96% of Companies Using AI Report "Little to No" ROI
Inadequate C-level leadership and strategic planning are cited as primary causes. Investment in AI will continue to increase as new paradigms mature.
When I first saw the headline above, my immediate reaction was “Wow. Now that is a rather misleading Web 2.0-ish headline. Either we have fallen for the most well-orchestrated trillion-dollar scam in human history and AI is really vaporware. Or, there are widespread shortcomings in the strategic leadership and knowledge needed to effectively leverage AI and its various frameworks.” It would take a lot of effort to produce a near-100% failure rate on return on investment (ROI). Let’s go with the latter and add some context.
Common traits contributing to unsuccessful AI initiatives include:
Lack of AI strategic leadership and planning across the enterprise. For example, “AI First” doesn’t happen by simply announcing it at a meeting. It’s a mindset and approach that requires ongoing support and commitment from key stakeholders.
Avoid conducting deep research on AI frameworks, resource allocations, KPIs, and case studies to meet defined goals.
Having the wrong stakeholders at the table, working with unclear goals and objectives that are not aligned with the overall business.
Unwillingness to pause an initiative (in progress) to recalibrate the project plan, risk, and trajectory as “stress fractures” begin to appear.
Ignoring Albert Einstein's words: “We cannot solve our problems with the same level of thinking that created them.”
Yes. Artificial Intelligence (AI) frameworks are rapidly evolving, and some tools are unreliable, ineffective, or simply benign. However, many AI breakthroughs are truly transformative and astonishing, influencing every stratum of our society. Given this new reality, we cannot blame technology alone for failures without also recognizing the strengths and limitations of humans in generating immediate, widespread ROI success.
A significant lack of essential AI knowledge and leadership skills is a big challenge. Two widely publicized reports, The GenAI Divide: State of AI in Business 2026 (MIT) and PwC 2026 Global CEO Survey, show that up to 96% of companies have not yet achieved significant ROI. However, 94% of CEOs plan to continue or increase funding, viewing AI as a long-term strategic priority rather than a quick profit.
Primary Causes of Negative ROI
The lack of ROI in AI is rarely due to the models themselves. Instead, it is often the result of systemic structural, and organizational challenges:
The Data Foundation Crisis: 73% of data leaders identify data quality and fragmentation as primary barriers.
“The Perpetual Pilot: So, Close. Yet So Far.” Approximately 80-85% of AI initiatives fail to reach deployment. Many start as “standalone” prototypes that cannot scale or lack a clear way to fit into existing enterprise workflows. Hmmmm... one might say that “Building prototypes should be formal parts of application design and development.” Stakeholders can then “kick the tires” on key assumptions, timelines, and functionality in a safe, stable environment. The problem with prototyping is:
Stakeholders may not understand the “cause and cascade effect” of an issue and move on to something easier to fix;
Continuous changes to the project scope are not reflected in the final prototype design, leading to conflicts where they shouldn’t be.
Who likes delivering bad news? Nobody. But failing to inform your team of an issue and to seek help understanding its cause or limitations before a prototype goes live is, at the very least, your professional responsibility. At worst, it could lead to a pink slip when things go “boom.”
The Productivity Paradox: While 96% of execs expect productivity gains, 77% of employees report that AI has actually increased their workload due to the time required for reviewing AI output and learning new tools.
Workflow Misalignment: Only 21% of organizations have redesigned their business processes to align with AI capabilities; most simply “overlay” AI on top of inefficient, legacy processes.
What the “Successful 5%” Are Doing Right
The small number of companies seeing measurable ROI shares specific traits:
Hyper-Focus on Pain Points: They steer clear of broad “AI for Everything” strategies and instead focus on one high-value, specific business bottleneck (e.g., reducing vendor disputes or automating invoice processing).
Customization Over Commoditization: They move beyond generic “out-of-the-box” bots, building customized systems that learn from company-specific data and adapt to internal workflows.
Strategic Partnerships: Successful firms are twice as likely (67% vs 33%) to use external partnerships for developing solutions instead of relying only on in-house development.
Back-Office Prioritization: Prioritize AI efforts on unglamorous back-office automation where ROI is easiest to measure, instead of focusing on high-visibility marketing or front-end tools.
In closing, I would love to hear your thoughts and feedback on this topic. I am personally a big fan of prototyping, and I believe it is one of the most essential (often overlooked) phases of successful AI initiatives.