Most (80%) enterprise AI projects fail.
Most of these failures are can be prevented by thinking about four simple questions.
Most of the remainder that fail, fail because of “leadership problems” and enough different types of leadership problems to require a separate article or even an article series.
Question 1: Do you have enough data?
You probably don’t. For what you are trying to do, do you know how much data is enought?
Question 2: Is your data of sufficient quality?
It probably isn’t. It is likely to have insufficient variety and too many errors. More than can be compensated for be high quality data engineering even if you haven’t skimped on data engineers in order to get more data scientists and ai specialists.
Question 3: Do you have sufficient data infrastructure?
You probably don’t. So even if it “works” you won’t be able to put into production. Even if it were technically possible to put it into production, you won’t have sufficient data governance in place and trying to add it later will be too hard. Of course, you probably won’t be able to get it to work because even in the unlikely event that you had had enough data of sufficient quality, it would have been too difficult to get it all where it needed to be.
Question 4: Is it actually possible?
There may have been an academic paper showing that a toy version of this problem worked. There may have been a version of something somewhat similar that a salespersons showed you. Or maybe it was just a concept demo.
The Other Questions
If your project can already survive seriously answering these questions, or be remediated so that it can survive these four questions, your risk of failure is better than cut in half!
You can now ask the more interesting questions that are categorized as leadership failures know that each one of these four represents a greater risk than all of the leadership risks put together. The harder, more interesting questions can be found here: Leadership failures leading to AI project Failures. While less likely to kill your project, they are still very likely to kill your budget and timeline.
Want to Avoid AI Project Failure? Let’s Talk.
If you want to dramatically reduce your time, cost, and risk, let’s discuss how your AI project can actually succeed.
✔ It’s free.
✔ It’s fast.
✔ It’s easy.
Schedule a time to talk. Only you can prevent catastrophic failure.
