Large-scale agile tensions in traditional banks
The disruption caused by agile-born fintech and contactless digital banking incumbents is pressuring banks to innovate and deliver products and services faster.
Despite pockets of success, traditional banks struggle to balance the tensions involved in transitioning to large-scale agile.
The disruption caused by fintech and contactless digital banking incumbents capable of launching new features every few months poses a significant threat to traditional banks.
At the same time, an established regulatory framework controls banks in South Africa. The regulator closely supervises local banks to ensure they comply with international regulatory standards and best practices.
For example, the regulator approves the bank’s external auditors responsible for reporting irregularities or suspected irregularities. At the same time, the internal audit unit monitors and reviews the bank’s internal controls, including the electronic information system and electronic banking services and risk management procedures. They provide regular reports to senior management and the board of directors.
The challenge for traditional banks is to develop a dynamic capability that affords them more agility while complying with regulatory, audit and risk requirements.
Delivery with speed and stability simultaneously
Delivering high-quality products and services quickly at scale and keeping bureaucratic controls at a minimum is a major challenge for traditionally highly-regulated banking organisations. During such a large-scale agile development (LSAD) transformation journey, one of the major dilemmas is balancing the ensuing tensions between autonomy and control.
Traditional governance and agile work designs and development practices focus on quality software, customer satisfaction and value delivery; however, the main concerns of agile centre on adapting to change and accelerating time to market.
In contrast, governance structures emphasise strategic alignment, budgetary controls, project governance, regulatory, audit and risk compliance. Furthermore, while traditional governance frameworks that emphasise stability and control are slow, costly and tedious to follow, agile emphasises the delivery of software with speed and stability simultaneously.
Some experts question whether traditional governance models are appropriate or sensible in an agile approach. While this sentiment may apply to digital start-ups and agile-born organisations where regulations have not kept pace with digital technologies, skirting regulations would not hold for already highly-regulated and mature organisations burdened by their legacy systems.
After all, dodging substantive governance controls can be detrimental to the reputation of traditional organisations and their customers.
Tailoring large-scale agile development frameworks
The use of agile on large-scale projects is still relatively new. A recent survey found that many traditional organisations are undergoing large-scale transformations. They are employing LSAD frameworks, such as Disciplined Agile Delivery, Large Scale Scrum, Scaled Agile Framework, Spotify, Nexus, and Recipes for Agile Governance in the Enterprise.
While these frameworks tend to emphasise more autonomy in their set of agile practices when compared to traditional plan-driven development approaches, they have not paid sufficient attention to the processes underlying autonomy-control tensions in a highly-regulated organisational context.
One should be wary of the dogmatic agile evangelist who proposes an ‘all or nothing’ approach.
While large-scale agile approaches build more controls into agile development practices, ensuring that this integration does not jeopardise the agile team’s autonomy is a major concern for practitioners and researchers, as this can dilute agile benefits.
Banks are also learning that forcing employees to follow the prescribed agile guidelines rigidly will be counterproductive, so an increasing degree of tailoring occurs. Tailoring often highlights the challenges of fitting new agile cultural norms with an organisation’s traditional cultural norms.
For example, the cultural tensions between talented young incumbents and the inertial forces of the knowledgeable but disenchanted old guard are impeding valuable aspects of agile practices. Meanwhile, tangible benefits of agile can be realised even in small doses.
One should be wary of the dogmatic agile evangelist who proposes an ‘all or nothing’ approach.
Rationalising legacy systems while building a digital platform
The core systems of traditional banks were originally designed to support their brick-and-mortar operations. These are made up of dated legacy applications written in COBOL and running on mainframes, while more recent online banking platforms are built on Java and .Net.
Running these disparate platforms over the years means that traditional banks have accrued a significant amount of technical debt. The process of standardising and rationalising several core banking systems that banks built up over decades is a significant challenge.
Banks have also begun ‘aggressively’ decoupling from multiple legacy and mainframe IT systems, while building a platform to enable faster build and delivery of new digital products and services. Legacy capabilities are still being migrated onto the ‘enterprise digital platform’ of traditional banks.
Understanding the benefits of autonomy and control
The agile approach is predicated on the assumption that team performance increases with less, not more, control.
However, large-scale agile transformations in traditional and highly-regulated organisations need to incorporate the benefits of both autonomy and traditional control elements to strike an optimal balance.
