CIOs need to take back control of their IT roadmap

Long-term enterprise technology planning is changing. Beset by market disruptions and rising operational costs, CIOs are being asked to make tough, high-consequence decisions faster than ever before.

Boards want to drive AI, but don’t always appreciate the difficulty in delivering value from it. Yet, the biggest obstacle to enterprise agility might just be the major vendors who are trying to set the pace and write the rules of this new landscape.

At the recent Gartner IT Symposium/XPo in Barcelona, James Harvey, CTO at enterprise software support company Rimini Street, told IT Pro that CIOs are currently being asked to do more with less.

“They're looking to simplify their environments, consolidate, transform, and automate securely,” he said. “They're looking at less lock-in from large vendors. They want flexible partners.”

As the enterprise world becomes more fluid, Harvey says there are several trends pushing organizations away from traditional service integrators (SI) to SRE and DevOps, as “more leaders are doing tech themselves rather than buying it”.

He says CIOs are also struggling with how to adopt AI amid executive pressures to drive innovation and value. “[CIOs are] seeing a lot of siloed and shadow AI in their organizations that isn't really delivering much value”, he says, where they should be linking it to enterprise processes.

Surging costs are also a factor, says Martyn Hoogakker, GVP and theater GM, EMEA, at Rimini Street. He says rising energy prices are putting pressure on businesses to update their IT estate, while many spend a lot of their budget on just maintaining what they have.

Sign up today and you will receive a free copy of our Future Focus 2026 report - the leading resource for IT decision-maker insight on priorities and investment areas in AI, security and more.

“At the same time, they're being forced to make decisions quickly on security, ERP, digital, all those areas. And they're struggling to make decisions,” he tells . “Do they keep the old estate, do they invest in huge upgrades with their existing platforms, or are they going to go best of breed across ERP, digital, security, or AI?”

He says these are big decisions to make, which aren’t being made any easier by the fact that SIs, who have their own interests, aren’t giving out the best advice.

Overcoming vendor squeeze

There’s an ongoing myth that companies can only adopt AI within their vendor’s roadmap and that modern AI can’t be built onto legacy software, which “just isn’t true”, says Hoogakker.

While it can feel like companies are locked in to certain platforms due to the large investment and approvals needed to build an ERP, for example, the case for staying with a single vendor is becoming less clear as business needs change.

“The business case is not very clear because no one knows how expensive the migration will be,” says Hoogakker. “There's the custom code, there's no view on what AI is embedded in the new functionality, what token costs are involved as well, and what they're going to have to pay for all the API calls in a new estate.”

He noted that there’s more pushback from customers now around sticking with “monolithic vendors”, especially as new startups are quickly taking over certain areas of the market.

This is particularly visible with VMware, for example, whose steep price hikes last year sent customers into panic.

“If you look at the composable architectures people are now deploying in the application space, you're no longer getting fully packaged applications,” Harvey noted. “You're getting much more architectures around containerized microservices, delivering multiple capabilities to individual outcomes.”

The infrastructure landscape is changing, he said, adding that “it’s giving people much more flexibility around where they park different workloads within their stack”.

“Things are becoming much more fluid. The optionality out there both in infrastructure and software is compelling, without getting that vendor lock-in, which is going to be a struggle for some of the vendors that don't innovate quickly.”

Avoiding disruption

For most companies with large ERP estates, upgrading to the next release offers relatively little additional functionality despite the huge undertaking a single upgrade could take.

The process could mean a complete reimplementation, taking three to five years to complete and tying up resources that could otherwise be spent on innovation.

“It’s disruptive,” says Hoogakker.

“You have to set aside a huge team to do the custom code, maybe some self-support, you have to hire an SI to do everything, manage the SI, and hopefully, it's going to happen in the next three to five years,” he explains. “And you have to do it because they're going to stop the support.”

Even when an existing platform is stable and working well, businesses can find themselves having to invest in an upgrade or migration simply because the vendor is withdrawing support.

Harvey argues that CIOs need to look instead at their own technology needs across the board, and consider where targeted improvements could deliver genuine value rather than replacing entire core systems.

“It’s about turning a technical debt into a technical asset,” he says, adding that software doesn’t need to be ripped up and replaced every time.

When he speaks to CIOs about ERP or enterprise software, he says he always asks three key questions: Why do anything at all? Why do it with that particular company or product? Why do it now? “I find that seven to eight times out of 10, they can't answer all three questions. So we go back through the cycle,” he says.

“It's about taking an outside-in approach. So building the technology solutions and stack from business outcomes back through the technology, rather than starting at the base in the technology.”

Rather, businesses will find they can actually orchestrate over the top of existing functionality, enabling them to redo processes, sort out data, and build an agentic AI capability to give them that new extensible integration layer perspective, without spending huge amounts replacing core stacks.

“That is transformation without disruption,” says Harvey.

The ‘AI last’ approach

There are many parallels between current AI adoption and the rush to go cloud-first a few years ago. At the time, many organizations moved their systems without properly considering whether doing so would deliver genuine business value, according to Harvey.

“We saw what happened when people went cloud first, costs went through the roof,” he explains. What’s clear is he’s now seeing a lot of retrenchment from cloud because of data sovereignty issues and geopolitical challenges.

For Harvey, the issue is just as much about governance as anything else. Rather than starting with AI, he argues organizations should be addressing underlying problems with their data, processes, and governance, in order to identify where the technology can actually deliver genuine business value.

“If you deploy AI in a dysfunctional organization on top of bad data without a clear strategy, it's a waste of money and a waste of time,” he says. “And it's just a race to the bottom.”

Rimini Street’s focus is on helping customers with that governance question. Harvey says the company has just launched Rimini Govern, which he says allows customers to lock down agentic AI within their enterprise to ensure non-human actors have the same controls as human actors. Harvey believes this capability is something any company should be able to provide as the space evolves.

“I think that's where a lot of CIOs have got stuck where the board have said ‘we need AI, AI, AI’ and they've just deployed AI, but they can't tangibly show why or what benefit. This is why we're taking that AI last approach rather than first, because you have to do some of those foundational things to make it give you that return on investment.”

“If you understand what the business outcome is, you break down siloed deployment of AI, and you understand what you're trying to do with it, then it will make some great leaps forward for you.”

Original source CIOs need to take back control of their IT roadmap

Back to home