AI is reaching more of the business, and investment plans are growing. Leaders want to know what value will come from that commitment. The answer starts with how we develop and evaluate each idea.
“We should use AI for this.”
If you lead an AI team, you’ve probably heard that a few times this week. It usually arrives with plenty of enthusiasm and rather less detail about what “this” actually involves.
That enthusiasm is welcome. People across the business can see opportunities to improve how they work, and you want them to bring those forward. The challenge is that each idea needs some work before your team can decide whether it deserves more of their time.
Someone needs to understand the problem, find the data owner, clarify the expected benefit, and work out what would have to change. Often, that someone is already on your AI team, with a delivery backlog and a calendar that appears to have lost the concept of empty space. A promising suggestion can turn into several meetings before anyone is ready to evaluate it.
More AI. More investment. What are we getting?
The research puts some numbers behind the pressure. McKinsey’s August 2026 survey found that 44% of respondents said AI was scaling across their organization, up from 38% a year earlier. BCG’s AI Radar 2026 found that surveyed companies planned to roughly double AI spending as a share of revenue, from 0.8% to 1.7%. Those are spending intentions, but they give a clear sense of the commitment being made. McKinsey; BCG.
The value story is still catching up. In the same McKinsey survey, 80% reported improved personal productivity, but 37% reported a positive effect on company-wide earnings before interest and taxes, broadly unchanged from the previous year. That doesn’t mean the rest are getting no value: better service, innovation, and benefits within individual functions also matter. It does explain why a growing list of AI initiatives may leave the leadership team asking for a little more than a very busy progress slide. McKinsey, 2026.
For an AI leader, that adds another job to an already full calendar. Alongside deciding what can be built, you need a credible explanation of what it should improve, what it will cost to deliver and run, and how the business will capture the benefit. That explanation is much easier to develop while shaping the idea than after launch, when someone asks for the ROI and everyone suddenly remembers another meeting.
Help people develop the idea
The person raising the idea probably knows something useful about the problem. They deal with the process, see where it breaks down, or spend Friday afternoons doing something they’re fairly sure a computer should be doing. What they may not know is how to turn that experience into a case your team can assess.
A longer submission form only helps if people know how to answer it. Asking someone to “quantify the business benefit” can produce a thoughtful estimate, a blank box, or a very confident number whose origin nobody wants to discuss. We need to help people work through the questions, including where they don’t yet have an answer.
Take an AI assistant for client briefings. Before discussing the solution, it helps to understand how briefings are prepared today, which information people use, and where the effort goes. That gives the business a starting point for explaining what the improvement would be worth, alongside the cost of building, running, and adopting the solution.
That’s where we see KangaROI helping. Our AI assistant, kAI, guides people through developing the idea, clarifying expected benefits, and structuring the case. The aim is to give the AI team a clearer starting point, with assumptions and gaps visible, rather than quietly filling them with optimism.
Give reviewers a case they can compare
Once there’s enough context, the team can spend its time on the questions that need its expertise. Is the approach feasible, is the data ready, and does the expected benefit justify the effort? Technical, security, and responsible AI assessment still need specialist judgment, however neatly the submission is written.
A shared structure also helps when several ideas are competing for the same resources. One team might be talking about hours saved, another about customer experience, and a third about something a senior executive saw in a demo. All three deserve a fair assessment, even if only one arrived with the word “urgent” in capital letters.
KangaROI brings the investment information together so reviewers can consider value alongside costs, effort, readiness, and the business’s ability to act. A useful outcome might be approval, a request for specific information, or a decision to wait. Recording the rationale gives the person who submitted the idea something more helpful than watching it disappear into the backlog.
Then find out where the saved time went
Approval should leave the delivery team with a clear explanation of what it’s trying to achieve. The problem, expected benefit, and success measures need to survive the move from business case to project. That gives leaders something concrete to review alongside adoption and spending, instead of trying to reconstruct the value story at the next budget meeting.
Go back to the client-briefing assistant. Suppose it reduces preparation time, which is a useful result. Before turning those hours into a financial saving, the business needs to understand what actually changed as a result.
Perhaps people spend more time with clients, respond faster, or need less paid overtime. Perhaps they simply finish their existing work more comfortably, which can still be valuable. It just shouldn’t become a cost saving because someone multiplied hours by a salary and the spreadsheet looked pleased with itself.
This is why we keep benefit owners, actions, forecasts, and evidence alongside the case in KangaROI. The business can review what’s happening, explain the gaps, and decide what action to take. That gives leaders a better basis for expanding an initiative, adjusting it, or directing resources somewhere else.
Make the next idea better informed
The results also tell you something about the next investment. Adoption might have taken longer than expected, a data dependency might have been more significant, or the benefit might have appeared in a different part of the business. Those lessons are useful when the next enthusiastic request arrives, and it probably won’t keep you waiting long.
Bringing that experience back into planning helps teams ask better questions and use more realistic assumptions. It also makes the work of developing a case more valuable over time. You’re building an understanding of what works in your organization, including the changes needed to make it work.
Give good ideas a clearer starting point
For us, the value conversation starts with the work your AI team does before it can evaluate a request. Helping the business develop that context earlier gives reviewers a stronger case and leaders a clearer expectation to follow through delivery. If your team is managing growing AI demand while being asked to demonstrate the value, let’s chat about how KangaROI could help.
Get in touch with KangaROI
