how people actually choose a lender is harder to visualize but far more useful to understand.

Your funnel isn't wrong in the way a broken clock is wrong. It's wrong in a more dangerous way — it's approximately right enough to be convincing, but specifically wrong enough to mislead your resource allocation. Every dollar you spend optimizing the top of a funnel that doesn't exist is a dollar you could have spent responding to a buying signal that does.