Case Study 2: Automating Mortgage Document Collection and Application Preparation
What a mortgage brokerage could achieve by reducing document chasing
The opportunity
Mortgage brokers frequently spend significant amounts of time requesting, downloading, renaming, checking and chasing client documents.
Applicants may provide:
Incorrect statement periods.
Expired identification.
Missing payslips.
Partial screenshots.
Documents with unclear file names.
Incomplete proof of deposit.
Information spread across several email threads.
The issue is not usually a lack of client willingness.
Clients may simply be unsure about which documents apply to them or whether the information they have submitted is complete.
If documents are collected across email threads, shared folders, CRM records and separate systems, employees can spend considerable time checking what has been received, identifying what is missing and requesting information again.
Automating the document-collection and application-preparation journey could help a mortgage brokerage reduce administrative effort, improve the client experience and prepare more complete cases for broker review.
What could be improved?
A manual document-collection and onboarding process may involve:
A broker completing an initial conversation.
An administrator sending a standard document checklist.
The client emailing documents individually.
Employees downloading each document.
Renaming files manually.
Saving documents into the correct client folder.
Updating the CRM manually.
Checking whether all required documents have been received.
Identifying missing or incorrect information.
Sending further requests to the client.
Reviewing newly submitted documents.
Rechecking the entire case.
Extracting relevant financial figures.
Copying information into the CRM or sourcing system.
Confirming that the case is ready for broker review.
Common challenges include:
Generic checklists that do not reflect the applicant’s circumstances.
Documents arriving across multiple channels.
Information being stored in different locations.
Delays between document submission and review.
Repeated client chasing.
Duplicate documents.
Incorrect or incomplete information.
Poorly named files.
Missing pages or statement periods.
Brokers checking administrative work themselves.
Application preparation being delayed.
Clients being unsure whether their case is complete.
Increased administrative pressure as application volumes grow.
What could automation look like?
1. Applicant data assessment
Before automating document collection, the information already held about the applicant could be assessed.
The client record could be checked for information such as:
Number of applicants.
Employment status.
Income type.
Property purpose.
Deposit source.
Residency status.
Existing property ownership.
Credit circumstances.
Self-employed status.
Company-director status.
Existing mortgage commitments.
Information already held by the brokerage.
This would help determine which documents are actually required.
It would also reduce the risk of sending every applicant the same generic checklist.
2. Dynamic applicant questionnaire
Following the initial conversation, the applicant could receive a secure digital questionnaire.
The questions could establish:
Whether there is one applicant or multiple applicants.
Employment status.
Income type.
Property purpose.
Deposit source.
Residency status.
Existing property ownership.
Credit circumstances.
Whether the applicant is self-employed.
Whether the applicant is a company director.
The answers could automatically determine the document requirements for that particular case.
3. Personalised document requirements
Instead of sending every applicant the same checklist, the workflow could create requirements based on their circumstances.
For example, an employed first-time buyer might be asked for:
Proof of identity.
Proof of address.
Payslips.
Bank statements.
Deposit evidence.
Details of financial commitments.
A self-employed applicant might also be asked for:
Tax calculations.
Tax-year overviews.
Company accounts.
Business bank statements.
Accountant details.
Evidence of salary and dividends.
This could reduce confusion and prevent applicants from being asked for documents that are not relevant to them.
4. Secure document portal
Applicants could upload documents through a secure portal rather than sending them across multiple email threads.
The portal could show:
Documents required.
Documents uploaded.
Documents accepted.
Documents awaiting review.
Documents requiring replacement.
Items still outstanding.
This would give applicants and employees one shared view of the case.
It could also reduce the need for clients to email the brokerage asking whether documents have been received.
5. Automatic document classification
AI could identify the type of document that has been uploaded.
This could include:
Passport.
Driving licence.
Payslip.
Personal bank statement.
Business bank statement.
Mortgage statement.
Tax calculation.
Tax-year overview.
