Digital Factories: Accelerating Developer Velocity
McKinsey describes a 'Digital Factory' approach for Digital Transformation, moving to self-organizing Agile Squads to increase 'Developer Velocity'.
Two McKinsey articles are ideally combined when considering the challenge of growing the speed and success of your IT & DevOps teams.
In ‘Welcome to the Digital Factory‘ McKinsey review the new organizational models being employed by disruptive digital leaders – The ‘Digital Factory’.
Complimenting this is ‘How software excellence fuels business performance‘, where they explain how a headline metric of ‘Developer Velocity’ can ensure digital investments.
The two are an ideal combination as a Digital Factory approach identifies new ways of working that tap insights from the world of manufacturing, applying a production line metaphor for software workflow management, that can be monitored and managed through KPI metrics under a headline of ‘Developer Velocity‘, so that output can be improved via the same type of process improvement regimes.
As they observe in companies like Goldman Sachs, software engineers now make up 1/4 of the total workforce. In addition to mastering the nuances of their industry, businesses today need to excel first and foremost at developing software, and their articles form a helpful implementation model for getting started.
In their in-depth article McKinsey explores the dynamics and successes of the ‘Digital Factory’ model, a “construction site” where change happens.
“A Digital Factory is the “construction site” where change happens. It comprises dedicated, cross-functional teams that work together on change-the-business programs. They resemble factory workers in that they employ reusable tools and repeatable processes to build specific “products” in the form of new experiences, services, or solutions. The secret to the Digital Factory’s success is that its small teams, working closely with the business side, function as a start-up accelerator.”
“We see reductions in management overhead of 50 percent for technology teams in the DF, 70 percent in the number of business analysts needed to write technology requirements, and, as test automation becomes the norm, a drop of 90 percent in the number of testers. Finally, we see top engineering talent performing at eight times the level of their peers, as measured with metrics such as code commits.”
Through the story of a visit to the Digital Factory of a Telecomms business, McKinsey introduce and explain an approach to IT organization that has yielded benefits including:
- bring products to market faster (in six months versus two years)
- do more with existing resources (eight product launches per year versus one or two)
- create dramatically reimagined experiences (opening an account in five minutes versus ten days)
- reduce tech development costs by a third (fewer managers per engineer)
- attract the great talent required to compete in a digital world
Given a context that only 16% of executives rate their digital transformation projects as successful, this recipe for a better way to improve IT delivery overall is therefore very appealing. Especially as they describe how it can be treated as a ‘sidebar’ unit, a new innovation-centric digital team that doesn’t disturb and ‘bolts on’ to the existing legacy business, or as a complete new broom, wholesale transformation.
Ultimately the important success factor is building them around key strategic goals – For example reducing mortgage loan rate application turnarounds from three weeks to minutes. These kind of major leaps in the market deliver real strategic advantage.
Digital Factories mainly address the large scale enterprise complexity required to achieve those kinds of improvements, notably breaking down into self-managing, autonomous teams. These ‘Agile Squads’ are typically 10-15 people in size and bring together business and IT stakeholders to action, and validate, ideas much more quickly.
They are guided by a central Centre of Excellence, which acts as the ‘control tower’ and provides for central functions notably skills and tools assets and services, such as:
“One of the most important roles in the nerve center is the lead product owner. This person is not a manager but a “doer-in-chief,” who works with the local product owner to ensure quality of execution, provide weekly coaching, pressure-test business cases, and review team progress with the goal of understanding bottlenecks and helping to break them. The lead product owner is particularly involved during the first month that the squad is in place and ideally acts as its mentor.”
If done right, the Digital Factory is a powerful engine to enable and accelerate not only the business but also the IT transformation agenda. That’s because the missions that Digital Factory teams work on have sufficient scale and clarity of focus for IT to develop next-generation capabilities such as automated testing, cloud-based applications, secure coding practices, and application programming interfaces (APIs) that can be put to immediate and practical use.
Accelerating DVI – Developer Velocity Index
McKinsey developed the Developer Velocity Index (DVI) from an in-depth survey of senior executives at 440 large enterprises, more than 100 expert interviews, and extensive external research.
It pinpoints the most critical factors (related to technology, working practices, and organizational enablement) in achieving Developer Velocity. It expands the scope of how Agile traditionally defines Velocity, to connect it with a general ability for IT to deliver strategic value.
They identify top-quartile DVI scores correlate revenue growth that is four to five times faster than bottom-quartile DVI scores.