Organisations may be better off configuring different combinations of autonomy and traditional controls depending on their context (type of project, technology platform, risk appetite, culture).
Leaders should understand the interplay of opposites as an essential feature of large-scale agile transformations and operating at large-scale aim to satisfy both oppositional demands.
Effective agile transformations are not characterised by eliminating control or resolving the contradictions brought about by the seemingly opposing features of autonomy and control.
In striving to accomplish high-performance in large-scale agile, team members have simultaneous responsibilities for autonomy and control.
Leveraging the benefits of autonomy and control
Leaders of agile transformations have to be cautious of not framing agile principles in “either/or” ways.
For example, where the discourses of autonomy are taken for granted, it is more likely to worsen the autonomy-control tension, as agile teams tend to over-focus on autonomy to the neglect of control.
Rather than minimising the importance of traditional governance approaches or overemphasising the benefit of agile autonomy, those leading the agile change should emphasise the synergetic benefits of autonomy and governance.
Since leaders need to pay attention to both the organisation’s governance requirements and the agile team’s need for autonomy, they should use generative images to transcend polarised thinking about control.
For example, framing the autonomy-control tension as ‘controlled autonomy’. The ideal amount of autonomy may gradually emerge over time as the organisation matures in the agile journey.
However, agile teams in a bureaucratic organisation may have to accept excessive control tensions during early phases of the agile journey and adopt practical solutions to derive optimal value from large-scale agile.
The implication is that experienced agile team members have to endure a more painful transition process in such a context.
The ongoing balancing of agile and control tensions is a vital capability
The autonomy-control tension can be used as one of the primary drivers of a large-scale agile transformation.
During a large-scale agile journey, leaders will need to work out the optimum autonomy-control tension. There are no guarantees of eliminating the autonomy-control tension.
Dialogue may soften the negative impact of this tension. Over time, this tension may lessen as bureaucratic silos are replaced by self-governing product teams, but eliminating this tension is unlikely.
IT and business unit leaders should adopt a dialogical and collaborative approach and offer a more democratic and humanistic response to the autonomy-control challenge facing large-scale agile transformations.
Understanding these large-scale agile tensions can provide IT and business leaders with new insights on how to help agile teams and traditional units cope better with change.
The dark side of a large-scale agile transformation is that expectations about autonomy can be thwarted if it is only superficially introduced. This will leave valuable agile team members and software developers feeling frustrated.
I suggest that leaders set the expectations that agile is more about individuals and teams taking the initiative, not absolute control.
To sum up: collaboration in agile cannot exist without both autonomy and control, especially at traditional banks – and finding the appropriate balance is the key.
* Based on a paper presented with co-author Sydwell Rikhotso at the 29 International Conference on Information Systems Development (ISD2021 Valencia, Spain).
Professor in Information Systems (IS) at the Wits School of Business Sciences.
Rennie Naidoo is a professor in Information Systems (IS) at the Wits School of Business Sciences. An established NRF-rated researcher, his focus areas include data science, sustainable IT, artificial intelligence and cyber security. Naidoo earned a Master of Commerce in Information Systems (with Distinction) from the University of the Witwatersrand and a PhD in Information Technology from the University of Pretoria. His research appears in leading scholarly journals, such as the Journal of Strategic Information Systems and the European Journal of Information Systems. He has presented at premier academic conferences, including the International Conference on Information Systems and Hawaii International Conference on System Sciences. Devoted to lifelong learning and scholarship, he mentors students at the Master's and PhD levels at Wits. He actively serves the IS profession as a member of the Association for Information Systems and the South African Institute for Computer Scientists and Information Technologists.
Rennie Naidoo is a professor in Information Systems (IS) at the Wits School of Business Sciences. An established NRF-rated researcher, his focus areas include data science, sustainable IT, artificial intelligence and cyber security.
Naidoo earned a Master of Commerce in Information Systems (with Distinction) from the University of the Witwatersrand and a PhD in Information Technology from the University of Pretoria. His research appears in leading scholarly journals, such as the Journal of Strategic Information Systems and the European Journal of Information Systems.
He has presented at premier academic conferences, including the International Conference on Information Systems and Hawaii International Conference on System Sciences.
Devoted to lifelong learning and scholarship, he mentors students at the Master's and PhD levels at Wits. He actively serves the IS profession as a member of the Association for Information Systems and the South African Institute for Computer Scientists and Information Technologists.