Everyone Says AI Is the Future. Then Why Are Most Companies Seeing Zero ROI?
Everyone is adopting AI. Very few are creating real business value from it. This blog explores why most AI projects fail to impact ROI — and why the companies winning with AI are focusing deeply on a few real business problems instead of spreading AI everywhere.
I came across an interesting LinkedIn post by Kumar Gautam discussing two very different reports on AI adoption.
One came from McKinsey & Company, which claimed that some companies are getting nearly $3 back for every $1 invested into AI.
The other referenced research connected to the Massachusetts Institute of Technology, which found that nearly 95% of enterprise AI projects showed little to no measurable impact on profit and loss.
And honestly?
That contradiction is exactly what most people in business are feeling right now.
Every company wants AI.
Every investor wants an AI story.
Every sales deck suddenly has the words “AI-powered” written somewhere.
But behind closed doors, a lot of teams are still asking:
“Okay… but is this actually helping the business?”
As someone from a business development background, this entire situation reminds me of one very common gym phenomenon.
Most people don’t fail at fitness because they lack access to equipment.
They fail because they buy:
expensive shoes,
smart watches,
protein shakers,
resistance bands,
and maybe even post one “Day 1” Instagram story…
…but never consistently train.
AI adoption in companies feels very similar right now.
A lot of organizations are collecting AI tools like Pokémon cards.
One tool for meetings.
One for emails.
One for presentations.
One for customer support.
One for analytics.
One for “AI note-taking.”
Half the employees don’t even know why the tools were purchased in the first place.
It’s like giving an air fryer, microwave, coffee machine, and stand mixer to someone who still orders food every night.
The tools are not the transformation.
That’s why the most interesting part of the McKinsey finding wasn’t the ROI number for me.
It was the implementation pattern behind it.
According to the discussion around McKinsey’s “Rewired” framework, the companies seeing actual returns were not trying to sprinkle AI everywhere.
They focused deeply on a few high-impact areas.
That makes complete business sense.
Because real business value rarely comes from doing 47 things at 2%.
It usually comes from doing 2 things extremely well.
Think about a restaurant.
If the owner says:
“We now use AI for menu design, music playlists, staff scheduling, customer reviews, Instagram captions, inventory tracking, and table booking…”
That sounds impressive.
But if the waiting time is still 40 minutes and the food is inconsistent, customers don’t care.
Now compare that to another restaurant that uses AI only for one thing:
predicting ingredient demand accurately.
Suddenly:
wastage reduces,
ingredients stay fresh,
margins improve,
delivery becomes faster.
That’s measurable business value.
And that’s probably the difference between “AI adoption” and “AI implementation.”
One is experimentation.
The other is operational change.
I also think many companies are making another mistake:
they are implementing AI because they fear being left behind, not because they clearly understand where value will come from.
Which creates a dangerous cycle:
leadership wants AI,
teams rush pilots,
dashboards look exciting,
LinkedIn posts get published,
but nobody defines what success actually means.
At that point, AI becomes a branding exercise instead of a business strategy.
And to be fair, this isn’t entirely new.
Businesses have done this before with:
blockchain,
metaverse,
NFTs,
“digital transformation,”
and even apps during the mobile boom.
Sometimes companies don’t adopt technology to solve problems.
They adopt it because everyone else is doing it.
The companies that usually win are the boring ones.
The ones that quietly ask:
Where are we losing time?
Where are costs increasing?
Where are employees doing repetitive work?
Where are customers facing friction?
And then they use technology there.
Not everywhere.
Just there.
That’s probably why I personally trust the “focus” lesson much more than the flashy ROI headline.
Because whether the return is 30%, 200%, or somewhere in between, the core business principle remains the same:
Technology creates value only when it solves a real operational problem.
Not when it simply creates excitement.
Credit:
This article was inspired by a LinkedIn post by Kumar Gautam discussing the contrast between McKinsey’s AI ROI findings and MIT’s broader research on enterprise AI outcomes.