Technology has reshaped banking. Traditional financial institutions have grown their technology teams and are demanding more from them than ever before. Technology teams build innovative new products such as AI-enabled personalized customer experiences, digital payments, and analytics-driven cross-selling. They integrate acquisitions, automate back-office processes, and securely manage vast amounts of data. A look through a bank’s strategy reveals that most of the initiatives require extensive technology delivery.
Yet many banks lack the resources to fully invest in technology-enabled innovation. Budgets are constrained. As the IT estate grows older and more complex, more spend is needed just to “keep the lights on.” Most institutions have already applied the conventional levers for creating headroom in technology budgets, including offshoring and vendor renegotiation, and are looking for a new tool kit to free up resources.
To boost tech innovation, a few leading banks have focused less on cutting costs and more on increasing the productivity of their engineering teams. Taking a cue from digital-native software players, they are maximizing their portion of the technology workforce engaged in writing code, and then helping those engineers work as efficiently as possible. While this developer-first approach is common in the software industry, it’s not yet widespread among financial institutions. However, banks that have adopted the practice are seeing measurable impacts. As a result, the best-performing banks can achieve fully 50 percent more technology capacity—defined as the number of team hours available to create tech-enabled innovations—than average banks for the same budget. They also ship new features to customers faster and attract higher-caliber engineers, who bring better practices that lead to even greater speed and productivity, creating a positive reinforcing loop.
This article outlines the value of that productivity-centric approach, the internal levers that drive the greatest impact, and the actions leadership teams can take to get started.
Value at stake: Get more from your technology investments
For many institutions, spending on technology—including talent, hardware, software, and services—is their biggest cost base, yet it is also a black box. While they are clear on how much they spend and whether projects deliver on time, they have limited understanding of the productivity of their teams, or whether it would be worthwhile to invest in making them more efficient.
Over the past few years, leading banks have implemented scorecards to measure the speed and productivity of engineering teams by analyzing data from the tools the teams use (Exhibit 1).
When companies quantify productivity, they pinpoint areas where engineering speed and accuracy can be improved, as well as identify meaningful ways to foster creativity. By helping their engineering teams work smarter, leading institutions significantly increase their technology capacity compared with average institutions, with no increase in budget (Exhibit 2).
To see how this works, let’s compare a hypothetical average bank with a leading bank. At the average bank, half of the technology employees create software, and those employees spend half of their time on the core activities of writing, testing, and maintaining code. In other words, only 25 percent of the average bank’s total technology capacity is devoted to writing software. The other 75 percent involves teams doing orchestration, analysis, and control processes that are important but could be better automated or streamlined. In contrast, at the leading bank, 70 percent of the technology employees write or maintain code, and those employees spend 55 percent of their time on these core activities. As a result, the best-performing bank has 39 percent of its technology capacity devoted to software development—around 50 percent more than the average bank.
High-performing engineering teams are not only more productive, but they also ship new features to customers faster and deliver higher-quality products with fewer bugs. They build products more aligned to business and customer needs that, in turn, deliver the outcomes banks want: better user experiences, more new customers, better retention, and income growth.
Four moves to transform engineering productivity
Measuring engineering productivity, rather than just costs, tends to change management attitudes toward technology spend. Management teams start to ask what they can do to measurably improve productivity, evaluate the ROI of different actions they could take, and explore how to create a positive loop in which a better environment attracts higher-caliber engineers.
To start boosting tech team productivity, bank leaders can take four steps (Exhibit 3). In our experience, banks that implement at least one or some combination of these four strategies typically achieve 20 to 30 percent productivity improvements in 18 to 24 months. (Closing the full 50 percent gap between average and best performance is a multihorizon journey that takes a more coordinated approach.)
