Business leaders lament sluggish software development pace, according to a survey report
A new study by Forrester Research, sponsored by development consultancy ThoughtWorks, has shed light on the challenges facing IT system delivery and the mismatch in perception between business and IT leaders.
The study argues that the adoption of agile and lean development practices, continuous integration, dev-ops techniques, and principles of the lean start-up can significantly reduce the time it takes to release new software. However, it also indicates that large software upgrades rolled out intermittently, rather than small ones regularly, slow down the development process.
The study suggests that low maturity in continuous delivery is a major reason for slow software release times. This is further compounded by the fact that only 44% of IT leaders automate software deployment consistently, as per the study.
The study reveals an apparent mismatch in the perception of the software development function between business and IT leaders. Over half (51%) of business leaders want "strategic IT services or software products" to be delivered within six months. However, a larger proportion (41%) said it would take more than a year, a finding described as "worrying" by Forrester.
The study identifies several key reasons for this disparity. Legacy tech infrastructure and accumulated tech debt hinder rapid development and delivery. Forrester advises radical approaches such as "declaring tech debt bankruptcy" and outsourcing legacy stacks to free up resources for modern, adaptive, AI-powered systems.
Governance complexities also play a significant role. As data governance evolves beyond compliance to support trust, agility, and AI readiness, business expectations for quick delivery clash with the need to embed agentic AI for tasks such as policy enforcement and data product creation. This shift demands more sophisticated tools and governance practices that can slow down delivery initially.
AI adoption challenges also slow the pace at which IT systems can be delivered and trusted. While AI-driven tools promise efficiencies, regulatory risks, AI fatigue, and integration difficulties still pose significant hurdles.
To bridge this gap, the study suggests several practices. Investing in employee data literacy and AI readiness to ensure the workforce is equipped to integrate and manage AI-enabled systems effectively is one such practice. Adopting a cloud-as-necessary strategy rather than cloud-first, balancing cost, sovereignty, and compliance constraints to speed deployment where appropriate is another.
The study also recommends experimenting with agentic AI for task automation initially in read-only/analytical apps to reduce risk, enabling faster incremental delivery and trust-building around AI components. Implementing modern data governance solutions that empower business users with embedded guidance and autonomy is another key strategy.
Finally, the study advises addressing tech debt decisively by outsourcing or eliminating legacy systems to free resources for innovation and speedier system delivery.
In conclusion, the study highlights that unmet expectations for six-month IT delivery stem from legacy systems, governance evolution, and new AI adoption complexities. Bridging this gap requires modernizing infrastructure, enhancing AI readiness, rethinking cloud and governance strategies, and empowering business users.
The survey was conducted among 325 business and IT leaders from the US, UK, and Australia. Over the past few years, many forward-looking development shops have adopted these principles to enable continuous delivery.
The study proposes that adopting modern technology, such as AI-powered systems, can help bridge the gap between business and IT leaders' expectations for six-month IT delivery. However, challenges like AI adoption complexities, legacy tech infrastructure, and governance complexities slow down the process.
To speed up software release times, the study suggests practices like investing in employee data literacy, adopting a cloud-as-necessary strategy, experimenting with agentic AI for automation, and addressing tech debt decisively by outsourcing or eliminating legacy systems.