Company accounts.
Gifted-deposit declaration.
Proof of address.
The document could then be:
Classified.
Renamed.
Linked to the correct applicant.
Saved in the appropriate client folder.
Recorded against the document checklist.
This could significantly reduce manual document handling.
6. Automated quality checks
The workflow could perform initial checks before an employee reviews the document.
Checks could include whether:
The applicant’s name is visible.
Identification appears to be in date.
All required pages are present.
The correct statement period has been provided.
The document is readable.
Payslips cover the requested months.
The document appears to belong to the correct applicant.
Required sections are missing.
The same document has already been submitted.
The system would not determine whether a mortgage application should be approved.
Its role would be to identify whether the information required for human review appears to be present and complete.
7. Immediate client feedback
When an uploaded document does not meet the stated requirements, the applicant could receive immediate feedback.
For example:
Thank you for uploading your bank statement. We require the complete statement covering 1 April to 30 June. The document currently provided covers 1 May to 30 June.
Other automated requests could explain that:
Identification has expired.
A page is missing.
A payslip period is incorrect.
The document is unreadable.
Additional evidence is still required.
The wrong document type appears to have been uploaded.
This could prevent several days of avoidable delay between submission and correction.
8. Automated reminders and follow-up
The workflow could monitor outstanding items and automatically remind applicants when documents are still required.
A possible reminder schedule could include:
Initial document request.
First reminder after several days.
Second reminder if items remain outstanding.
Administrator task where no response is received.
Broker escalation where the case is becoming time-sensitive.
Employees would then focus on applicants requiring human assistance rather than manually chasing every incomplete case.
9. AI-assisted data extraction
AI could extract key information from approved documents.
This could include:
Gross income.
Net income.
Employer details.
Salary amounts.
Dividend amounts.
Account balances.
Mortgage payments.
Regular financial commitments.
Deposit amounts.
Business turnover.
Declared profits.
The extracted information could then be presented for validation by an administrator or broker.
This could reduce the amount of information being manually copied from documents into internal systems.
10. Broker-ready case summary
Once the required documents have been collected, the workflow could prepare a structured case summary.
This could include:
Applicant details.
Employment status.
Income.
Deposit source.
Financial commitments.
Documents supplied.
Documents outstanding.
Information extracted from documents.
Potential inconsistencies.
Questions requiring clarification.
Compliance steps still required.
The broker could then receive a notification that the case is ready for review.
This could create a more consistent handover between administration and advice.
11. Application preparation dashboard
Management and employees could view:
New applications.
Documents requested.
Documents received.
Documents outstanding.
Documents requiring replacement.
Cases awaiting client response.
Cases awaiting administrator review.
Cases ready for broker review.
Average document-collection time.
Number of client reminders.
Applications experiencing delays.
Records requiring manual intervention.
This could provide much clearer visibility of application progress across the brokerage.
What could be achieved?
Assume that a brokerage processes 35 mortgage applications each month.
1. Without a structured automated process
If document-related administration currently takes approximately three hours per case:
35 cases × 3 hours = 105 hours per month
2. With a structured automated process
If automation reduces employee involvement to approximately 70 minutes per case:
35 cases × 1.17 hours = approximately 41 hours per month
3. Potential monthly capacity released
105 hours − 41 hours = 64 hours
4. Potential annual capacity released
64 hours × 12 months = 768 hours
At an illustrative employment cost of £30 per hour:
768 hours × £30 = £23,040 of annual administrative capacity
This does not necessarily mean £23,040 of direct cash savings.
The value could instead be realised through:
Processing additional applications.
Improving response times.
Reducing broker administration.
Faster application preparation.
More proactive client communication.
Avoided recruitment.
Reduced pressure on existing administrators.
Greater capacity as application volumes grow.
4. Potential commercial value
The strongest benefit may extend beyond the direct value of the hours released.
If a brokerage has more administrative capacity, it may be able to process more applications without immediately increasing headcount.
