The Fact Find Doesn't Take an Evening Because It's Long. It Takes an Evening Because It's the Fourth Time.
One Australian broker's fact find runs to 335 input slots across four sheets. I mapped it field by field. About 60 percent of it is either sitting in a document the client already has or was said out loud in the meeting. Here is where the rest of the time goes, what is actually possible to take off a broker, and a three minute walkthrough of one file entered once producing both documents with the same $5,598.38 repayment on each.
What it looks like when one entry does the typing
Three minutes. One file entered once, then the funds to complete and the fact find, both filled, both agreeing on the repayment.
A mortgage broker's fact find does not take an evening because the document is long. It takes an evening because it is the third or fourth time the same client details have been typed that day: once in your notes during the meeting, again into the fact find, again into the funds to complete, again into the aggregator platform and the lender's portal. The document is the last place the information lands, not the first.
A fact find, or Fact Find Analysis, is the document an Australian mortgage broker completes for every loan application. It records the client's income, employment, assets, liabilities, living expenses, and their objectives and requirements for the loan. It is the evidence base for the credit assessment, and under Best Interests Duty the broker signs off on it personally and keeps it for seven years. Many brokers use the fact find built into their aggregator's CRM. Brokers writing higher volumes often build their own in Excel, which is the case this article measures.
Every broker who has built their own fact find knows the evening it costs. Below: what is actually inside the document, how much of it can be filled in for you, and the part that never should be.
- One Australian broker's fact find template carries 335 input slots across four sheets and 1,515 formulas
- The same broker maintains three versions of it for PAYG, self-employed and hybrid applicants, and those three are 96 percent identical
- Mapped across 140 field groups: 41 percent readable from a document the client already has, 19 percent said out loud in the meeting, 17 percent the broker's own judgement
- Published Australian estimates put file admin at 5 to 10 hours per loan, and brokers put the document build alone at 45 minutes to an hour when uninterrupted
- About one field in six is judgement. That is the ceiling, and it should be
The fact find is the document brokers complain about, but it is rarely the document's fault. By the time you open it, the information in it has already existed in three other places that day. What follows is what that costs, measured rather than estimated.
Why does the fact find take so long to fill in?
It is not a long document problem. It is a fourth-time problem.
Here is the sequence almost every broker who has built their own documents runs through, for every deal:
- In the meeting. You take notes while the client talks. Income, employment, liabilities, what they are trying to buy and why.
- That night. You open the fact find and type those same details in again, in a different order, into a different layout.
- Once a lender is chosen. The funds to complete gets the same names, the same figures, the same security address, typed a third time.
- Then the platforms. The opportunity in the aggregator, the contact record, the checklist to the client, and eventually the whole file keyed into the lender's portal by you or by someone you pay.
Nothing in that list is hard. All of it is slow, and all of it is done by the person in the business earning the most per hour. One broker I work with described a good day as four or five calls, and the cost of a good day as working until nine or ten at night to finish the paperwork behind it.
This is the part that matters, and it is not a paperwork complaint. The hours the documents take are the same hours you would use to write the next deal. That makes the fact find a ceiling on how many files you can carry, not an annoyance at the end of the day.
Published Australian estimates put broker admin at 5 to 10 hours per loan file from application to settlement, with messy deals running past 20 hours (CountHQ 2026; BrokerEngine 2026). Brokers themselves tend to put the document build alone at 45 minutes to an hour per file, uninterrupted. Uninterrupted is doing a lot of work in that sentence.
The client name, addresses and figures in this walkthrough are invented. No real client file appears in it. Full transcript below.
One sell-to-buy file, entered once across 33 fields. Thirteen of those feed both documents. One click produces the funds to complete: a $950,000 loan, a 65.52 percent LVR and a monthly repayment of $5,598.38. A second click produces the fact find, and the repayment on it reads $5,598.38, because it came from the same entry. Nothing is typed twice, so the two documents cannot drift apart.
What is actually inside a mortgage broker's fact find?
Measured, not estimated. One broker's template, parsed cell by cell.
Most writing about broker admin talks in adjectives. So I took one Australian broker's fact find templates and parsed every cell to find out what is really in there.
