No ROI with AI

January 13, 2026 / Leave a Comment

The hype around AI was that it was going to change how businesses succeeded “and it’ll be unthinkable not to have intelligence integrated into every product and service” (Sam Altman, CEO of OpenAI). The problem with this marketing was that it encouraged every leader in every type of business to rush to get AI into their tech stack in some way or form – regardless of whether it was actually solving anything. After the adoption of AI in businesses has doubled in 2025, we’re seeing reports from companies like McKinsey showing only 39% of companies are reporting EBIT value. With these kinds of results, it is clear that just onboarding AI into your business is not enough. You need to know what makes AI effective in your business and how to implement it to see true ROI.

What Causes the Lack of ROI

Using AI Only for Task Efficiency: improving employee task completion rates without business benefits tied to the top or bottom line.
No Clear Goal Defined: implementing AI for the sake of implementing AI without a clear goal for the use of AI tied to a problem or specific revenue or operating metric.
Poor Implementation: onboarding AI solutions, especially Agentic AI, without cleaning up data for input or providing business context for valuable outputs. Garbage in, garbage out.
Assumed Benefits Instead of EBIT Value Association: expecting benefits outlined by AI providers by just implementing cookie-cutter solutions instead of choosing solutions & implementing with EBIT value related metrics.

3 Ways to Get ROI from AI

The reality is that AI is not cost-efficient today because of the high cost of AI processing. But if you know how to implement it correctly and size the cost to the benefit accordingly, it’s worth it every time.

1. Understand the Problem

You need to know the specific problem you’re trying to solve and WHY an AI solution is applicable to solve it effectively. Not everything requires AI. Sometimes AI can add unnecessary noise to your tech stack, especially if the direction is not clear from leaders and aligned to the working level knowledge of how processes actually work.

2. Hire the Right Employees

Not everyone can implement AI because not everyone has the skillset required to implement it properly. You need people who know how to translate business context AND customer experience into technical requirements, while also understanding the data structures required for the each type of AI solution. Not everyone knows how to do this and it is the baseline requirement for AI algorithm inputs and prompts.

2. Match Your ROI to the EBIT Goal

Measure your success with an AI solution against the problem being solved with a direct to tie to EBIT value instead of expecting revenue to magically appear from AI implementation. Your KPIs need to show the extent to which the problem is solved AND how the problem being solved adds EBIT value – whether it’s direct topline value through innovation or bottomline savings through organizational efficiencies achieved.