Process automation is a mindset, not a tool
Every company wants "automation". Few can say which process, in what order, and how they will know it worked. Here is how I decide what to automate — across manufacturing, finance and enterprise .NET backends.
Process automation has become a checkbox on every company's strategy deck. But automation is not a product you buy; it is a way of seeing your own operations. The tool is the last decision, not the first. Across my work — a production scheduler at Dekali, a document-verification system at Studium Group, a CRM overhaul in subsidised finance — the winning move was always the same: understand the process deeply enough to know which part is worth removing.
The first question is never "what can we automate?" It is "where does a human currently wait, re-key, or re-check?" Waiting, re-keying and re-checking are the three signatures of an automatable step. A salesperson waiting two days for engineering to price a quote. An operator re-keying an order from an email into the ERP. A compliance officer re-checking a document against a database by eye. Each is a place where information that already exists is being moved or verified by hand. That is the target.
The second question is order. You do not automate the biggest process first; you automate the one with the clearest input, the clearest output, and the lowest cost of being wrong. That first win builds the trust and the data you need for the harder ones. At Studium Group the document-verification system started with the document types that were cleanest and highest-volume, not the exotic edge cases — because the point of the first release is to prove the pattern, not to handle everything.
The third question, the one most teams skip, is measurement. Before writing a line of code, define the number that will tell you it worked, in the operation's own units: quote turnaround, queue length, response time, error rate. If you cannot name that number, you are not ready to automate — you are ready to keep studying the process. An automation you cannot measure is an automation you cannot defend when someone asks whether it was worth it.
The mindset underneath all three questions is this: automation is subtraction, not addition. You are not adding an AI to the company; you are removing a delay from a process the company already runs. The best automation is invisible — the work simply happens faster, the queue simply stops growing, and six months later nobody remembers doing it by hand. That is the goal. Not a dashboard full of AI. A business that quietly runs better.