335 input slots. Spread across four sheets: the fact finder itself, a living expenses declaration, a year-to-date income annualisation sheet, and a long-run projection. Behind those slots sit 1,515 formulas, the large majority of them on the projection sheet alone. The copy used in the walkthrough above is trimmed to the three sheets the entry actually feeds.
That last number is the one worth sitting with. He keeps three versions of the document, one for PAYG applicants, one for self-employed, one for a hybrid couple. Compared cell by cell, the living expenses sheet is identical across all three. So is the income annualisation sheet. So is the projection sheet. Only the income block differs, and the hybrid version already contains both variants.
In other words, he built the general case years ago and has been maintaining two narrower copies of it alongside, by hand, ever since. He is not disorganised. This is what happens to a good document over ten years of Save As.
Your numbers will differ. But if you built your own fact find and you have been refining it for years, you are somewhere in this range, and you have probably never counted.
How much of a fact find can actually be filled in for you?
The honest answer has an upper limit, and the limit is the point.
I grouped those 335 slots into 140 field groups and sorted every one by where the answer actually comes from. Not by whether it would be clever to fill in, by where a human currently gets it.
| Source of the answer | Share | What that means |
|---|---|---|
| A document the client already has | 41% | Licence, payslips, notice of assessment, rates notice, statements. Readable |
| Said out loud in the meeting | 19% | Employment history, objectives, circumstances. Already in the conversation |
| Your judgement | 17% | Assessment calls only you can make. Should never be pre-filled |
| An external lookup | 17% | Duty calculators, registry fees, lender rate and fee sheets, servicing calculators |
| Calculated from other fields | 5% | Should never have been an input in the first place. Several currently are |
Figures rounded to the nearest percent.
The document and spoken groups together are about 60 percent of the fact find. That is the part that is worth taking off a broker, because it is pure transcription: the answer already exists somewhere, and a person is retyping it.
About one field in six is your judgement, and nothing should go near those. That is not a limitation to apologise for. A fact find that pre-fills your assessment calls is a fact find you cannot defend, and any broker holding a credit licence knows it.
The external lookups sit in the middle. Eleven distinct sources feed them, from state revenue office duty calculators to lender fee schedules, and each one is a separate piece of work. Some are worth connecting. Some are already handled perfectly well by the funding calculator inside your aggregator, and the honest answer there is to leave them alone rather than rebuild something that already works and is maintained for you.
Can a fact find and funds to complete be filled in from one entry?
Yes, and the part that matters is which document comes out.
Yes. Every detail on a file currently gets typed four or five times before lodgement: once in your notes, again into the fact find, again into the funds to complete, again into your aggregator, again into the lender's portal. This collapses the first three into one. You enter the file once and both documents come out filled, in your own template, in your own format, ready to check and sign. Not a summary. Not a report about the client. The documents you would have typed, already typed.
Thirteen of the fields on a typical sell-to-buy file feed both documents. Right now you type those twice, and the second typing is where the two documents quietly stop agreeing with each other. Entered once, they cannot drift. The walkthrough above shows exactly that: one entry, and the same $5,598.38 repayment printing on both.
Nothing is invented to fill a gap. Any required field still missing is flagged red before the document will generate, so a half-complete file announces itself rather than printing quietly. You fill those, review the rest, and sign.
If you want the longer version of what this looks like for a brokerage, that is on the mortgage brokers page, and the honest comparison against off-the-shelf software, no-code tools and hiring is on Why ChadScales.
If you are writing enough files for this to matter, the fastest way to find out what it would look like on your documents is half an hour with your fact find open.
Book half an hour with your fact find openDoesn't my aggregator's CRM already do this?
It does the standard parts well, and it will keep getting better at them.
Aggregator platforms are improving, and the sensible expectation is that they keep improving at the standard parts of a file. What they are not going to do is fill in the fact find you built yourself. They give you their fact find, in their layout, in their platform.
You can use those, and you will fill in their fact find their way in their portal. Which is the same reason you are not already using the one built into your aggregator.
That is not a criticism of any platform. It is the whole shape of the market. Every independent tool sold to Australian brokers replaces the broker's document with the tool's document. Collecting documents from clients is solved to death by half a dozen good products. Producing your own document, in your own format, is not solved anywhere, which is exactly why brokers who built their own fact find are still typing.
