Mortgage sourcing systems: broker guide

Sourcing is the part of the job clients think brokers do all day. In practice it occupies a small slice of the working week, and how well it goes depends less on the software than on what you feed it. This guide walks through a sourcing session as it actually happens, and where each system either helps or gets in the way.

Three things that get confused

Sourcing, criteria research and affordability are separate problems, and treating them as one causes most of the frustration brokers report.

Sourcing ranks products by cost. Give it a loan amount, a property value, a term and a rate type, and it produces a list ordered by true cost over a chosen period, fees included.

Criteria research answers whether a lender will consider the applicant at all. Zero-hours contract, recent default, a Tier 2 visa, a flat above a takeaway, a deposit gifted by a non-relative — none of that is a pricing question.

Affordability answers how much a lender will actually lend to this household. Two lenders on almost identical rates can differ by tens of thousands in maximum loan, which is why affordability often decides the case before cost does.

Some platforms cover all three. Many brokers use one tool for pricing and a different one for criteria, and that is a perfectly reasonable setup.

Working a real case in order

The efficient order is affordability, then criteria, then cost. Brokers who start with the cheapest rate and work backwards spend a lot of time discovering that the top three results were never available.

Start by establishing the maximum borrowing across a handful of plausible lenders. That gives you the real shape of the case and tells you immediately whether the client's target property is in reach.

Then filter on the awkward facts. Every case has at least one. Write it down before you touch the system, because it is easy to run a beautiful search that quietly ignores the thing that will sink the application.

Only then rank on cost, over a period that matches the client's likely behaviour rather than the default. A two-year fix compared over five years tells a different story from the same fix compared over two.

What the true cost figure hides

Sourcing outputs are calculations, not promises. They assume you fed in the right numbers and that the product data is current.

Watch for four things. Whether the arrangement fee is added to the loan or paid up front changes the ranking. Whether cashback is included in the calculation varies by configuration. Free valuation and free legals are worth different amounts to a remortgage and a purchase, and the system's assumed value may not match reality. And the reversion rate assumed after the fixed period is a modelling choice, not a fact.

None of this makes the tool wrong. It makes the output something you should be able to explain to a client in your own words, which is also what a file reviewer will expect.

Recording why, not just what

Under Consumer Duty the interesting question is not which product you selected but why it suited this client. A sourcing result printed and filed does not answer that.

Get into the habit of capturing, at the moment of decision, the filters you applied and the reason you moved past cheaper options. "Cheapest two required a minimum income the applicant does not meet; third was declined on the flat's cladding position" takes fifteen seconds to write and saves an hour reconstructing your thinking eight months later.

Many systems can save a search and attach it to the case record. Whether that saved search captures the exclusions as well as the results is worth testing before you rely on it.

Where these systems disappoint

Criteria data is compiled from lender guidance and lender guidance changes without notice. Any criteria search is a starting point for a conversation with a business development manager, not a guarantee. Experienced brokers treat a promising criteria hit as a reason to phone, not a reason to submit.

Product data feeds also lag. New products and withdrawals do not appear instantly, and during periods of rapid repricing the gap can be commercially significant. If the market is moving, check the lender's own page before quoting a rate to a client.

Specialist and complex cases are the weakest area across the board. Bridging, commercial, adverse credit at the harder end and unusual property types are still largely a matter of broker knowledge and packager relationships.

Choosing between systems

Twenty7Tec, IRESS and the sourcing modules built into broker platforms such as Acre are the names that come up most in the UK intermediary market. They differ in lender coverage, in how criteria and affordability are handled, and in what pushes through into your case file. Capability and pricing change, so verify both with the vendor rather than relying on any article, this one included.

Three practical questions cut through most demos. How many lenders are covered for affordability specifically, as opposed to product data? How often is criteria updated and by what process? When you select a product, what data flows into the case record and suitability document automatically, and what still gets retyped?

That last question is the one that determines whether the tool saves you time or just gives you a nicer screen to type into.

If you only change one habit

Log the cases where the system sent you the wrong way. A short running note of near-misses, mispriced results and criteria that turned out to be stale is worth more to your firm than any feature comparison, because it tells you where your particular case mix meets the limits of your particular tools.