How quickly and cost-effectively can your IT organization deliver new tech capabilities or update the ones you have? The answer might have greater impact on your bottom line than you expect.
Our latest research on IT productivity shows how significant that answer can be. Enterprises with high-performing IT organizations1 have up to 35 percent higher revenue growth and 10 percent higher profit margins (Exhibit 1). The research also finds that an average company has an opportunity to optimize and potentially reinvest 30 percent of its IT spend through improvements to IT productivity. Combined, these findings highlight how much the ability to manage IT efficiently is not just an operational concern but a business necessity. (See sidebar, “About the research.”)
About the research
We conducted surveys with 158 CIOs and CTOs across Asia–Pacific, Europe, the Middle East and Africa, and the Americas on the top-line effects of 58 indicators for technological excellence across seven areas—architecture, infrastructure, data and AI, engineering practices, processes and practices, tech organization, and talent—which are measurable drivers of technological maturity (exhibit). Our survey sample was evenly distributed across the following industries: banking, consumer, global energy and materials, advanced industries, insurance, telecom, and travel, logistics, and infrastructure; we also had a separate quota for 20 start-ups and growth companies across industries.
In addition to the survey, we also modeled the potential impact of these different productivity levers on a typical company’s IT spend based on McKinsey tech productivity practice data to understand what capacity can be freed by these levers for reinvestment in value-adding activities. We also validated the survey results against companies that have recently applied these levers in practice. This research expands on previous McKinsey research on developer velocity and operating model maturity; it spans both traditional and digital-native companies and provides new insights into the importance of simplifying developer journeys and decoupling the tech landscape.1
As experienced leaders know, IT productivity is a hotly debated topic with little consensus on where to start and how to measure improvements. Our research, however, is bringing more clarity to the issue with four new insights:
- Time-to-market is the metric that matters most in evaluating IT productivity. It correlates more strongly with higher profit margins than any other frequently used business performance metric.
- Three critical delivery capabilities—cross-functional delivery, vendor independence, and the use of public cloud—also correlate with higher profit margins.
- The most important practices for achieving these delivery capabilities are improving developer journeys and engineering practices; adopting a cross-functional, product-led operating model; and reducing dependencies between IT systems and applications.
- These capabilities and practices, properly implemented, have the potential to deliver solid returns on investment and should be a critical area of focus within a company’s broader efforts to rewire their organization for digital and AI.
Insight #1: Time-to-market is the metric that matters most in evaluating IT productivity
Of the five frequently used business performance metrics for measuring IT efficiency and effectiveness that we assessed, time-to-market on implementing changes had the strongest correlation with higher profit margins (Exhibit 2). Its correlation with higher revenue growth is three times stronger than that of customer satisfaction and seven times stronger than that of employee satisfaction.
What time-to-market should organizations strive for? For a significant share of high performers in our survey, the average time for completing a medium-size change request, such as adding a new product feature or changing pricing logic, was two to four months from initial idea to production. In contrast, less-advanced peers required up to a year to complete such changes (Exhibit 3).
Our research also found that two other factors are strongly aligned with better business performance: a higher share of IT investment relative to revenue and a higher frequency of production releases. For example, high performers were more likely to make monthly updates to core applications such as enterprise resource planning systems, while quarterly updates were standard among conventional peers.
One potential explanation for this might be that more-profitable companies can afford greater technology investments, which gives them a productivity advantage over competitors. However, profitability alone may not fully explain these patterns—other factors, such as strategic priorities, likely play a role as well.
Would you like to learn more about McKinsey Technology?
Insight #2: Three delivery capabilities—cross-functionality, low vendor dependency, and use of public cloud—correlate most strongly with high profit margins
The research also surveyed leaders on their use of 50 common delivery capabilities across seven categories: architecture, infrastructure, data and AI, engineering practices, organizational processes and practices, organizational characteristics, and talent. We then correlated the responses to company profit margins.
As expected, digital-native tech companies and start-ups—which have often not been included in such research in the past—surpassed their traditional industry peers in IT maturity. But some companies in traditional industries, including global energy and materials, banking, and telecom, are closing the gap (Exhibit 4).
Three particular delivery capabilities have a much stronger impact on profit margins than the others for companies with the greatest IT maturity, regardless of industry (Exhibit 5):
- Cross-functional delivery. Companies whose technology teams are comprised of 20 to 40 percent business professionals, subject-matter experts, and colleagues relevant to product delivery tend to have higher profit margins than those whose teams consist primarily of tech experts. The majority of top performers also have a broad range of technical skills, including security operations (SecOps) and development operations (DevOps).
- Low vendor dependency. High performers are more likely to use internal IT resources, using vendors and contractors to develop less than 10 percent of their applications. In contrast, conventional companies rely on vendors and contractors for over 40 percent of their application development.
- High share of workloads on public cloud. On average, enterprises run only 15 to 20 percent of their applications on cloud today.2 Our survey found, however, that companies with the highest profit margins are more likely to have more than 30 percent of their workloads running on a public cloud infrastructure. This finding is consistent with the latest McKinsey Global Survey of technology and business leaders, in which a majority of top performers3 report adopting cloud at scale.4 Combined, these survey results support a fundamental need for cloud services to support enterprise generative AI (gen AI) solutions and for developers to use gen AI to .
