Most companies have started their AI journey. Very few have designed it.
The pattern is familiar. One team builds a chatbot. Another automates a form. A third explores a new platform. There is a lot of activity. There is very little direction.
Six months in, you have more tools than before. And the same problems, just faster.
Here is the uncomfortable truth about most enterprise AI programs: they automate the wrong things. Not because technology is bad. Because the foundation was never fixed first.
Companies layer AI onto existing services without asking whether those services are actually working. Processes that are confusing become confusing at scale. Duplication that was annoying becomes duplication at speed. Fragmentation that was manageable becomes fragmentation that is hard to even see anymore.
The technology is not the problem. The strategy is.
Most enterprise services are organized around the company, not the employees’ intent.
HR owns HR services. IT owns IT services. Finance owns Finance services. Each function builds its own corner of the experience. The result is a maze of portals, systems, and contact points that make sense on an org chart but makes no sense to the person who just wants to get something done.
An employee does not think: “I need to contact HR.” They think: “I need to update my bank details.”
An employee does not think: “I need to raise an IT ticket.” They think: “My laptop is not working, please help.”
That gap between how companies are organized and how employees actually think is where most AI programs lose value.
The fix is to invert the logic. Stop designing services around your org chart. Start designing them around what the underlaying intent is.
When you do this, everything changes. You can see which services touch the most people, you can identify what to build first and you can create a single, consistent experience across all functions, even if ten different teams are delivering it behind the scenes. And your AI actually works. Because it is routing real human needs, not navigating internal bureaucracy.
One rule matters more than almost any other in AI adoption: do not automate a broken process.
It is tempting. Automation looks like progress. But if the underlying process is unclear, inconsistent, or unnecessarily complex, automating it just locks those problems in and makes them harder to fix later.
Before any service goes near an AI platform, ask:
Services that cannot pass these checks should be fixed first. Automation comes second.
This discipline, simplify before you scale, is what separates AI programs that deliver lasting value from ones that create expensive new problems.
Enterprise AI programs almost always cross multiple functions. HR, Finance, IT, Procurement, each with their own priorities, their own timelines, and no obvious reason to coordinate.
The instinct is to centralize. Appoint someone to own it. Issue standards. Require compliance.
It rarely works. Functions push back. Progress stalls. The politics become the project.
What actually works is creating the conditions for alignment:
When people understand the direction and trust the logic, they align. You do not need to force it.
The companies scaling AI successfully are not necessarily the ones with the best technology. They are the ones who did the strategic work first.
They mapped what they actually need, they fixed what was broken and they agreed on a shared direction. And then they built.
The result is not just a better AI platform. It is a better organization, one where functions work together, decisions are made on evidence, and employees actually experience the difference.
Do we have a shared strategy for how we make decisions about AI?
If the answer is no, more tools will not help. They will accelerate what you already have.
Get the strategy right first. Everything else follows.
Natalya Permyakova is an Independent Consultant with a-connect with over 20 years of experience in innovation strategy, AI-enabled products, and enterprise transformation. She helps organizations translate AI ambition into structured, executable programs, designing the experience logic, decision frameworks, and shared principles that make adoption real across complex, multi-function enterprises. Natalya has advised Fortune 500 executives across life sciences, financial services, and global business services. She previously led innovation initiatives at Accenture and is the founder of My Life Quest, an AI-powered life design platform.
Register for our newsletter to receive curated insights directly to your inbox.
