The CIO Playbook That Worked For 20 Years Is Breaking
Sue Odle is CEO of StorMagic, where she helps organizations simplify on-site virtualization for distributed IT environments.
gettyFor years, infrastructure planning was predictable. CIOs knew their vendors, understood licensing models and could plan around standard hardware refresh cycles. That predictability has changed as hardware costs and availability fluctuate, vendor models change and AI creates new infrastructure demands while organizations continue to manage centralized and diverse distributed IT operations.
Many CIOs are slowing down, not because they lack urgency but because they understand that the way IT planning has been governed for the last 20 years is no longer reliable. The old playbooks assumed a level of stability that many organizations don’t have anymore.
Hardware is one of the clearest examples of how infrastructure planning has changed. Many organizations historically followed a three-to-five-year refresh cycle and treated it as a standard part of IT planning. Now, CIOs have more reason to question whether every scheduled refresh is really necessary.
Hardware is more dependable today, while costs and supply are harder to predict. According to a July 2026 report from TrendForce, server DRAM contract prices were expected to rise 13% to 18% quarter over quarter in the third quarter, while there was still a hardware shortage. As a result, some organizations are extending the life of existing hardware and looking more closely at what they own before committing to a replacement.
That decision isn’t only about preserving budget. Hardware cost has a trickle-down effect on modernization plans, software choices and the pace of innovation. It’s a risk factor, not just a line item. Refresh cycles should increasingly be driven by workload and business requirements rather than simply by the calendar. If existing hardware still supports the business, keeping it in place often makes sense.
When conditions are difficult to predict, long-term commitments can carry more risk than they once did. A longer agreement may create cost certainty, but it can also limit flexibility if pricing, licensing, support or business needs change before the contract ends.
That doesn’t mean long-term planning should disappear. It means CIOs need to be more deliberate about where the business can comfortably make a long-term commitment and where a shorter scope gives the organization more room to adjust.
Vendor history, stability and behavior are now a part of infrastructure risk. CIOs aren’t only evaluating whether a platform works today. They’re also evaluating whether the licensing model, support experience and vendor relationship can be trusted over time.
Cloud-first strategies have shaped IT modernization for years, but a one-size-fits-all approach is becoming harder to justify as infrastructure and workloads spread across more locations. The cloud remains important, but some workloads need to operate closer to where data is created and where the business is running.
Retail locations have different requirements than centralized data centers. Many retailers run transaction, inventory and security applications locally, where connectivity and onsite IT support vary. Relying entirely on centralized infrastructure can increase operational risk.
Healthcare, manufacturing and other distributed industries face similar challenges. Healthcare organizations may run systems locally to support patient care, while manufacturers may keep applications onsite to maintain production when connectivity fails. Latency, availability, regulatory requirements, data sovereignty and data movement costs can determine whether workloads run in the cloud, at the edge or across both.
This is where CIOs need to be careful not to apply data center assumptions to edge environments. The edge isn’t simply a smaller data center. It has different constraints around space, staff, connectivity, cost and uptime. Treating those environments the same can lead to unnecessary complexity or infrastructure that exceeds what the location actually needs.
AI is creating new pressure on infrastructure decisions, but AI readiness isn’t only a GPU issue. Before committing a significant budget to new compute resources, CIOs need to understand where AI will be used, what data it depends on and whether the workload needs to run centrally, at the edge or both.
Some AI use cases may require specialized infrastructure. Others may be supported by existing CPU-based servers, especially in early-stage or edge environments where the objective is to process data closer to where it’s created. The point is to avoid assuming every AI use case requires the same infrastructure strategy.
CIOs should connect AI infrastructure planning to data maturity, workload placement, governance and operational requirements. AI investments should be tied to what the business can truly use, support and scale.
There’s still pressure on CIOs to move quickly as the C-suite pushes for modernization, AI efforts and measurable returns. Taking more time can be interpreted as hesitation, but a slower decision isn’t always a weaker one.
CIOs are weighing rising costs, changing vendor models and infrastructure needs before committing resources. Some projects can move forward now, while others may be better phased or delayed.
CIOs also need to explain that reasoning to the rest of the C-suite. If a delay reduces unnecessary cost or operational risk, that context matters. Careful sequencing can protect the business from overcommitting to technology, contracts or hardware before the requirements are clear.
CIOs need to build more flexibility into the decisions they’re already making. That means looking more closely at whether existing hardware can continue doing the job before refreshing it, choosing vendors and contracts that leave room to adapt as conditions change and matching infrastructure to the workload rather than defaulting to one environment.
The old playbook assumed more predictability than CIOs have today. Hardware cycles, vendor models, cloud strategies, data sovereignty and AI requirements are all changing at once. CIOs can’t control what the market does next, but they can build infrastructure strategies that give the business more room to adapt.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