Insight #3: Three practices are priorities for improving IT productivity
Developing effective delivery capabilities to drive IT productivity requires strong practices. Of the dozen we analyzed, three seem most promising, though they all had various levels of interdependency (Exhibit 6).5
Even small improvements in developer journeys and changes to engineering practices can help a lot
In our experience, “change” programs (as opposed to IT maintenance) often account for roughly half of the total budget spent on applications. As a result, even minor productivity improvements to streamline manual and error-prone processes have the potential to visibly impact team productivity.
One large bank, for instance, set out to double its delivery capacity over two years by removing obstacles in developer environments, such as highly manual processes, and adopting continuous integration/continuous delivery (CI/CD) practices. In launching this effort, the bank tested the new engineering practices across two groups of engineers, totaling 500 people, to confirm the processes would deliver the expected productivity gains; refine workflows, tools, and documentation based on real-world feedback; and make a case for additional investment in tooling and engineering skills. The engineers and technology leaders who participated in the pilot served as champions and trainers during the full rollout to help ensure the program’s success. As part of its next phase in improving developer journeys, the bank is evaluating gen AI coding and quality-assurance tools to accelerate daily tasks, which in some cases have resulted in a 10 to 15 percent productivity lift.6
Typically, the work of streamlining developer journeys and engineering practices in a mid- or large-size company can be led primarily by IT and completed in less than a year. However, CEOs can facilitate this change by ensuring sufficient resources and shielding the tech organization from excessive business demands, such as supporting change freezes during the transformation.
As more gen AI tools to improve developer productivity come to market, business leaders may need to consider new skill-building strategies and tooling to take full advantage of them. At another leading bank, executives found that the team’s pilot using gen AI capabilities to refactor and migrate legacy code on a sample application enabled 40 to 50 percent faster completion times than conventional manual methods, which is expected to lower the cost of modernization by 30 to 40 percent at scale.7
Embracing a product and platform operating model empowers teams to work at scale
Our research finds that leading companies empower cross-functional delivery teams to work independently to achieve a given business outcome, such as improving user experiences or building internal services for reuse. The product and platform operating model provides this level of independence, while preserving sufficient oversight.8 And the latest McKinsey Global Survey of technology and business leaders confirms that top-performing respondents9 are more than twice as likely as others to report use of product- or platform-centric operating models.10 The most mature product and platform models are developed through collaboration with business and technology leaders across the C-suite (including the CEO) to restructure all facets of how work gets done, including talent, processes, and technologies.
The first step one financial services company took in adopting a product and platform operating model was having business and technology leaders meet to align their product teams along service domains, such as payment operations, loan operations, and customer relationship management. This put the organization on track to accelerate time-to-market for new capabilities and reduce expenses related to legacy and other non-strategic applications by more than 50 percent over two years.
Decoupling core-system capabilities that change frequently is key to facilitating greater cloud adoption and operating model maturity
Transitioning to a product and platform model may not deliver full business benefits if teams must still navigate siloed approval processes and outdated systems. Our research finds that leading companies kept pace with digital natives in decoupling applications (reducing dependencies between applications to allow for faster fixes). One bank, for instance, found that, despite its shift to a new operating model, it at first couldn’t reach its goal of cutting time-to-market on projects from up to 20 months or more to less than a few weeks. The culprit? Three outdated, complex, monolithic core systems that required extensive updates with every product change. As a result, nearly two-thirds of developer time was spent maintaining and managing complex dependencies in these core systems, leaving little time for more strategic work. By decoupling frequently changing capabilities embedded in the core systems and turning them into cloud-based microservices, the organization was able to increase the number of releases per year from four to more than 10,000 and reduce the time-to-market to as little as hours for small changes, such as A/B testing of new customer journeys.
Effective governance—clear policies, metrics, and processes—is important in choosing which capabilities to decouple and how to manage them on an ongoing basis. To facilitate its shift to decoupling applications, a digital-native retailer adopted policies for managing the application life cycle, tracking technical debt, assigning priority and engineering capacity, and sunsetting features that no longer delivered business value. These steps have enabled the organization to more effectively prioritize which capabilities to invest or divest in, decouple the architecture, and simplify the overall delivery process, putting IT on track to speed application changes and cut engineering time spent on maintenance from 50 percent to 20 percent.
Insight #4: When done well, companies making these changes can expect a solid return on investment in productivity
In all, our research estimates that increased technology maturity could unlock almost 30 percent additional value through productivity gains. The impact within each sector will depend on its current level of IT adoption and maturity.
On average, organizations can expect at least a 24 percent reduction in their total IT spend. Some industries, such as travel, logistics, and infrastructure companies, could realize up to 34 percent higher marginal returns from their efforts, given their lower IT maturity. But the real value comes not just from the savings themselves but from their reinvestment into further IT advancements, positioning companies for even greater IT productivity, innovation, and sustained growth.
The dynamics of IT productivity are complex. While each organization will need to tailor these insights to its own unique context, this research offers a framework for evaluating existing efforts and prioritizing strategies that can increase both revenue and profit.