I tried something like this and it didn't stick. Why would this be different?
Most tools read the data fine. That was never the failure point.
The common experience is a tool that read everything correctly and then had nowhere useful to put it. It produced its own version of the file, so the broker checked the tool's output and then typed their own document anyway. Two jobs instead of one, and the tool quietly stopped being opened.
Reading is the easy half. Landing the result back in the document you actually send is the half nobody builds, because there is no off-the-shelf version of your fact find. That is not a gap in the market by accident. It is a gap because it can only be built around one broker's documents at a time.
I've got a processor. Doesn't that solve it?
It moves the typing. It does not remove it.
If you have a support broker or a processor, the same details are still being typed four times. They are just being typed by someone you pay, which means the cost has moved from your evenings to your payroll and it recurs every month whether the pipeline is full or quiet.
The size of it usually surprises people. One broker described the job of keying a completed fact find into the aggregator and the lender's portal, along with the compliance documents and signature pages, as taking his support person up to a full working day per application. He described his own process, thirty seconds earlier, as single entry. Most brokers do not count their staff's re-keying as re-keying.
When the file arrives complete and internally consistent, that day compresses into review and entry. The person stays. The typing goes.
Is a pre-filled fact find a problem under Best Interests Duty?
The document that comes out is a draft for review, never a finished file.
What comes out is a draft. Missing required fields are flagged before it generates, and everything else is yours to check. You review every field, you make every assessment call, and you sign. Nothing changes about who is responsible for the credit assessment, and nothing changes about the seven year record-keeping obligation under the NCCP.
You have stopped typing. You have not stopped reviewing. That distinction is the whole design, and it is why about one field in six is deliberately left alone. Your judgement is not a data entry problem and should never be treated as one.
Your files also stay yours. Every client record is exportable at any time, in whatever format you ask for, because a broker carrying a seven year obligation cannot afford to have records trapped anywhere.
The fact find is the loudest job. It isn't the only one.
The quiet ones cost more and get noticed less.
The fact find is what brokers name first, because it happens at nine at night and it is impossible to ignore. Three quieter jobs usually cost more:
- Chasing documents off clients. Every broker has the client who sends exactly one document a day and forgets the rest. Done properly, the chase runs from your own inbox in your own voice and nothing outstanding depends on anyone remembering it.
- Enquiries that went quiet. The referral that lands mid-appointment, sits until five, and books whoever answered first.
- Fixed rates about to roll off. A spreadsheet holds the book until it passes a hundred or so clients, and then the misses start. If you are not the one calling first, someone else is.
Most brokers need one or two of these, not all of them, and which one depends entirely on where your own week actually leaks.
Does this replace anything I'm already using?
No migration, nothing new to log into, nothing new to learn.
No. Your CRM keeps holding your data. Your spreadsheets keep doing their maths. Your aggregator keeps doing what it does. This sits on top and does the part in between, which right now is you.
Because it is built around documents that already exist, the setup runs in weeks rather than months. There is no data migration, because nothing moves. Your team opens one more page in the browser, fills it in once, and the documents they already use come out the other side.
Who this isn't for
Better said here than on a call.
The average active Australian broker lodges around 3.2 applications a month. At that volume, taking 40 minutes off a file returns about two hours a month, and two hours a month does not justify the cost of building anything. If that is you, keep your spreadsheet. It is doing its job.
This is worth doing if all three are true:
- You are writing roughly eight files a month or more, where the per-file hours compound into real weeks
- You have built your own fact find and funds to complete and you are not giving them up
- You have staff, or you are about to, and the same details are being re-keyed downstream by someone you pay
If you are happy using the fact find built into your aggregator, or what you actually want is a CRM, this is not that and there are good products that are. If you are earlier than that and just want the map of what can run without you, the free 38-page handbook for Australian brokers covers all five stages of a brokerage.
Frequently asked questions
What is a fact find in mortgage broking?
A fact find, or Fact Find Analysis, is the document an Australian mortgage broker completes for every loan application. It records the client's income, employment, assets, liabilities, living expenses, and their objectives and requirements for the loan. It is the evidence base for the credit assessment, and under Best Interests Duty the broker signs off on it personally and keeps it for seven years. Many brokers use the fact find built into their aggregator's CRM. Brokers writing higher volumes often build their own in Excel.
