Mortgage lenders have spent the last few years plugging into every available external source, from Automated Valuation Models (AVMs) to payroll verification.
We have built the pipes to pull in more information than ever before but when you look closely at how that information is actually used to underwrite a loan, a frustrating reality emerges. We are hoarding data but we are still largely relying on manual intervention to make sense of it.
The Fignum Mortgage Tech Pulse 2026 report paints a clear picture of this disconnect. Despite all the investment in data acquisition, the actual process of credit decisioning remains stubbornly analogue. The majority of lenders admit that fewer than a quarter of their applications are assessed automatically. For specialist lenders and regional building societies, the reliance on human underwriters is often even heavier.
The problem is not a lack of data. The problem is that the data is not flowing to where it is needed, when it is needed. As an industry, we have tended to bolt new data sources onto legacy architecture, creating silos of intelligence.
Accessing external data earlier in the process to provide greater certainty around decisions is a goal we all share. Delivering it requires overcoming historic divisions within lending businesses. We still see situations where credit risk departments are reluctant to devolve ownership of key data points.
When the data is fragmented across different systems and controlled by different teams, the underwriter is forced to act as the human bridge, manually stitching the picture together.
This fragmentation also explains why the much discussed shift towards individualised pricing has stalled. In theory, having a richer, data-driven view of borrower risk should allow lenders to price loans on a highly granular basis.
In practice, our research found very few lenders are using data to support individual pricing. The funding mechanisms that underpin the market make anything other than tranche pricing difficult to execute. More importantly, the Consumer Duty requires us to evidence fairness and transparency at scale.
When pricing becomes dynamic and highly individualised, it becomes much harder to explain. For now, most lenders are sensibly keeping their pricing structures straightforward while focusing their data efforts on improving the speed and consistency of the decision itself.
So where should we be focusing our efforts? The answer lies in workflow orchestration.
We need to stop viewing data integration as an IT project and start viewing it as an operational imperative. The goal is not to remove the underwriter from the process but to remove the friction that slows the underwriter down.
By automating data collection, policy checking and case management, we can present the underwriter with a clean, comprehensive and verified case file.
The real value of data in the mortgage market is not in replacing human judgement. It is in equipping experienced professionals with the right intelligence at the right time, enabling them to make faster, more consistent and better-informed decisions. That is where the true efficiency gains lie and that is where the focus must remain.
