When you spend some time on LinkedIn or any other business forum and you will see a very clear pattern. Everyone is talking about AI. Executives are excited and budgets are getting approved. But when you ask what has AI actually done for your company? The honest answer from most of them is… not much. For most companies it may be a chatbot that answers HR policy questions or something that auto-responds to emails or takes notes in online meetings. But are companies really getting the ROI on these AI investments? Simple answer is no.
Tech companies are a different story. For them the use case is pretty simple – Writing & reviewing code, and generating docs. But for others such as manufacturers, retailers, logistics companies etc, it’s mostly been frustration and a lot of money spent with not much to show for it.
So what’s actually going wrong?
The FOMO is driving the wrong kind of urgency
Every conference has AI on the agenda. Every industry publication has an AI cover story. AI is being discussed in company boards, CEOs are talking about it and nobody wants to be the company that “missed AI.”
So what happens? Organizations rush to implement. They announce AI projects and rush to hire consultants. Six months later the project get shelved quietly.
The problem isn’t AI. The problem is that everyone’s trying to skip the boring parts.
There’s a rule that everyone knows but nobody follows
There is a popular methodology for any digital transformation to be successful:
Digitize → Standardize → Optimize
AI is part of Optimize stage. It sits at the end of this chain, not the beginning. But companies are trying to jump straight to the last step and wondering why it’s not working.
“Trying to build AI on top of your existing mess is like constructing the fifth floor of a building when you haven’t even dug the foundation.”
The implementation will fail not because the tool was wrong or people dint put the effort, but because the foundation was not ready.
This is honestly the core of the whole problem.
Your data is probably a mess and nobody wants to work on it
AI doesn’t think. It finds patterns in data. And for that it needs data that is reasonably clean and in one place.
For most companies half the data is in some ERP, half in Excel sheets that one person maintains, some of it still in physical registers, and the same information entered differently across different locations. If you ask two locations for the same report, you’ll get very different reports where not only formats and data quality is different but the calculations and formulas may be different as well.
When you feed this to an AI you don’t get smart output. You get very confident wrong output. Which is actually worse.
Leadership usually believes that AI will come in and magically clean all this up. It won’t. That’s not what it’s for.
AI is not plug and play
Even if data is fine, AI doesn’t work on day one. It needs to be trained on your specific context, your processes, your terminology. It will get things wrong initially. It needs people to correct it with proper feedback and help it understand your business.
Most companies expect it to be like switching on a light. When it doesn’t work like that in the first few weeks, the project loses momentum. Then the official story is AI doesn’t work.
Are we prepared for AI
Before taking on any AI project, just ask yourself two things:
- Is our data actually in our systems, or is it scattered across spreadsheets and emails or in someone’s mind?
- Is that data consistent and clean enough that even a human analyst could work with it easily?
If the answer to those questions is a no, you’re not ready.
What you need to do is go back to step one:
Get your processes into proper systems. Digitize what’s still in Excel and registers. Make sure data entry is consistent. Do the boring foundational work.
Then and only then, think about AI.
To wrap up
AI isn’t failing because the technology doesn’t work. It’s failing because companies want the fifth floor without wanting to dig the foundation. The sequence matters and the preparation matters and no amount of budget or enthusiasm will change that.