Skip to main content

Your mortgage website’s Apply Now button sends borrowers to a different domain. The application provider handles the next steps, while your analytics report records activity on the marketing site.

That handoff creates a measurement question: are you counting people who clicked the button, people who started an application, or people who completed one?

Those are different outcomes. A useful reporting setup names them clearly and measures only what the systems can verify. This guide focuses on the boundary between your website and an external application portal, including what to do when you cannot install tracking on the destination.

Map the Actual Borrower Journey First

Write down the path a visitor takes. Include the marketing page, the application link, any redirect, the external domain, and the point where an application is considered completed by the provider.

Check whether the process changes by loan officer, branch, device, or loan purpose. A homepage button and a profile-page button may lead to different destinations.

Record who controls each system and who can authorize configuration changes. Website administrator access does not establish permission to install analytics inside a third-party application platform.

A simple journey map helps identify what your team can observe directly and where provider cooperation is needed.

Define Three Different Events

Use event names and report labels that match the evidence available.

An application-link click means a visitor activated the outbound link. An application start requires evidence from the application process. A completed application requires a provider-defined completion signal or an appropriately reconciled operational record.

Do not label all three as “applications” in a dashboard. Doing so can make the website look more productive while obscuring what actually happened.

For example, 40 clicks and 12 verified submissions are not 52 applications. They describe two stages, and the counts may not represent a perfectly matched cohort without further reconciliation.

Define the reporting period and the meaning of completion before comparing results.

Measure the Outbound Click When That Is All You Control

If the external provider does not permit your tracking, measure the handoff on your own site and label it as an application-link click.

Use approved event parameters that describe context without identifying the borrower. Examples might include a generic page category or button placement. Review the implementation so it does not capture typed form values or sensitive query strings.

Do not rename the click as a successful application simply because there is no downstream measurement available. A limited but honest metric is more useful than a stronger claim unsupported by evidence.

LoanSites’ GA4 mortgage website guide covers broader event and traffic measurement. This handoff needs its own definition because the next activity happens outside the marketing site.

When Cross-Domain Measurement May Be Feasible

Google’s cross-domain setup documentation explains how compatible measurement can preserve identifiers across domains. The documented setup requires the same web-stream tag ID on participating pages and uses a linker parameter. Redirects or incompatible destinations can remove that parameter. Google also notes that enhanced-measurement outbound clicks are ignored for domains configured as cross-domain destinations.

For an application portal, technical capability is only part of the decision. The provider and your organization must permit the proposed setup, and the approved measurement must respect consent and data restrictions.

Listing the portal’s domain in GA4 settings alone does not establish end-to-end tracking.

Ask the Application Provider Specific Questions

Before asking someone to add a tag, request the provider’s supported approach. Useful questions include:

  • Can our organization use approved analytics on this portal?
  • Which pages or events are eligible for measurement?
  • Is an approved completion event, integration, or report available?
  • Do redirects preserve permitted linker parameters?
  • How does consent affect the setup?
  • Are there restrictions on shared infrastructure or lender-specific accounts?
  • What test environment or controlled test process is available?

Ask for documentation and an appropriate implementation contact. Avoid assuming that another lender’s configuration is supported for your account.

If the provider cannot support the requested setup, use a reporting method that reflects the available evidence.

Keep Borrower Information Out of Marketing Analytics

Mortgage application systems can contain information that does not belong in a general analytics platform.

Google’s PII guidance for Analytics prohibits sending information Google can recognize as personally identifiable, including names, email addresses, and Social Security numbers. Review both explicit event fields and automatically collected URLs or page information.

Do not place borrower names, loan application identifiers, financial details, or contact information into tracking links for convenience. Do not use an application completion URL containing sensitive information as an unreviewed analytics trigger.

If approved measurement is possible, use the minimum non-identifying event information needed. Have the technical, privacy, and provider owners review what is collected before deployment.

Test the Journey With Controlled Data

Use the provider’s test environment where available. Otherwise, agree on a controlled test process that avoids creating an unintended live credit application or triggering an unauthorized credit inquiry.

Check the approved events, destination, and report labels. Test the actual buttons borrowers use, including mobile navigation and loan-officer profiles.

Google’s developer guide to cross-domain measurement describes the linker mechanism. A technical reviewer can use that documentation to investigate the approved configuration rather than treating a button click as proof the entire journey was linked.

Also test what happens when optional analytics consent is not given. Report the resulting measurement limits instead of attempting to bypass them.

Check for Broken or Misleading Handoffs

The analytics can be configured correctly while the borrower experience still causes confusion.

Verify that the destination is the right lender or loan officer and that the button’s wording matches the action. Explain the transition in plain language where helpful, especially if the domain or branding changes.

Avoid collecting a detailed application on the marketing site only to require the borrower to enter the same information again without a clear reason.

If the external portal is unavailable, the marketing site should provide an appropriate contact path. Do not silently redirect borrowers to an unrelated form that creates a different expectation about applying.

LoanSites’ mortgage website service can support requested button, navigation, and content updates. Provider-side changes require coordination with the provider.

Reconcile With Operational Reports Carefully

An approved application-platform report can provide information the website cannot observe. Compare it with website handoff data while accounting for timing and definitions.

Someone may click this week and complete later, return on another device, or begin directly in the portal without a recorded marketing-site visit. Consent settings and blockers can also limit observed website activity.

Avoid forcing every operational application into a website channel when the evidence is missing. Keep unattributed or unmatched outcomes visible as limitations.

Where a CRM integration is involved, agree on what constitutes a new record, a duplicate, and a completed application. The goal is consistent reporting, not a count inflated by retries and repeated visits.

An Illustrative Reporting Example

A mortgage team records 80 application-link clicks during a month. The application platform reports 25 completed applications, but only some can be associated with the approved website journey.

The team should report those stages separately and explain the matching limits. Dividing 25 by 80 may produce a rough operational comparison, but it should not be presented as a precisely measured website conversion rate without a valid relationship between the groups.

This example is hypothetical. It shows why clear definitions matter before interpreting a percentage.

Frequently Asked Questions

Is an Apply Now click a completed application?

No. It shows the handoff action, not the outcome inside the portal.

Can cross-domain measurement work without provider cooperation?

Do not assume so. The destination must support the approved implementation, and your organization must have appropriate authorization.

Does excluding the portal as a referral create completion tracking?

No. Referral handling does not create an application event or prove the borrower completed the process.

Should every page of the application portal have a marketing tag?

No blanket instruction is appropriate. Only use explicitly approved measurement and data collection, with the provider and relevant internal owners involved.

Measure the Handoff Honestly

A useful setup distinguishes a website click from a provider-confirmed outcome, respects the data boundary, and explains what cannot be measured reliably.

Schedule a call with LoanSites to discuss your mortgage website’s application links and reporting needs. Bring the portal provider’s documentation so the proposed approach can fit the systems and permissions actually available.