“Did you see the article about blockchain on the paper’s home page? I think we need blockchain, too.”
“We should give all our IT to an offshore system integrator—that will save a lot of money. Our competitors are doing the same thing.”
“We need to do something about generative AI right away!”
Many who work in IT have heard variations of these statements from well-meaning senior colleagues. And given how much technology underpins so many trends, as well as the increasing pace of innovation, IT leaders can expect to hear more of them and feel pressure to act on them.
Sometimes it makes a lot of sense to act quickly on a trend. Innovation, after all, is a hallmark of successful companies. But not all trends are created equal. The annals of many IT organizations are littered with bold initiatives in trends that created a lot of excitement but not much value. Moving quickly to chase a trend often bogs an organization down, leads to wasted spend, and takes attention away from important priorities.
For this reason, the ability to evaluate trends quickly and communicate their relevance to the business is becoming a critical capability for the modern CIO. In practice, it is rarely as simple as saying “yes” or “no” to investment in a particular trend. Trends are unpredictable, change with time, and their relevance to a given business often waxes or wanes. Instead, CIOs need a clear set of parameters to rely on both for evaluating trends and for determining which posture to take for engagement with them: first mover, fast follower, slow adopter, or sometimes non-partaker.
Register to join our next McKinsey Live
The four guides for determining a trend’s relevance
We researched trends of the past to find patterns that can help us better understand how to evaluate them in terms of their relevance to a business. There are many reasons why attempts to turn a promising trend into a profitable one for any given business may fail—insufficient talent or leadership nonalignment come to mind. But we have also found that trends that add value have inherent attributes that make them particularly valuable to any business. We’ve boiled them down to the following four:
- Disruptive business value: The trend can result in measurable value to the business.
- Independence: The trend allows the organization to work in smaller, more independent units.
- Connectivity: The trend reduces friction in the organization’s connectivity.
- Extensibility: The trend can broadly shape and improve the organization’s technology and management practices.
You don’t necessarily need a green light across each of these four parameters to make investment in a trend worthwhile. But if one of them comes up red or yellow, it’s worth taking a closer look before making a significant commitment. And while trends often initially aim to address just one of these areas, their relevance and impact rise dramatically when they can be applied coherently to all four.
It’s worth emphasizing that this evaluation isn’t a simple checklist. It requires rigor in the analysis, a willingness to review the analysis on its merits (without being influenced by how it might serve a narrow but favored “pet” project”), and creativity—some trends aren’t particularly meaningful on their own, but in combination (virtualization and cloud, for example) can have greater potential.
1. Disruptive business value
Almost any trend or development has the potential to improve something in an organization. The question is whether that improvement is worth what it costs. It’s important to understand the trade-offs. You may save money or generate value in one area by adopting a certain trend, but will it cost you in another?
The most important points to determine are whether the value is to IT alone or to the business overall, whether that value is merely incremental or significant, and whether success is clearly measurable in KPIs. If a trend improves an IT process but can’t be directly linked to a business advantage, then it’s probably not worth a significant investment. For example, much of the benefit of cloud computing comes not from improvements to IT productivity but from how it accelerates and enables business processes and innovation.
The value of some trends will depend on the sector where they are applied. Blockchain’s value potential, for example, has differed from industry to industry. While in some areas of financial services, such as cryptocurrency and collateral management, blockchain has the potential to create a lot of value, in other industries, such as consumer packaged goods, the value potential is less clear.
Key questions
- Does this trend provide disruptive business value? If the value is incremental or isolated to some operational improvement in IT, then the trend isn’t really worth a “big bet.”
- How will you measure value capture? Every initiative has metrics, but they are often hazy or do not clearly reflect business value. Be disciplined in developing measurable objectives and key results (OKRs), and determine which KPIs contribute to them. This clarity is critical during implementation to see whether what you’re working on is delivering value and why or why not.
Would you like to learn more about McKinsey Digital?
2. Independence
One of the tech executive’s key challenges is that IT often has too many interdependencies, which leads to technical debt and administrative processes such as alignment meetings and process coordination—for example, when changes to the billing system depend on numerous other systems and development teams, meetings and delays are inevitable consequences.
The concept of “modularity” has been in vogue for almost two decades now, but the enthusiasm for it has generally not been matched in its implementation. APIs have certainly helped, but they don’t address the important organizational changes that also need to happen to reduce dependencies. That’s where moving toward a product and platform operating model—where independent teams work on user-facing products while platform teams build capabilities to support them—can have significant impact. Assessing how a trend can be profitably adopted and supported by either product or platform teams should be a critical criterion for evaluating its potential.
Many of us are familiar with the big trend some 10 to 15 years ago toward large-scale outsourcing of IT to an external system provider. The rationale was that external providers would reduce IT’s overall cost. For some industries with low margins, large-scale outsourcing deals were at least partially successful. But the system dependencies didn’t change. So, as the need for speed and flexibility in technology grew, the savings from outsourcing were increasingly outweighed by the system dependencies that remained in place and slowed the pace of development. With modern software as a service (SaaS), on the other hand, only a specific, relatively independent, and clearly contained functionality is outsourced to a vendor, making this approach much more scalable.
