The opinions expressed here are those of the authors. They do not necessarily reflect the views or positions of UK Finance or its members.

The research examines how lenders are using technology and data across the full mortgage journey, from origination through to completion, covering strategy, operations, risk management, and customer experience. It found a sector that knows what needs to change, but is increasingly struggling to deliver it.

2025 tested lenders on multiple fronts. Gross mortgage advances surpassed £80 billion in a single quarter, margins stayed tight, and the Financial Conduct Authority's (FCA) Mortgage Rule Review continued to shift the regulatory baseline. Technology investment decisions made two or three years ago are already being reassessed.

The divergence is not between lenders who want to invest and those who do not. It lies between those who can translate strategy into delivered change and those who cannot.

Where investment has gone, and what it has left behind

Investment has concentrated at the front of the mortgage journey. Scores for affordability and decision in principle consistently outperformed those for servicing and core banking. The result is a market that has modernised origination but left servicing behind. In the product transfer market, some lenders are managing post-contract variations through workarounds, re-running the full origination process as a patch for inadequate servicing infrastructure.

Steve Carruthers, Growth Director at Fignum, said: "The gap between origination and servicing investment has persisted because origination change is traditionally seen as more widely visible and more commercially urgent, while servicing transformation is structurally harder and carries perceived operational risk. By shifting their lens from front-end optimisation to end-to-end processing, lenders are beginning to treat servicing architecture as a strategic asset, not a legacy constraint."

Strategy is not the problem. Delivery is.

More than half of lenders consider themselves well-prepared. But preparedness correlates more strongly with architectural flexibility than with size. A large lender with fragmented data ownership may be less ready than a smaller one with clean integrations and a clear roadmap.

Lisa Neary, VP for Business Development at Firstsource, explained: "Execution strain shows up when lenders try to modernise on top of fragmented data, unclear ownership, and legacy integrations that can't adapt quickly enough to regulatory or operational change. Those who manage it well build disciplined sequencing and protect change capacity. The difference is not budget, but clarity, accountability, ownership, and focus."

Running transformation and business-as-usual in parallel is now a core organisational capability, and not every lender has built it.

AI is already here, just not where most expect

Fraud detection, document verification, and voice analytics are all in active use. What AI is not doing is making credit decisions. Under Consumer Duty obligations around explainability and fair outcomes, most lenders are not yet willing to cross that line.

Carruthers added: "Lenders should be investing now in accurate, complete, and timely data with clear governance frameworks, and controlled experimentation environments. AI maturity will favour those with structured data and clear accountability. The winners will not be those who wait longest, but those who prepare their foundations so adoption, when scaled, is deliberate, defensible, and commercially aligned."

The cost of standing still

Nearly three in ten firms report that less than a quarter of their regulatory reporting is automated. For a growing number of lenders, maintaining the status quo is the biggest risk.

Neary concluded: "Lenders move beyond workarounds when they focus on the foundations: stronger data, clearer ownership, and dependable core integrations. But the real accelerant is culture, teams that collaborate openly, challenge old assumptions, and treat change as shared responsibility."

The firms best placed for the next cycle are not the largest. They are the ones who understand their own data, control their integrations, and sequence change with discipline.