People
It was not too long ago that we witnessed the emergence of the digital-first challenger banks. These digital natives were free from the shackles of legacy technology and bolstered by robust investor backers, big ambitions, and a revolutionary mindset. They seemed poised to upend the competitive dynamics in financial services.
Many of us will also recall Atom Bank formally registering as a bank in 2015. It was one of the first to obtain a banking licence, rapidly followed by others, including Starling, Revolut, and N26. Was this the point at which banking was going to change forever?
Now, noting the significant investments across the industry in technology and generative AI, we take a fresh look at digital investments in banking and ask: How much should banks invest in digital and AI today, and for what purpose? How much is too much? And should the industry incumbents be concerned about their digital-native challengers in 2025?
1. Digital challenger banks: How much of a threat are they now?
Ten years ago, the gap in digital capability and competency between incumbents and challengers was staggering. It wasn’t only the latter’s technical accomplishments that stood out; it was also their focus, speed to market, and level of ambition, often supported by major PE and VC backers. At this time, the incumbents were bogged down by the restructuring of their legacy technology and unwelcome attention from the regulators.
Today, challenger banks hold more than 8% of current accounts and their share of the market is increasing year on year, though the pace of this rise has been slow, increasing by just 1%-2% annually.
Can we expect the challenger banks’ modest upward trajectory to radically change?
A recent study by Robeco Asset Management concluded that “only 23 of 453 digital challenger banks globally are operationally profitable”. Independently, a review of private capital funding has also revealed that funding has markedly decreased by 90% since its peak in 2021 (figure 1).
The size of the deposits held by the bank is also important, driving Net Interest Income (NII) and overall profitability. Deposits held by the challenger banks have risen – for example, Revolut’s deposits rose from $15bn in 2022 to $19bn in 2023, and Monzo from $4bn to $6bn over the same period. Yet challenger bank deposit levels overall remain well below those of the incumbents. For comparison, Lloyds Banking Group holds $1,201bn, Crédit Agricole holds more than $1,240bn, JP Morgan $2,401bn.
This will be influenced by their different business models and product ranges – full-service banks will naturally hold greater deposits – and it is important to acknowledge the post-COVID effect on bank deposits. Traditional banks were able to capture and maintain proportionately large deposits (and market share) over the past few years, due in part to their ability to offer high-interest savings accounts, enabled by the sharp post-pandemic rises in inflation that challenger banks were unable to match through their own savings and current account offerings.
In addition, trust remains an issue for challenger banks. A number have attracted unwelcome public and regulatory scrutiny; Starling was issued a £29m fine in 2024 from the FCA for failures in AML, while N26 has faced regulatory challenges and a fine amounting to €9.2m from BaFin for deficiencies in its own AML controls.
Branches, too, which have been in rapid decline (5,000 branch closures in the UK between 2015 and 2022, and 3,000 in France between 2015 and 2020, for example), may still have their place. Jamie Dimon, CEO of JP Morgan, is increasing the organisation’s branch estate as part of its growth strategy, placing clear value on their importance in serving communities and attracting new customers as a key component of an omni-channel approach.
Given the limited options available to challenger banks, their primary strategy must be one of driving profits through scalability. This, too, has proven difficult in the retail space historically, and many are exploring diversification of product offerings into higher-margin services as a result, such as wealth management, crypto, commodities/gold, investment trading (e.g. Nordnet), or foreign exchange (e.g: Wise and Revolut).
As it stands, the incumbents may still have the upper hand. Digitally native challenger banks face a much greater cost of customer acquisition, their business models are not proven, and investment has diminished dramatically. But they do remain a threat. As demographics shift in their favour, the gap will continue to close – incumbents must avoid complacency, especially now that some of the strongest threats have emerged from fellow competitors.
2. What are the banks investing in, and for what purpose?
At the macro level, we observe three phases of digital investment:
Phase 1 – “Protection”, pre-2016: The initial moves to digital banking, motivated by protecting market share, and keeping up with emerging threats to churn. This phase witnessed the creation of the first mobile apps and online banking capabilities. Investments in tech rose rapidly, and pay-offs in labour productivity were not observed.
Phase 2 – “Attack and grow”, 2016 to 2022: Increased competition driven by the challengers and digital disruption across incumbents drove innovation to grow revenues. Examples include new tools that allowed customers to monitor their financial “health” and contactless payments. While investments in tech continued to rise sharply, labour productivity remained consistent with the pre-digital age.
During this phase, incumbent banks made digitalisation gains, including strengthening data management, streamlining legacy platforms, migrating systems to the cloud, integrating AI and machine learning, and enhancing the customer experience. For example, JP Morgan decommissioned 2,500 legacy platforms between 2017 and 2023, Commonwealth Bank created a single mobile app by consolidating existing digital offerings into one while investing in integrated data and talent capability, and UBS decommissioned more than 39,000 legacy systems and 600 applications.