1. Streamline the software development life cycle for engineering excellence
Executives are sometimes surprised by the level of manual effort required when their technology teams write software. At one major bank, for example, a developer writing software had to start by submitting requests to several other teams to get an environment set up. Once they finished coding, they then needed to do extensive manual testing, since the automated tests covered only a small portion of cases. Further integration and security testing needed to be done by other teams only available on certain weeks. Deploying the software required approvals by multiple individuals and committees. Once deployed, the application was then transferred to an IT operations team, requiring an extensive handover. Each team controlled only a small part of the process. Yet instead of automating and streamlining these processes, the bank hired extra project managers and test engineers to manage the workload—adding cost and complexity.
One leading financial institution avoided this type of pitfall by conducting an “X-ray” of each of its engineering teams to understand existing practices and performance. This data-based approach allowed management to identify the cross-cutting initiatives that would deliver the biggest productivity increases and greatest ROI. The company overhauled its engineering tools, automated its security controls, and created a coaching team to train frontline engineers in the new approach. Once the centralized tools were built, engineering teams adopted them in structured three-month waves over 12 months.
As the bank implemented this transformation, leaders also established a culture of continuous improvement. Similar to lean manufacturing operations, the bank’s technology teams began regularly reviewing their own productivity data and asking themselves what they could do to become faster and more productive. Ultimately, this organization created 30 percent more capacity in its existing technology team, without any increase in budget.
2. Deploy generative AI along the whole product life cycle
Generative AI-enabled software development tools have rightly received a lot of attention, as they save valuable time for engineering teams, above and beyond the automation outlined above. Many organizations are running pilots of generative AI (gen AI) tools that enable their teams to code faster than before. (And in companies that aren’t running pilots, many engineering teams are experimenting with these tools anyway.) But while these gen AI tools are often successful in pilots, there are several challenges in scaling them safely and effectively. First, banks tend to treat these pilots as a tooling effort and underestimate the change management required. They see strong adoption from the first 100 enthusiastic users but more limited uptake after that. Second, many of the tools on the market are point solutions. For example, they might aid code generation, but coding is only a small part of what engineers do. Third, many of the tools are not trained on a company’s code base, so they generate code that does not meet organizational standards.
Some leading institutions have approached the rollout of gen AI tools differently. They provide gen AI tools to everyone involved in the software development process, not just to coders. Product owners and managers, data analysts, user experience designers, and others get access to the gen AI tools they need to be more efficient. For example, product owners use gen AI tools that can automatically generate the stories and architecture diagrams for a new feature within seconds—an activity that otherwise would take days. These banks invest heavily in change management, recognizing that gen AI is a fundamental shift. They do not just roll out gen AI tools and hope teams adopt them, but instead invest significant time and resources in establishing organizational practices to ensure these tools are integrated into everyday workflows. They fine-tune the gen AI tools on their own code base and use different large language models (LLMs) for different tasks. For example, they might use one LLM for new code generation but a different one that performs better for documentation. Institutions that deploy gen AI throughout the whole product life cycle can see a 20 to 30 percent uplift in development team productivity.
3. Integrate tech and business teams
Banks that want to increase technology productivity typically must change how engineering and business teams work together. Getting from an idea for a new customer feature to the start of coding has historically taken three to six months. First, business and product teams write a business case, secure funding, get leadership buy-in, and write requirements. Most engineers are fast at producing code once the requirements are clear, but when they must wait six months before they even write the first line, productivity stalls.
Taking a page from digital-native companies, a number of top-performing banks have created joint teams of product managers and engineers (and sometimes including operations, too). Each integrated team operates as a mini-business, with product managers functioning as mini-CEOs who help their teams work together toward quarterly objectives and key results (OKRs). With everyone collaborating in this manner, there is less need for time-consuming handoff tasks such as creating formal requirements and change requests. This way of working also unlocks greater product development speed and enables much greater responsiveness to customer needs. While most financial institutions already manage their digital and mobile teams in this product-centric way, many still use a traditional project-centric approach for the majority of their teams.1
One large bank moved to a product-centric model and made far-reaching changes in the way thousands of its people worked, led by the CEO, COO, and CIO. After a short pilot, the bank reorganized into a new set of integrated teams with new roles and shared incentives. Leaders replaced the annual IT investment allocation processes with quarterly business review cycles. They embarked on extensive coaching for everyone from executives to frontline developers and created a capability-building program for product managers. As a result, the bank realized capacity improvements of 20 to 30 percent, as well as a renewed focus on customer experience.
