Questions To Ask To Better Understand Nontechnical Stakeholders’ Needs

What Problem Disappears If This Succeeds?

It’s critical that technical leaders start with the question: “If this works perfectly, what stops being a problem?” The goal is to move from a preselected solution to a deep understanding of the problem and who benefits. Otherwise, you ship something that technically works and changes nothing, which is exactly how organizations end up with a pile of AI pilots and no outcomes. – Kathryn Harrison, Concentrix

What Are The Expected Outcomes And Goals?

The question to ask is: “What are the expected outcomes and goals to be achieved?” For example, is it faster processing, lower cost, new features or something else? This helps them to focus more on the outcomes to be achieved rather than the tools available. The solution to achieve the outcome may or may not leverage AI, as there may be better ways of achieving the same objectives with a simpler solution that is faster and cheaper. – Naveen Jayaram, UniCredit

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?

What Work Is Piling Up, And Why?

Tech leaders should ask, “What work is piling up, and what input or coordination bottleneck is causing it?” AI can speed up outputs, but the real business need is often upstream: scattered inputs, slow handoffs, unclear ownership or communication overhead creating backpressure. – Harsh Singhal, Glean

Is The Goal Automation Or Reinvention?

The most important question is whether the stakeholders want to automate the current process or reimagine it for a future that accounts for work being performed by agents while keeping humans in the loop. An inefficient “automated” process will not change the outcome; a reimagined and agentified process will. – Manoj Mishra, Deloitte

Has The Underlying Process Been Optimized Yet?

Technology leaders often ask, “What business problem are you trying to solve?” for good reason. A clearly defined problem helps identify the right solution. However, it is equally important to determine whether the underlying process is already optimized. Applying AI, automation or digital transformation to a flawed process often accelerates inefficiency rather than creating value. – April Ho-Nishimura, Infineon Technologies AG

What Result Must Change, And How Is It Measured Now?

Ask, “What result are you hoping to change, and how do you measure it today?” People often ask for AI or automation when they really just have a problem they haven’t clearly named. This question gets past the buzzword to the real need, sets a clear and measurable goal and often shows that a simpler, cheaper fix would work better than the technology they first asked for. – Luboslava Uram, Solvd Group

Is This The Right Technology To Achieve The Goal?

The key is to identify the goal we want to achieve through AI, automation or digitization. The next step is to verify whether these technologies can make it happen. For example, AI in the form of LLMs (which is often what “AI” means today) is not suitable for solving all complex problems. Only composite AI employing symbolic and subsymbolic AI can do it, as it follows logic principles and mimics the human brain. – Filip Dvorak, Filuta AI

What’s Slowing The Decision You Want To Make Faster?

Ask, “What decision are you trying to make faster, and what’s slowing it down today?” Most transformation requests are really about a bottleneck someone is tired of working around. That question cuts past the technology hype and surfaces the actual workflow pain. Once you understand the friction, the right solution often turns out to be simpler than the AI-powered platform they came in asking for. – Kuldeep Chowhan, Walmart

Why Are You Asking For This?

Sometimes the most powerful question is the simplest one: “Why are you asking me for this?” It forces the stakeholder to articulate the real need behind the request, separates concrete use cases from hype-driven ones and tells you everything you need to know about where to go next. – Nino Letteriello, FIT Group

What Business Action Do You Want To Improve?

Ask, “What action do you want the organization to take faster, better or more consistently?” AI and automation create value when they improve execution. Understanding the desired business action often reveals the true opportunity behind the request. – Lori Schafer, Digital Wave Technology

What Is The Cost Of Being Wrong?

Ask, “What’s the cost of this being wrong?” the honest answer is “we rerun it,” AI as assistance is the right fit. If the answer involves customer harm, regulatory exposure or unrecoverable decisions, AI as autonomy is being requested at the wrong scope. The cost of error tells you the deployment model the request actually needs. – Andrew Siemer, Inventive

Will This Empower Employees Or Replace Them?

When nontechnical stakeholders request AI or automation, I believe tech leaders should ask, “Will this solution empower employees or replace them?” Many initiatives stem from repetitive tasks that limit growth. Digital transformation succeeds when technology complements people and equips them with new skills, rather than creating uncertainty or resistance to change. – Prashanthi Kolluru, KloudPortal Technology Solutions Pvt Ltd.,

Is AI Governance Already In Place?

Ask, “Is your AI governance baseline established?” Most requests surface after employees have already used ChatGPT unsanctioned. The real ask is governance: making shadow adoption visible and auditable. Without visibility, ownership, risk registers and change controls, you’re scaling rogue behavior with approval. The business need isn’t the tool. It’s the infrastructure to operate it safely. Fix the governance first. Then scale. You’d be surprised by the change. – Dan Sorensen, Nexus Security Advisors

What Evidence Supports The Need For AI?

Most AI requests are solutions in search of a problem. The strong ones arrive with evidence. Ask, “What signal do you already have that points to this?” It could be customer escalations, workflow data or manual workarounds that broke. If your stakeholder can’t name the evidence, they’re not ready for AI. They’re chasing a trend. AI doesn’t fix the absence of evidence. It amplifies it. – Varun J. Vincent, FalconFirst AI

Who Will Need To Change How They Work?

Ask who needs to change their work style if this works out, and if they have been involved in the discussion. Most transformation plans spot a problem and ignore the person causing it. AI won’t fix an approval process or a team that wasn’t asked for input. It just makes the issues cost more. The real business issue is usually a coordination problem. The honest answer to that question shows whether technology is the solution or just a distraction. – Maitrik Patel, Apple

Who Is Already Using AI Without Approval?

Ask who on their team is already using AI for this without telling you. Most “AI initiative” requests come months after employees have quietly adopted ChatGPT or Claude to do the same work. The real ask is usually governance to make the unsanctioned workflow safe, audited and scalable. This question surfaces the actual problem in one move. – Nidhi Jain, CloudEagle.ai

What Happens If We Do Nothing For Six Months?

A specific answer to this question—we lose this client, miss this deadline, burn X per quarter on manual work—means a real problem exists. Vague anxiety about competitors or board optics means you are building a solution for a problem that does not exist. We open every client engagement with this question. It kills more bad projects than technical discovery ever will. – Denys Vorobyov, EltexSoft

Who Is Accountable When This System Gets It Wrong?

Amazon CTO Werner Vogels built Amazon on “you build it, you run it.” The same test applies to AI requests: If no one will own the failures, no one actually believes in the initiative. Naming an owner before a line of code is written separates institutional commitment from enthusiasm. – Manas Chaudhari, Meta

What Business Assumption Are You Trying To Test?

Ask, “What assumption about your business are you hoping this technology will prove or disprove?” Most transformation requests are really attempts to validate a belief about customers, employees, markets or operations. This question surfaces the hidden hypothesis behind the initiative. Once leaders understand the assumption being tested, they can determine whether AI is the right tool, a simpler experiment would suffice or the premise itself needs rethinking. – Jagadish Gokavarapu, Wissen Infotech