Top-quartile companies also have 60 percent higher total shareholder returns and 20 percent higher operating margins. In addition, top-quartile players appear to be more innovative, scoring 55 percent higher on innovation than bottom-quartile companies. These businesses also score higher on customer satisfaction, brand perception, and talent management.
Importantly the bulk of the article focuses on the best practices they learned from their interviews, that describe what those successful businesses have done to achieve those improvements:
By far the most impactful area is developer tools, somewhat obviously – Better empower developers, better speed quality and throughput of code produced.
“According to our research, best-in-class tools are the primary driver of Developer Velocity. Organizations with strong tools—for planning, development (for example, integrated development environments), collaboration, and continuous integration and delivery—are 65 percent more innovative than bottom-quartile companies.”
This empowerment is powerful exemplified through new trends notably ‘Low Code’ development.
“Top-quartile companies give developers a degree of choice—usually between two and five options to account for different needs and preferences—but restrict ad hoc tools from being added. Leading companies also use tools to unleash Developer Velocity by investing in low-code and no-code platforms. These platforms enable the average business user to develop applications without any software experience, freeing up seasoned developers to focus on the most challenging tasks.
For example, one pharmaceutical company grew its low-code platform base from eight users to 1,400 in just one year. Business users outside of IT are now building applications with thousands of monthly sessions. The companies in our survey that empower “citizen developers” in these sorts of ways score 33 percent higher on innovation compared with bottom-quartile companies.”
DevOps Flow Automation
Another key consequence of better developer tools is improved management of the whole environment and elevated maturity levels, such as the ability to recover from failures, as a function of a more capable, controlled environment.
“Companies that perform best at this aspect of cultural change also invest in systems that can absorb and minimize the cost of failures. These investments include capabilities such as controlled releases, feature flags (the ability to turn features on and off without redeploying code), and automated rollbacks, as well as postmortems and retrospectives that allow teams to reflect constructively on what worked and what did not.
A software leader at one top-quartile company said, “You need to implement safeguards in order to embrace failure, so we build contingencies as part of the software-development process. For example, we install a new version side by side with the stable version.”
A backbone enabling component of this new tool sets is the use of Public Cloud:
“Public-cloud adoption as a catalyst of Developer Velocity is especially strong for nonsoftware companies—public-cloud adoption has four times the impact on their business performance than it does for software companies. Developer Velocity benefits are also sharply correlated with the degree of adoption: companies in the top quartile of public-cloud adoption have DVI scores that are 32 percent higher than companies in the bottom quartile.”
Hand in hand with Cloud adoption is also more adoption of open source tools:
“For organizations that already have a strong DVI score, open-source adoption acts as a major accelerator. The data show that top-quartile company adoption of open source has three times the impact on innovation as compared with companies in other quartiles.
Top-quartile DVI companies are especially active adopters, scoring 36 percent higher on open-source adoption than the next quartile—the highest delta on any dimension studied. We found that building an open-source culture is about more than using open-source software within the code; it extends to encouraging contribution and participation in the open-source community as well as adopting a similar approach to how code is shared internally—that is, strong InnerSource adoption.”
Talent and Skills
Companies seeking a competitive edge through technology place a keen emphasis on learning and skills development.
“Leading companies are resourceful when it comes to keeping software talent happy and motivated. One leading telecom company offers a wide range of skills certifications or “microbadges,” from beginner’s-level mobile development to machine learning. It also created a Developer University to provide developers with fresh learning opportunities and the chance to apply these skills in their workplace.”
Product Management as a Core Capability
Hand in hand with this improved infrastructure management are better organized related practices, notably the Product Management that guides what code is developed, when and why.
“The product-management team not only needs relevant business and market knowledge but also a strong technical background. Companies with above-average performance across the six dimensions have DVI scores 1.5 times higher than companies with top-quartile performance in just one or two dimensions. It is important to note that excellent product management is also not about the discrete product-management team; developers and other members of an agile team are increasingly wearing the product-manager hat to understand how their work is aligned with business priorities and customer needs.”
They also look to future trends that will further impact on this scenario, in particular AI.
“The analysis also identified emerging drivers with the potential to accelerate DVI scores over the next three to five years. Top-quartile companies are increasingly exploring the use of artificial intelligence (AI) and machine learning in developer tools. For example, some have begun using AI to perform aspects of pair programming (in which typically one developer writes code while another almost simultaneously reviews it), providing automated code reviews and using natural language processing in low-code tools.
Additional areas that executives believe will accelerate software innovation and impact in the future include increased usage of product telemetry to make product decisions and automation in detecting and remediating production issues.”