Key questions
- Does this trend allow me to make my teams more independent—for example, by automating communication at interfaces or reducing the need for coordination and up-front planning?
- To what degree will this trend enable a more modular IT estate? Look for specific benefits, like a trend’s ability to separate different applications from each other or abstract layers of the technology stack.
- How many parts of the IT estate are needed to make this trend run well? The greater the number, the greater the complexity and the more likely you’ll run into inevitable workarounds that will add to technical debt.
3. Connectivity
While independence is important, a trend that operates like a satellite on its own or divides the organization up into disconnected units isn’t going to lead to at-scale impact. There is a difference between dependency, which isn’t good, and leverage, which is critical for delivering value.
That difference was clearly illustrated in enterprise service buses (ESBs), which afforded independence but required additional alignment and orchestration, sometimes through a single central team, which created a bottleneck.
Virtualization is another trend that highlights the need to balance independence and connectivity. Initially focused on improving the usage of hardware assets—specifically, the CPU—virtualization also made applications less dependent on each other and on larger systems by not forcing everything to run on a single server. But it provided limited connectivity between individual virtual machines. There was no structured capability in place that made it easy and convenient to move workloads, scale them, and manage the connections. This changed with cloud computing and containerization, which connected individual machines through a central management layer.
In contrast, modern microservices patterns aim to both improve independence and clarify the interfaces and thus improve connectivity by reducing the need for communication or a central coordination mechanism.
“Improved connectivity” should not be misunderstood as more communications—that’s a recipe for creating more dependencies. Improved connectivity is often reached through more clarity on the interfaces, reducing the need for communication.
Key questions
- Does this trend increase clarity at the interfaces and reduce the need for coordination? You’re looking for a trend that can use stable and reusable interfaces between technologies.
- Does this trend lead teams to work with each other more effectively due to less “waste” in communications?
- In what way can this trend leverage existing technology assets? It will be important to understand what kinds of interfaces will make the necessary connections, and how easily they can be implemented and maintained.
The CIO Agenda
4. Extensibility
The impact of a technology trend increases with its applicability and coherence across the IT estate. A technology trend that touches only one part of the IT estate in isolation or is managed simply as a “tech product” does not often have the widespread impact that comes with true innovation. Impact often requires companies to think through both the technology and operational implications to open up the potential. Broad applicability is where you get to scale.
Virtualization, microservices, or SaaS-based services, for example, were less successful when they were managed as a pure tech product. Only when combined with enabling factors on the operating side, such as agile, DevOps, and the necessary support structures (such as budgeting roles and HR support) did they have much greater impact.
Similarly, agile works best when all relevant parts of the company work in agile ways. (The impact of agile developer teams is improved when operations, for example, also work in agile ways, which is how we came to DevOps.) The key question for the CIO, then, is how to adjust the technology estate to extend the range of application across technology and the business.
Key questions
- How can IT management practices adapt or innovate with the trend?
- Will this trend require the business operations and decision processes to change in order to develop its full potential?
- Are all the support processes (such as budgeting, HR, architecture) in place to support the trend?
Adopt or not: A strategic decision
As they say in comedy, timing is everything, and that’s just as true when it comes to trends. Going in on a trend too early or too late can sometimes be worse than doing nothing at all. Just as important as evaluating a trend’s “bona fides” according to the four parameters described above is determining how best to engage with it. CIOs should consider four possible engagement approaches:
- First mover: This engagement makes the most sense in the case where the trend has significant impact on the company’s core business model (which is to say, it’s a matter of business survival). This approach generally involves investing significant resources of time, people, and money. The big trap here is convincing yourself that the first-mover approach is the only viable option.
- Fast follower: This approach is best applied when the trend could have an important impact on your business model or open up a significant revenue stream. This engagement posture can work when the trend is still too early in its maturity cycle to understand how it can best be harnessed, or when the business doesn’t yet have sufficient capabilities in place to act on the trend. A good signal of a trend’s importance is when big players lead the innovation and go all in. The danger here is taking a “wait and see” approach too long and finding yourself much further behind your competitors than you expected. Greenlighting pilots can make it seem that you’re taking action, but unless those pilots are closely tracked and—most importantly—scaled, then this activity has no real impact.
- Slow adopter: This approach is an acceptable option when the trend isn’t directly relevant to the core business or is not mature. It might make sense for niche applications of a trend. The danger here is that complacency and an overly narrow view of the trend’s relevance, such as focusing only on how existing competitors might use it and failing to consider emerging players, leave your company too far behind to catch up.
- Non-partaker: Some trends just don’t make sense for a business to adopt. This is often an unpopular approach when the board or senior management is demanding action. A clear view of business goals and the impact of a trend on those goals is critical. But complacency is also a danger here when legacy-driven thinking argues that a given trend “doesn’t apply to us.”
The call on whether and how to adopt a trend is not a one-time decision. It requires continual review as technology matures and evolves, new implementation models and supporting services scale, or the market situation shifts. But CIOs who use the four parameters as a compass to determine relevance and think through their engagement options can better guide their organizations toward turning trends into value.
