Marketing Attribution for Mortgage Brokers: Which Channels Actually Bring In Settled Loans

13 August 2026·5 min read

Quick answer: Attribution for a mortgage broker isn't about which ad got a click — it's about tracking a lead's source all the way through to a settled loan, months later, across a dozen touchpoints. Last-click attribution routinely credits Google Ads for a deal that was actually won by six months of Instagram content and a referral conversation. The fix is simple tagging discipline at intake, not expensive software. 📈

Most brokers can tell you their cost per lead down to the dollar and have almost no idea what their cost per settled loan actually is by channel — and that gap is where marketing budget quietly gets wasted. 💖 A lead is not a loan. Between enquiry and settlement there's pre-approval, property search, finance approval and settlement itself, often stretching three to six months, and whichever channel happens to touch the file last usually gets credit for the whole journey.

What most mortgage brokers get wrong

The most common mistake is judging channels purely on lead volume or cost per lead, without ever connecting that lead back to whether the loan actually settled. A channel producing cheap, plentiful leads that mostly fall over at pre-approval can look brilliant on a lead-cost report and be quietly the worst-performing channel in the business. The second mistake is last-click attribution: crediting whichever channel the client interacted with right before submitting an enquiry form, which systematically favours retargeting and branded search over the referral partner or the Instagram post that actually built the trust six weeks earlier.

The broker attribution loop

Track every lead through four stages in your CRM, tagged at the point of intake, not reconstructed later:

Source — how you'd actually describe it in conversation (Meta ads, agent referral, past client referral, organic Instagram, Google search), tagged manually for phone and referral leads, not just digital form fills.
Campaign / detail — the specific ad, post or referral partner, using a UTM convention you don't change every month.
Funnel stage — enquiry, pre-approval, unconditional approval, settled, fell over (with a one-line reason).
Settlement value and date — matched back to source and campaign once the loan actually settles.

Once a month, pull settled loan value by source for the trailing quarter and divide by spend for that channel — that's your real cost per settlement, not cost per lead.

A broker specialising in medico lending: Leads mostly come from LinkedIn content and referrals from accountants who service medical professionals. Because these referrals rarely fill out a web form, the broker manually tags every phone enquiry with the referring accountant's name at intake — without that step, this entire channel would show up as “direct” and appear to generate itself.

A first-home-buyer focused broker in Robina running Meta ads and content: Meta ads generate the most leads but the lowest settlement rate, because a meaningful share are early-stage browsers who aren't finance-ready for months. Tracking settlement, not just lead volume, showed organic content actually converted at a higher rate despite far fewer enquiries — a result invisible in a standard leads report.

A broker network relying heavily on real estate agent referrals plus a small paid search budget: Uses a simple first-touch-plus-last-touch model rather than picking one, since a client often first hears the broker's name from an agent, then searches the broker's name on Google weeks later — crediting only the last touch would erase the agent relationship that actually built the entire pipeline.

How to set this up without hiring a data analyst

Start with UTM parameters on every digital link you control, using a fixed naming convention you write down once and never improvise again. For everything that isn't a digital click — phone calls, referrals, walk-ins — add one required field at intake: “how did you hear about us,” captured as a specific source, not a checkbox for “other.” Then the discipline that actually matters: when a loan settles, someone updates that same record with the outcome, or the tagging just produces a detailed list of leads that never connects back to revenue. Please note: this is general marketing information, not credit or financial advice — check current ASIC/NCCP obligations before relying on it.

💡 Cost per settlement beats cost per lead, every time. A channel with expensive, low-volume leads but a high settlement rate can genuinely outperform a channel flooding your CRM with cheap leads that mostly stall at pre-approval. Reporting on lead volume alone rewards the wrong channel almost by design.

Mistakes to avoid

Tracking form-fill leads only, while referrals and phone calls default to “unknown” or “direct”
Reporting cost per lead as if it were cost per settled loan
Changing UTM naming conventions every campaign, breaking historical comparison
Relying purely on last-click, which systematically favours retargeting over the channel that actually built trust
No monthly cadence for reconciling settled loans back to source — the data decays fast if it's only reviewed annually

Frequently asked questions

Which attribution model should a broker actually use?

For a sales cycle this long, a simple first-touch-plus-last-touch view usually tells you more than any single model, because it captures both what started the relationship and what triggered the final enquiry. Perfect multi-touch attribution exists in theory more than in most brokerages' actual CRM setup.

Do we need expensive attribution software?

No — most CRMs used by brokers already have the fields needed for source, campaign and outcome tracking. The gap is almost always process (tagging consistently at intake and updating on settlement), not tooling.

How do we properly track referral partner sources?

Give each referral partner a distinct, trackable tag — a unique landing page, a specific UTM, or simply a required field at intake — and reconcile it against settled value regularly. Referral relationships are often a business's best-performing channel and the least measured one.

Can we ever track this perfectly?

No, and it's honest to say so. Some word-of-mouth and direct type-in traffic is genuinely unattributable no matter how good your tagging is. Aim for a reasonably accurate approximation of cost per settlement by channel, not a perfect model — a good-enough number you actually use beats a perfect one you never build.