For years, enterprise technology decisions have been framed as a binary choice: build custom solutions or buy packaged software. With the emergence of AI, agentic engineering capabilities, and changing economics, that binary is shifting.

Custom solutions are becoming more viable, with lower build and maintenance costs, greater speed, and more flexibility as agentic capabilities reduce the effort required to develop and sustain them. At the same time, SaaS platforms continue to evolve, with vendors embedding AI and adapting their delivery and pricing models as capabilities expand and costs scale over time. Together, these shifts are expanding the range of options and require a more deliberate approach to determining where SaaS is the right answer and where custom solutions now make more sense.

Rather than replacing existing platforms outright, organisations are increasingly combining SaaS and custom solutions, retaining platforms where they provide scale and reliability, while extending, augmenting, or selectively replacing them where differentiation, control, or economics justify it. These decisions are rarely made at the enterprise level, but instead at the domain level, with different choices across the technology landscape.

As a result of these decisions, enterprise environments are becoming more distributed, with work spanning vendor and enterprise systems. This increases the importance of clear accountabilities, strong data foundations, and cross-functional governance across technology, data, and business teams.

As organisations revisit build vs. buy decisions, they’re asking more targeted questions aligned to business outcomes:

If you’ve been running VMware for the better part of a decade, you probably didn’t expect to be rethinking that choice any time soon. Broadcom completed its $69 billion VMware acquisition in late 2023, and the licensing changes that followed have put a lot of IT shops in a difficult position.

As someone has previously put it, “There is no longer any such thing as a reasonably priced support option for VMWare. Broadcom has taken over and has basically told everyone except enterprise customers that they do not want our business.”

A January 2026, CloudBolt survey found that 86% of companies are now reducing their VMware footprint, though only 4% have fully migrated. Most are somewhere in between, working through their options at a pace that fits their environment.

In recent years, the technology landscape has been a whirlwind as AI reimagines nearly every longstanding process. Globally, teams are expected to navigate market shifts, drive innovation, and respond to sudden disruption, all while juggling competing priorities and proving ROI. Sounds easy, right?

For most, navigating the AI landscape is a challenge. With CFOs focusing on ROI, CIOs managing governance, technology leaders championing tech optimisation, and boards and investors prioritising outcomes, everyone has their own agenda and key metrics to march toward. Yet this growing list of competing demands has become difficult to manage cohesively, leading to increasingly rash pivots and continued missteps by large organisations.

As we continue riding the wave of the AI hype cycle, it’s imperative that businesses look beyond siloed success metrics and focus on the overall organisational impact. A market-wide reset is needed: one where leaders realign priorities and prioritise company-wide value. We’re entering an era defined by a new term: valuemaxxing.

Managing endpoint security for a remote team is one of the trickier problems in IT right now. With so many organisations now running fully remote operations, security teams are managing fleets of personal devices spread across home offices, public networks and sometimes even spanning continents.

For these people to be able to work effectively, all of these devices need to be connected to the same sensitive systems, leading many cybersecurity teams to prioritise the risks associated with these connections.

Today, about 26% of organisations now run on a fully remote model, according to Gallup. Each one of these workplaces carries a unique endpoint security challenge. Microsoft Intune has emerged as a popular fix for this, offering centralised control over devices, apps, and access policies from one console.

However, there’s another side to the picture. Intune’s licensing structure is layered in ways that often catch organisations off guard, usually after they have already committed. So here’s a clear look at where those surprise costs come from and how to get ahead of them before they hit your budget.

Many VMware customers assumed the most disruptive effects of Broadcom’s acquisition were already behind them. Licensing changes and pricing shifts required attention, but infrastructure strategy largely stayed the same. That assumption is about to be tested.

By October 2027, VMware customers must migrate to VMware Cloud Foundation (VCF) 9. What sounds like a routine upgrade has much larger implications. The timeline is compressed. Hardware and operational changes may be required. And organizations are being pushed to rethink platform strategy at a time when they are balancing modernisation, cost control, and new demands such as artificial intelligence (AI) and cloud-native development.

Infrastructure transitions rarely happen quickly. Moving workloads, retraining teams, validating interoperability, and maintaining resilience all take time. When migration is mandated rather than optional, the risk profile shifts. CIOs must act within a certain window of time while still supporting critical systems and ongoing initiatives.

Cloud spending continues to rise, but waste remains a stubborn issue. Global spending on cloud infrastructure services reached $102.6bn in the third quarter of 2025, growing 25% year on year as enterprises moved from early artificial intelligence experiments towards large-scale deployment.

At the same time, priorities are changing. As cloud environments grow in scale and complexity, organisations are placing greater scrutiny on how infrastructure spending is managed and controlled. The market for FinOps tooling has grown in parallel, but more choices do not always translate to clarity.

The question for 2026 is not whether a company has a FinOps tool, but whether that tool is fit for purpose.

Software makes the world go round, and more often than not you have to pay a pretty penny for the biggest and most popular software packages. Which is more than a little ironic when you consider that just about any software you pay for today is built at least partly on the work of free and open source developers.

Enterprises expected SaaS to simplify their IT infrastructure. In many ways, it did – but it also created an entirely new attack surface. Most departments now rely on a mix of sanctioned and unsanctioned SaaS applications – and even the sanctioned ones are often unmonitored for risk. This uncontrolled growth has evolved into a new form of Shadow IT: SaaS sprawl.

The rise of AI-powered SaaS tools has accelerated this trend even further. Users can now connect generative AI platforms to business applications like Google Drive, Slack, Microsoft OneDrive, CRM systems, and developer tools with just a few clicks – often without understanding the permissions being granted or the data being exposed.

While most organisations have adapted their identity and network controls for cloud usage, they are still unprepared for what has emerged behind the scenes: SaaS-to-SaaS integrations, AI-driven automations, OAuth permission chains, risky third-party plug-ins, and misconfigurations inside sanctioned apps.

The result is a scenario where security teams have minimal visibility and even less control.

Leading software development company Codehesion has identified a major industry shift where companies are moving from using off-the-shelf SaaS subscriptions, to building their own custom software.

“Businesses outgrow SaaS when software stops being a tool they use and becomes the engine their business runs on,” said Codehesion founder and CEO Hector Beyers.

“At that point, customer software isn’t a cost – it’s a competitive advantage. You’re building something your competitors can’t buy off the shelf.”

AI’s ability to accelerate delivery timelines has made the case for going custom even stronger.

“With the right partner, you get purpose-built technology at a speed that would have been unthinkable even a year ago,” said Beyers.