For example, the released capacity could be redirected towards:
New client enquiries.
Application progression.
Client communication.
Lender follow-up.
Broker support.
Compliance activity.
Protection opportunities.
New-business generation.
If the improved process enabled the brokerage to complete only three additional mortgages each month, the potential revenue generated could exceed the direct value of the administrative time released.
Actual results would depend on factors including:
Application volumes.
Client responsiveness.
Existing systems.
Data quality.
Broker capacity.
Case complexity.
Conversion rates.
Average case revenue.
Before and after
1. Administration per case
Before: Document-related administration could require approximately 180 minutes for each application.
After: Automation could reduce the administrative requirement to approximately 70 minutes per case.
2. Monthly administration
Before: Processing 35 cases could require approximately 105 hours of document-related administration each month.
After: The same volume could potentially require approximately 41 hours of employee involvement.
3. Client checklist
Before: Applicants receive a generic checklist that may not reflect their employment status or circumstances.
After: The workflow generates a personalised checklist based on the applicant’s information.
4. Document collection
Before: Documents arrive across multiple email threads and may need to be downloaded and organised manually.
After: Applicants upload documents through a secure portal connected to the case.
4. Document naming
Before: Employees manually download, rename and file documents.
After: AI can classify, rename and route documents to the appropriate client record.
5. Quality checking
Before: Problems may only be identified when an employee manually reviews the document.
After: Initial quality checks can identify obvious issues immediately after upload.
6. Missing-item reminders
Before: Employees manually identify and chase outstanding documents.
After: The workflow monitors outstanding items and sends automated reminders at agreed intervals.
7. Client feedback
Before: Applicants may wait until an employee reviews their documents before discovering that something is incorrect.
After: Applicants can receive immediate feedback when a document appears incomplete or does not meet the stated requirements.
8. Data extraction
Before: Relevant figures are manually copied from documents into internal systems.
After: AI extracts key information for an administrator or broker to validate.
9. Case readiness
Before: Employees check emails, folders and CRM records to determine whether a case is complete.
After: The workflow shows whether the required information has been received and whether the case is ready for broker review.
10. Broker preparation
Before: Brokers may spend time reviewing administrative records to understand what has been supplied.
After: The workflow prepares a structured summary of applicant information, documents and outstanding questions.
11. Pipeline visibility
Before: Management may have limited visibility of where document-related delays are occurring.
After: A central dashboard shows outstanding documents, delayed cases and applications ready to progress.
Controls that should remain in place
A controlled document-collection and application-preparation workflow could include:
Secure applicant authentication.
Encryption.
Restricted access to client documents.
Role-based system access.
Human validation of extracted figures.
Data-retention controls.
Clear privacy information.
Audit logs.
Processes for correcting inaccurate information.
Controls around the use of applicant data.
Validation of automatically classified documents.
Human review of uncertain cases.
Regular testing of AI extraction accuracy.
Clear escalation procedures.
Human mortgage-advice and compliance review.
Automation should support application preparation.
It should not replace broker judgement, mortgage advice or required compliance review.
The potential business impact
Automating mortgage document collection could help a brokerage move from manually checking every upload to managing exceptions.
Applicants could receive faster feedback when information is missing or incorrect.
Administrators could spend less time downloading, renaming and chasing documents.
Brokers could receive more complete cases before beginning their review.
Management could gain greater visibility of cases that are delayed and why those delays are occurring.
The result could be:
Reduced administrative effort.
Faster document collection.
Fewer manual client chasers.
More complete applications.
Improved client communication.
Better broker preparation.
Greater visibility of case progress.
Increased application capacity.
Reduced pressure on support teams.
A more scalable mortgage onboarding process.
How much time is your brokerage losing to document chasing?
Neuranet helps mortgage brokerages assess their existing document-collection process, identify unnecessary manual activity and create automated onboarding and application-preparation journeys around their current systems.