AI IMPLEMENTATION / FIELD NOTE
The 95% question: how do you work toward the 5%?
AI can be added to a business and still fail to improve the work. The useful question is what your team will do differently, how you will check it and who stays responsible for the result.
What the report says
A July 2025 Project NANDA report described preliminary findings from more than 300 public AI initiatives, interviews with 52 organizations and surveys of 153 senior leaders. It reported that 95% of organizations in its research saw no measurable return from enterprise generative AI investment.
That figure belongs to the report's research and should not be read as a universal failure rate for every AI project. It is still a useful warning: adding a tool without a defined job, owner or check can create activity without business value.
Why implementations stall
Teams often start with a tool and work backward. Context is missing, the permitted tools are unclear, rules are left implicit and nobody has a loop for checking what happened after launch. The system may look busy while the original process stays just as slow or uncertain.
- ContextGive the system the facts and boundaries the work needs.
- ToolsChoose the smallest set of tools that can do the job.
- RulesKeep review, permissions and exceptions visible.
- LoopsCheck the result and improve the process after use.
How ALTDEL works toward the 5%
We start with one repeat process and make the next decision clear. CTRL Blueprint maps the work and tests the fit before a build. CTRL Build turns the approved plan into a system with known checks. AI education and team training gives people a practice task and a guide they can return to. CTRL Ops reviews what happened after use and keeps improvements inside the agreed scope.
The goal is not to add AI everywhere. It is to show a useful change, keep a person accountable and save enough of the work that the next improvement starts from evidence.
Start with one process
Working toward the 5% begins with a job your team already repeats and a clear definition of “better.” Tell us where the work gets stuck. We will help decide whether AI belongs there, what should remain with a person and what a responsible first step looks like.