I’m of the opinion that most organizations have to revisit their decision-making process for making technology and AI investments. Some organizations are too slow, seek consensus with too many people, and negotiate low-level contract points. Others shoot from the hip by commissioning meaningless POCs, then buying more than they need because the vendor is offering a great deal.

For the sake of this article, I’m going to assume that sponsors have defined a vision statement around the investment. Technology and AI investments are initiatives that require defining target customers, value propositions, and strategic value. In the StarCIO one-page vision statement template, we ask for twelve clarifications before a small team starts agile planning their solutions.

During this planning process, questions arise about whether solutions should utilize existing platforms and when to replace them with new technologies. AI is driving a big part of these questions, i.e., should the organization use AI from existing platforms or procure new AI capabilities. StarCIO’s workshop on the 10 Steps for Technology Selections is a guidebook that organizations adopt when they desire a fast, smart, and compliant process.

When new technology investments are under consideration, I’ve seen organizations follow strict procurement-guided stage gates that take too long. Others don’t have a clearly defined process, and so teams don’t know what’s required of them.

How CIOs are evolving their technology selection process

Whether it’s a data security automation platform, a unified AI search capability for developing AI agents, or new AI customer service capabilities, CIOs are finding smarter ways to lead technology selections. Instead of spreadsheets tracking hundreds of nuanced features, they ask agile technology selection teams to specify outcomes, then work backwards to contrast the relevant technology capabilities in each platform under consideration.  

“CIOs must shift from feature-led evaluations to outcomes-driven selection,” says Jaime Meritt, chief product officer at Verint. “Prioritize technologies that deliver measurable business value – improved efficiency, stronger security, or enhanced customer engagement. Adopt agile procurement models that support rapid pilots, usage-based pricing, and fast iterations.”

Focusing on outcomes helps agile selection teams address a second issue – the hype, marketing, and sales promises that may be leaps beyond actual capabilities.  

“Modernizing how we prioritize digital and AI investments means moving beyond the hype and toward a more strategic, value-led approach focusing on real business outcomes that responsibly impact people,” says Sean Wechter, CIO at Boomi.

Avoid finding better versions of how things work today

A third consideration is when technology teams focus their selections on finding better technologies to perform today’s workflows. That mindset leads to technology transitions instead of transformational results. The approach also doesn’t work when evaluating new capabilities, such as AI agent development platforms.

“Executives today must embrace an agentic transformation that embeds intelligent AI agents into all business functions,” adds  Wechter. “The results can be so much more than just potential ROI. It opens up the possibilities of changing everything that fuels a company’s upward trajectory, including better decision-making, scalability, cross-functional alignment, and positive ethical considerations.”

CIOs seek “and” technologies: AI capabilities and security

Let’s recap: Agile selection teams should

• Define outcomes for a future state – not just to improve today’s workflows.
• Explore platforms for outcomes, not features.
• Let outcomes guide them in optimizing which capabilities to review and zero in on the required flexibilities.
• Develop meaningful POCs with acceptance criteria.

But CIOs aren’t stopping there because they know innovation without security leads to problems. It may create compliance gaps and security risks when deploying to production. Or, it may lead to the need to procure additional technologies to address these gaps.

“When it comes to AI, CIOs need to think about managing innovation and responsibility simultaneously, how the technology advances, and how to secure it every step of the way,” says Mohit Kalra, CISO at Typeface. “The pace of technological change today demands a fast, cross-functional approach to platform selection. CIOs should empower teams to pilot new solutions, while upholding core principles of security and compliance, to prove business outcomes and develop more effective AI evaluation frameworks before full deployment.”

Transforming procurement into a faster process

These suggestions may sound promising to CIOs, architects, and business sponsors who recognize the importance of leading faster evaluations. However, this raises a concern, as they are aware of the challenges in driving change within the procurement team, their processes, and compliance procedures.

Miles Ward, CTO of SADA, says to accelerate tech testing, especially in AI and security, CIOs need to empower small, agile teams to run rapid experiments. “Tell procurement that fast tests, not slow evaluations, is the name of the game, and partner with them to push vendors for easy terms, free trials, and deep support for your initiatives.”

Whether you’re seeking a low-code platform or a data analytics solution to structure data science programs, the agile selection team will want to demonstrate the solution’s scale. I’m not talking about scalability (which is important) – I’m referring to showcasing multiple business needs, use cases, or stakeholder groups that require a solution. Procurement, like IT, has its issues with point solutions and SaaS sprawl.

“CIOs must show procurement that investments offer comprehensive, strategic capabilities that satisfy multiple needs and don’t offer a mere point solution,” says Chris Joynt, Director of product marketing of AI security at Securiti.

To accelerate the process without overwhelming people or cutting back on key requirements, organizations should separate evaluation from procurement into two phases:

• Evaluation – The vendor is reviewed for meeting financial viability, information security, and other compliance requirements needed before tackling any POCs or initial spending.
• Procurement – Completing the full procurement process only when a vendor’s technology will be used in operational pilots or in production.

What due diligence is non-negotiable?

Whether you’re in evaluation or in procurement phases, the process, deliverables, and minimal due diligence requirements need definition. I begin with non-negotiables in DevSecOps and data governance, which I have written about in previous articles.

Here are two other considerations

• “When it comes to due diligence, we absolutely must see demonstrable integration with existing systems and ironclad security models. If it can’t plug in and it can’t be trusted, it’s a non-starter,” adds Ward of SADA.
• “The most important aspect of due diligence is to validate a vendor’s alignment with emerging AI Security frameworks such as NIST AI RMF, OWASP Top 10, and HITRUST,” states Joynt of Securiti.

I suggest CIOs partner with CISOs, procurement, legal, and other risk management functions to define non-negotiable by phase (i.e., Evaluation, Procurement). The nice thing about using the term “non-negotiable” is that it leads to a highly prioritized list, and what’s not on the list is negotiable!  

Smarter technology selections: Concepts to practice

I’ve shared a few concepts here in this short article. For additional concepts, members of the StarCIO Digital Trailblazer Community can listen to the recent Coffee With Digital Trailblazer episode on Modernizing How Executives Prioritize Digital/AI Investments.