Why does the fact find take so long to fill in?
A fact find does not take an evening because the document is long. It takes an evening because it is the third or fourth time the same client details have been typed that day: once in your notes during the meeting, again into the fact find, again into the funds to complete, again into the aggregator platform and the lender's portal. The document is the last place the information lands, not the first.
How many fields are there in a mortgage broker fact find?
One Australian broker's fact find template, parsed cell by cell, carries 335 input slots across four sheets and 1,515 formulas. The same broker maintains three versions of it for PAYG, self-employed and hybrid applicants, and those three are 96 percent identical to each other.
How long does a mortgage broker fact find take to complete?
Published Australian estimates put total file admin at 5 to 10 hours per loan, application to settlement, with messy deals running past 20 hours. Brokers themselves put the document build alone at 45 minutes to an hour per file when uninterrupted, which is rarely how it happens.
How much of a fact find can actually be filled in for you?
Mapping one broker's fact find across 140 field groups: 41 percent can be read from a document the client already has, 19 percent is stated out loud in the meeting, 17 percent is the broker's own judgement, 17 percent needs an external lookup such as a duty calculator or lender fee schedule, and 5 percent should be calculated rather than typed. The 60 percent that is document or spoken is the part worth taking off a broker. The 17 percent that is judgement should never be touched.
What is a funds to complete worksheet?
The document showing what a client needs at settlement and where it comes from: purchase price, duty, registration and legal fees, deposit, sale proceeds, loan payouts, loan amount, and the resulting shortfall or surplus. Most of its figures are the same figures already sitting in the fact find, which is why the two documents are usually typed twice and quietly disagree.
What is the difference between a fact find and a funds to complete?
The fact find is the formal record of the client's situation: income, employment, assets, liabilities, living expenses and their objectives for the loan. It goes to the lender as the evidence base for the credit assessment. The funds to complete is the settlement day cash picture: purchase price, duty, registration and legal fees, deposit, sale proceeds, loan payouts, loan amount and the resulting shortfall or surplus. They share a large number of the same figures, which is why brokers type the same client details into both.
Can a fact find and funds to complete be filled in from one entry?
Yes. The file gets entered once and both documents come out filled, in the broker's own template and format, ready to check and sign. On a typical sell-to-buy file, thirteen of the fields feed both documents, and that second typing is where the two quietly stop agreeing with each other. Entered once, they cannot drift.
How long does it take to fill in a fact find and funds to complete from one entry?
Both documents are produced from a single click each once the file has been entered. In the walkthrough on this page, one sell-to-buy file is entered across 33 fields, and both the funds to complete and the fact find come out filled with a matching $5,598.38 monthly repayment. The time that remains is review, which does not go away and should not.
Doesn't my aggregator's CRM already fill in the fact find?
Aggregator platforms keep getting better at the standard parts of a file, and they will keep improving. What they will not do is fill in the fact find you built yourself. They give you their fact find, in their format, in their platform. That is the same reason most brokers who have refined their own document ignore the one already sitting in their aggregator.
Is a pre-filled fact find a problem under Best Interests Duty?
The document produced is a draft for review, never a finished file. Missing required fields are flagged before it will generate, and nothing is invented to fill a gap. The broker reviews every field, makes every judgement call and signs. Nothing changes about who is responsible for the credit assessment or about the seven year record-keeping obligation under the NCCP. The broker has stopped typing, not stopped reviewing.
Does a broker still need a loan processor if their documents fill themselves in?
Yes. The re-keying moves off the processor. Review, lodgement, lender liaison and condition follow-up all stay. The change is what the processor spends the day on, not whether you need one.
Is this worth it for a broker writing three or four files a month?
No. The average active Australian broker lodges roughly 3.2 applications a month, and at that volume the time returned does not justify the cost. This is worth doing for brokers writing around eight files a month or more, who have built their own fact find and funds to complete, and who have staff re-keying the same information downstream. More on how this works with brokers.
Bring Your Fact Find
Half an hour, your fact find open, and we go through where your hours actually go and what would come out the other side of your own documents. You will get a straight answer on whether it is worth doing at your volume, including if that answer is no. Nothing to prepare.