Phase 3 – “Efficiency”, 2022 onwards: As economic and inflationary pressures continue, alongside a lack of global stability, we see continued investment in technology but with much greater focus on trimming costs and driving tangible efficiency outcomes. GenAI and AI technologies are now past “peak hype”, and the incumbents are attempting to realise their potential within the mechanics of their banks.
GenAI and AI technologies promise the “holy trinity” – cost savings, revenue growth, and quality improvements. The narrative shift in AI from hype to real business results was reflected in executives' views in the AlixPartners’ 2024 Digital Disruption Survey: 22% told us they have raised their growth targets for this fiscal year because of AI-enabled products and features.
In this third phase, we expect to witness great improvements in labour productivity. However, as we’ll explore below, this is no easy task.
3. How much is enough?
Presently, huge sums are spent on technology. As expected, the bigger the bank, the bigger the spend. Citibank invested $9.1bn in IT during 2023, and JP Morgan spent $9.2bn. Estimated spend in Europe’s smaller banks is less, but not insignificant; for example, technology spend is thought to be $1.3bn for UniCredit and $2.3bn for BBVA.
The proportion of revenue spent on technology will depend on several important factors, including, but not limited to, the level of technology debt, client profile and demand, board appetite to invest, and the banks’ operating models. Based on our experience and analysis of the market, we see technology spending broadly range from 6% to 12% of revenue.
Global spending levels are forecast to increase further, with particular focus on cyber and GenAI investments. A survey by Forbes confirmed that “none of the executives surveyed had plans to cut their IT investments in 2025. In fact, most – 88% – expect to increase tech and IT spending by at least 10%”. In addition, Gartner suggests CAGR of 9% over the next five years.
While technology investment is essential (particularly in tech modernisation, leveraging AI, and protecting a business via cybersecurity), we observe many examples in the market of “investments with no outcomes”.
Technology transformation in a banking context is difficult. Though many claim expertise, this is an area where investments in technology routinely fail to meet their ROI. Many studies have confirmed that the majority of digital transformation projects fail to meet their objectives and that digital transformations intended to improve efficiency have an even higher rate of failure. Common causes of failure include cultural resistance, lack of controls (e.g., benefits cases), technical over-complexity, inappropriate program operating models, and over-ambition.
A study by Ayadi et al. in The International Review of Financial Analysis found that “excessive investment in IT technologies may be detrimental to both profit and cost-efficiency.” There is an inflection point, where efficiency gains stall or reverse.
Our experience confirms this. AlixPartners is frequently engaged to support with technology cost reduction, typically to support reinvestment in priority areas, and we regularly see excess spend ranging from 20% to 45% of overall technology spend.
This excess spend can often be related to personnel costs (design, engineering, and run teams, for example), and non-personnel costs (including infrastructure, application, hosting, and third-party spend). Many levers exist to reduce costs, ranging from stricter demand management, tighter cost control of third-party spend, and, of course, greater use of AI- and GenAI-related technologies to drive cost reduction, increased efficiency, and revenue growth.
To what extent does technology spend pay off?
As Figure 2 illustrates, the proportion of technology spending on R&D across all geographies is increasing year on year, and spending on R&D in North America is consistently higher than in Europe.
Comparing this to the Top 10 ranking of banks’ AI maturity in figure 3 below, we can see that seven of the top ten global banks ranked for their use of AI are North American and Canadian, with just two European banks (and one Australian). These investments in technology are reflective of stronger market technology leadership and a concerted push to attract the best talent, drive innovation, and build trust through transparency.
Investment in technology is rising and accounts for an ever-increasing proportion of revenue. This can pay off with increased revenue growth potential and greater brand recognition. However, the optimal investment level will vary according to many factors – and not all new spend is good spend. The market has an opportunity to reduce investment inefficiency with a sharp focus on priority technologies, backed by optimised cost cases and stronger implementation capabilities.
Concluding thoughts
For the incumbent banks, technology investment has entered a new phase. They must become leaner, faster, and more cost-efficient. A high level of waste and inefficiency remains common, and now is the time to shed it.
- Product and Technology teams in the incumbents include high degrees of waste and inefficiency that must be carefully identified and removed. Some of the most common causes of waste that we observe include low engineering productivity, sub-optimal infrastructure spend, a misaligned transformation portfolio, and inefficient operating models.
- Challenges with industrialising AI. There is widespread acceptance and engagement at the leadership level on the potential of AI-related technologies, and use cases are easy to identify, but there are few examples of these concepts being industrialised in a way that impacts the P&L. Operationalising AI technology in banking is not easy, but it can be achieved.
- A new mindset for governance. We observe interesting cross-fertilisation of talent between incumbents and challengers (for example, trading leadership excellence on agile productivity for experience of risk and compliance). However, cultural change has been too slow to engender. Building high-performing teams with speed and quality of execution, balanced with careful risk management and compliant activity, requires a bold vision, strong leadership, and a robust roadmap linked to tangible outcomes.
Please contact the authors if you would like to speak to AlixPartners about any of the themes explored in this article.
