I don’t think businesses have an automation problem. I think we’ve misunderstood what ERP systems were supposed to do.
For twenty years, companies have poured millions into ERP platforms to centralise information. And it worked. We got very good at recording the business. But somewhere along the way, we stopped asking the next question: recording it for whom, and to what end?
Because information was never the real bottleneck. Execution is.
Right now, in construction businesses everywhere, thousands of invoices, purchase orders, progress claims and approvals still depend on someone manually moving information between a document and a system. The ERP records the business. People still run it.
That gap is the thing I want to talk about.
The system that succeeds technically and fails operationally
Most ERP implementations I’ve seen work exactly as designed. They go live, the data is structured, the reporting looks clean. Technically, it’s a success.
Operationally, it’s a different story. Employees are still typing invoices. Still entering orders. Still matching POs by hand, chasing approvals over email, correcting mistakes that never should have made it into the system in the first place.
That’s not automation. That’s digitised administration – and it’s easy to mistake one for the other because the ERP looks so capable sitting there in the background. But a database that waits for people to feed it isn’t executing anything. It’s just recording what people already did.
The enemy isn’t your ERP
I want to be clear about something: NetSuite, Premier, and the other platforms construction companies run on aren’t the problem. They’re good at what they were built to do.
The enemy is the manual work that has grown up around them: invoice processing, order entry, approval chasing, data validation, the spreadsheets and email threads that exist purely to keep the ERP updated. None of that work creates value for the business. It exists to compensate for a gap the ERP was never designed to close.
Operational debt
There’s a concept I keep coming back to, because I think it’s the most useful way to see this problem clearly: every manual transaction is operational debt.
One invoice processed by hand isn’t a problem. Twenty thousand a year are. One order entered manually isn’t a problem. Five thousand a month, compounding across every project, every cost code, every approval layer, are.
Just like technical debt, operational debt doesn’t announce itself. It accumulates quietly: in the extra hire you make instead of fixing the process, the manager added to catch errors, the reconciliation that takes three extra days at month end, the compliance risk nobody notices until it’s expensive. Companies think they’re saving money by putting off automation. In reality, they’re financing a liability that compounds every single month.
I’ve spent enough time in construction finance conversations to know this isn’t hypothetical. It’s the reason project profitability is hard to see in real time. It’s the reason month-end close takes longer every year instead of shorter. The root cause was never the ERP. It’s the manual work surrounding it.
The shift
The old thinking was: we automate to save time. I think that framing has run its course.
The new thinking is: we automate to eliminate operational debt. Saving a few minutes per invoice was never really the objective. Building a business that can grow without adding administrative headcount for every extra project is.
Today, most ERPs still run on a simple path: documents go to people, people go to the ERP. Tomorrow, that path should look different: documents go to AI, AI applies your business rules, and only the exceptions that genuinely need judgement land on someone’s desk. Everything else executes on its own.
That’s the shift construction finance teams are in the middle of, whether they’ve named it yet or not.
Where SmartUi fits
This is why we didn’t build SmartUi as an OCR tool or a piece of AP automation software bolted onto an ERP. We built it as an operating layer, the thing that turns your ERP from a system of record into a system of execution.
Instead of your team moving information into NetSuite or Premier, SmartUi captures the documents, validates them against your business rules, manages the exceptions, and executes the transaction. Cost codes, retainage, progress claims, approvals, the parts of construction finance that generic AP tools were never built for.
The finance team’s job was never supposed to be typing numbers between systems. It was supposed to be making decisions with them. That’s the job automation gives back.
The businesses that win
I don’t think the businesses that win over the next decade will necessarily have better ERP systems. I think they’ll have ERP systems that execute work autonomously, and they’ll have started building that capability now, while their competitors are still hiring to keep up with the manual work.
ERP systems were designed to record work. The next generation of software will execute it. The companies that eliminate operational debt today will be the ones that scale fastest tomorrow.
