A week of outages gives CIOs a window to fund resilience
- Madeleine Streets
- 858 views
Recent Microsoft, FAA, Workday and JPMorgan Chase disruptions show how tightly systems now connect. CIOs have a short window to argue for engineering time and resilience.

Line up the recent disruptions and the common thread isn't one bad change. It's how tightly everything is connected. A Microsoft outage knocked out Outlook and Teams for more than a day. A Federal Aviation Administration network failure grounded flights across the country. A Workday disruption cut off access to HR and payroll systems, and JPMorgan Chase customers ran into banking delays. In a few days, several unrelated systems failed, and every CIO now has the same uncomfortable question: how much of the business stops if one provider does?
The reported causes differ, and investigations so far have found no evidence linking the incidents to each other or to cyberattacks. That absence is exactly what worries Lisa Palmer, CEO of AI firm Neurocollective and a former CIO. "No one has tied these outages to an attack, and that should worry CIOs more," she said. "Routine failures now hit thousands of companies at once, while AI makes real attacks faster every month."
Why the attention is worth using
The outages made technology dependencies, infrastructure decisions and operational weak spots visible to business leaders and customers who rarely think about them. That's useful. Budget requests for resilience usually lose to AI investment and cost-reduction targets, and a week like this one changes who's listening.
Here's the catch: the window is short. Attention fades quickly once services return, so the work of turning that attention into a plan has to start now, not at the next planning cycle.
Staffing and the knowledge that walks out the door
Technology companies have announced extensive layoffs this year, including cuts to experienced engineering teams. There's no evidence that staffing reductions caused any of these outages, and tying individual incidents to those decisions would be speculation. The loss of institutional knowledge is still a legitimate concern, though.
Palmer put it plainly. "Those are the people who know how the systems really connect: which application depends on which network, and what breaks when you change it. That knowledge rarely lives in documentation. Outages tend to start in exactly those hidden seams."
Tomás O'Leary, founder and CEO of enterprise software maintenance provider Origina, made a related point. Experienced engineers understand why systems were configured the way they were, where dependencies sit and what goes wrong when something changes. If those engineers leave before that knowledge has been documented, transferred and tested, the organization loses operational understanding it may not be able to rebuild.
Engineers pulled toward vendor timetables
Retaining expertise is only half the problem. Origina research found that 84% of surveyed senior IT decision-makers said experienced engineering talent was always or often diverted toward vendor-driven upgrades rather than building future capabilities. That creates a hard trade-off. The engineers who remain are scheduled around external product timelines instead of the internal work of mapping dependencies and fixing operational weak points.
This is where CIOs have an argument to make. Engineering expertise and time are operational requirements, not just line items on a staffing chart. That doesn't mean more headcount would have prevented this week's failures. It does mean showing leadership what the organization risks when too few people understand its systems and the people who do are busy with work that does little for operational control.
The help desk is the wrong place to start
Customer support will be the first thing people notice when services fail. During an outage, customers need accurate information, realistic expectations and a way to keep essential work moving. If support has been cut recently, that gap will be felt sharply, and CIOs can use it to justify renewed investment. But expanding the help desk alone won't fix the underlying problem.
Palmer was blunt about it: "Hiring more help desk staff helps you apologize faster, [but] resilience lets your customers keep working while you fix the problem." She pointed to communications plans prepared in advance, status pages hosted outside the affected platform, and manual workarounds for essential processes such as payroll, payments and flight operations.
Bernadette Nixon, CEO of network automation company NetBrain, made the same argument from the infrastructure side. A bigger help desk won't shorten the next outage or make it less likely. The stronger investment sits upstream: making changes safer, finding root causes faster and spotting weaknesses before they cause disruption.
O'Leary has seen the pattern in his work with enterprise customers. Recommendations from earlier outage reviews often go unimplemented, so the same underlying weaknesses cause trouble again. "Otherwise, we're just simply becoming better at apologizing for the same problems," he said. The practical fix is corrective work driven by technical findings, with clear ownership and follow-through, rather than a bigger support queue.
Making the case to the board
Palmer's experience is that boards tend to focus on resilience for a few weeks after a high-profile outage before moving to the next priority. She recommends arriving with a concrete picture of exposure: the single vendors and aging systems that could stop the business, the estimated cost of an hour of downtime, and a plan to reduce that risk. A map showing how payroll, payments and customer logins depend on the same provider will persuade more people than a slide full of risk percentages.
"Deferred maintenance isn't savings. It's a loan, and outages are how the bill comes due," Palmer said.
For some organizations, the case will be for more infrastructure investment or experienced engineers. For others, the first move is to look at how existing technology budgets are spent. Origina's research found that, on average, 28% of enterprise software budgets goes to vendor-driven upgrades, migrations and licensing changes that would not otherwise have been chosen.
"Every unnecessary upgrade consumes engineering capacity and introduces operational risk," O'Leary said. Before approving a change, teams should ask whether it genuinely improves the business or simply funds a vendor's schedule. Independent maintenance and validated interoperability can sometimes offer an alternative to an immediate, expensive upgrade.
Nixon's advice is to let the numbers carry the argument. What does an hour of downtime cost? Which changes would have the biggest impact? How quickly can the team identify a cause? "The pitch that works is not, 'We need more budget,'" she said. "It's, 'Here's the exposure, here's what it costs us and here's how we shrink it.'"
Those arguments won't lead every organization to the same decision. Some need to modernize infrastructure, others need to free up engineering time, and others need to scrutinize vendor commitments more closely. My read is that the teams who come out ahead will be the ones who walk into the next budget discussion with the dependency map already drawn and the cost of downtime already calculated.