4. Create teams of higher-proficiency people
Operating in a faster, more productive technology environment requires a different kind of engineering talent than many banks have traditionally employed. Instead of maintaining large teams of less experienced people—including nonemployee contractors—leading institutions are moving to hire smaller in-house teams of highly proficient engineers accustomed to moving quickly to develop high-quality software. Although these engineers often command high salaries, their greater productivity outweighs the higher talent costs and creates a virtuous circle where these top engineers attract their peers. Highly technical and performant engineers are also critical for leveraging new AI-enabled development tools.
Attracting and retaining a different cadre of talent in a competitive market is not straightforward. It typically requires a major overhaul of the employee value proposition for technology roles, changing the recruiting processes, and investing in employee development.
One bank emphasized how working at the company offered prospective hires the opportunity to help millions of customers, work with the latest technologies, collaborate on cutting-edge teams, and learn in an environment that values engineers. IT and HR leaders created dedicated career paths where engineers would be automatically promoted when they were ready, rather than having to apply for a new role. They also created a talent accelerator staffed by full-time engineers and HR professionals to dramatically scale up the talent pipeline, set a high bar for candidates, and create a better experience for new hires by streamlining the onboarding process. As a result, in just over three years, the bank’s engineering staff went from 70 percent external to 70 percent internal—significantly improving productivity, creating a closer integration of business and technology road maps, and markedly reducing turnover.
How to make productivity happen
Every bank we speak with is in some stage of adopting these four principles; the alternative of remaining a technology laggard is simply not an option. However, many banks have not yet managed to drive measurable performance improvements through a productivity-focused technology development model.
In part, this is because these transformations are never straightforward. Many institutions struggle to measure the baseline productivity of their teams and make the case for improvement. There is a wide range of actions that executives can choose from, and some actions have uncertain returns, so it’s sometimes hard to know what to prioritize to achieve a strong ROI. The teams needed to implement these transformations are already busy delivering products, often running constantly to keep up with customer demands. As a result, many banks have not prioritized these transformations, even when certain teams are convinced of their value.
The institutions succeeding with a productivity-focused approach all have one thing in common: they got early support from the CEO and executive committee. Typically, C-level buy-in requires meeting five prerequisites:
- A bold aspiration. Best-performing banks encourage their technology leaders to get curious about what companies in other industries have achieved and then to set ambitious goals that assume barriers can be overcome.
- Quantifiable impact. A comprehensive productivity baseline is imperative to know which actions could have the highest ROI and to measure improvement. The best institutions start by defining baseline performance against some of the metrics outlined in this article and then quantifying the potential uplift and resulting benefits.
- Engagement outside IT. Engineering teams will devote time to improving themselves only if the transformation is a priority for their counterparts on business teams. In addition, a higher-speed delivery model also requires business leaders to operate differently to get value from it. Therefore, it’s critical to explain the value of the engineering transformation to business leaders and find early advocates who can sponsor the effort.
- Clear accountability on the IT leadership team. Transforming engineering is a team sport; each member of the IT leadership team will have their own unique part to play, from divisional IT leaders to the head of infrastructure to the chief information security officer (CISO). This requires orchestration, so several financial institutions have nominated one member of their IT leadership team to drive the transformation as a “first among equals.” This collaborative leader creates cascading actions and expectations and coaches teams to deliver.
- Early business wins. The best way to build confidence in a productivity-focused approach is to demonstrate early impact to business leaders in an area that they care about, such as a mobile app or central data domains.
Once these prerequisites are in place, changes can be rolled out at pace, starting from the center and then moving team by team. Unlike some technology efforts, productivity-focused engineering transformations often have a tangible impact quickly, since they do not rely on complex or time-consuming actions such as swapping out entire core platforms or migrating customers. This means institutions that act boldly to supercharge their engineering teams’ productivity can quickly deliver more for the business and customers.
