Who will be eliminated by the "small, fast, light, precise" trend?

If you are an industrial software company, "small, fast, light, precise" may not be good news.

Why?

Because in previous issues we have been saying:
software must become smaller; go-live must become faster; deployment must become lighter; and ultimately results must be proven.
It all sounds like helping enterprises reduce costs.
But from the perspective of an industrial software company, this is actually continuously compressing the living space of the traditional project model.

How did industrial software companies make money in the past?
A customer comes, I do research;
requirements are complex, I customise;
there are many on-site problems, I send implementation staff;
the longer the project cycle, the higher the contract value tends to be.
So what many industrial software companies truly sell is not actually software.
But rather:

Software + People + Projects.

This model was not wrong in the past.
Because the industrial site is indeed complex, and many things cannot be used directly out of the box.
But the problem is, if every new customer requires new research, new development, and new implementation, and you have done 100 projects, you may have 100 different sets of code, 100 different processes, and 100 different experiences.
The more projects you do, the busier the team may become, but not necessarily more valuable.

But the direction that "small, fast, light, precise" truly promotes is the exact opposite.

Small: Don't do everything – first go deep into one scenario.
Fast: Don't deliver from scratch every time.
Light: Turn more implementation work into the product's own capability.
Precise: Ultimately, prove that the customer truly achieved results.

So what industrial software companies should be most vigilant about in the "Small, Fast, Light, Precise" Digital Products and Services Cultivation Guidelines (2026 Edition) is not any particular technical requirement.
Rather, it is pushing the entire industry to answer one question:

Are you a project company, or a product company?

On the surface, both types of companies sell industrial software.
But the business is completely different.

What does a project company rely on to make money?People.What does a product company rely on to make money?Replication.

What a project company fears most is standardisation.
Because every customer is different, and the more customisation, the higher the revenue.
What a product company fears most is precisely the inability to standardise.
Because if every customer requires redevelopment,
there is simply no economy of scale.

So the truly competitive industrial software company in the future may not be the one that "can do everything."
But rather:

the one that can extract what repeats across 100 projects and distil it into a single product.

01The Real Moat Is Changing

At this point, the real moat begins to change as well.
In the past, the competition was:
whose development team is bigger, whose implementation team is stronger.
In the future, it will increasingly be: who better understands an industry.

Because a truly excellent industry product appears to be software on the surface,
but behind it lies a vast amount of substance:
process knowledge, equipment knowledge, quality rules, industry experience, business processes, and even data accumulated from numerous real projects.

So in the future, what is most valuable may not be how many projects you have done, but how much "replicable industry capability" you have distilled from them.

02AI Will Further Accelerate This

And AI will further accelerate this.
Previously, the experience of one industry expert could only serve a few customers.
In the future, by embedding that experience into software, knowledge bases, and AI Agents,
it may serve hundreds of enterprises simultaneously.
Previously, one implementation engineer could only solve a few on-site problems per day.
In the future, AI may productise large amounts of configuration, training, diagnosis, and knowledge services.

So where AI truly changes industrial software
may not be adding an "AI button" to existing software.
Rather, it is:

turning industry capability that previously could only be delivered by people into replicable products.

03From Projects to Products

Looking back at "small, fast, light, precise" now, you will find:
it is not actually asking industrial software to be "smaller, faster, lighter."
What it truly drives is an industrial upgrade:

From projects to products;from products to industry capability;from industry capability to AI capability.

So who will ultimately be eliminated?
Not "small companies." Nor "big companies."
Not even necessarily traditional software companies.
What is truly dangerous is:

those companies that can only ever make money through individual projects, yet cannot distil project experience into products.

Because the key to future competition may no longer be:
"Can I build it for you?"
But rather:

"Can I turn it into a product, and then let 100 enterprises use it?"

This is the change most worth watching behind "small, fast, light, precise."
On the surface, it is changing how SMEs buy digitalisation.
But at a deeper level,
it is forcing the entire industrial software industry to answer anew:

What we sell – is it software?Or is it the industry capability behind the software?

If this trend continues,
the truly valuable industrial software of the future
may not be the largest one,
but the one that best understands an industry, replicates most effectively, and continuously creates results.

This may be the true next stop for industrial software after "small, fast, light, precise."