I've built automation into every layer of Fulfill.com's operations, from document processing for onboarding new warehouse partners to data extraction across thousands of SKUs. Here's what I've learned about making automation work at scale. A Center of Excellence isn't about creating another bureaucratic layer. It's your insurance policy against automation chaos. Without one, you end up with ten different teams building similar bots, none talking to each other, all breaking when something changes. I've seen this firsthand with brands we work with who automated their inventory management in three different ways across three departments. The result was more confusion than efficiency. Do you need one? If you're running more than a handful of automation processes or planning to scale beyond basic tasks, yes. We established our CoE when we hit about fifteen automated workflows. Before that, we were spending more time fixing broken automations than they were saving us. For minimum viable structure, start lean. You need three roles, not three full-time people. First, a business process owner who understands the work being automated. This person should spend about 30 percent of their time on CoE activities. Second, a technical lead who can actually build and maintain the automations, maybe 40 percent commitment. Third, a governance person who documents standards and manages the pipeline of automation requests, around 20 percent time. At Fulfill.com, these were existing team members who took on CoE responsibilities alongside their regular roles. The IT versus business team question is critical, and most companies get it wrong. Business teams must lead this, with IT as a strategic partner. I'm adamant about this because automation is fundamentally about business process improvement, not technology deployment. When IT leads, you get technically perfect solutions that don't solve real problems. When business leads with IT support, you get automations that actually move the needle. In our document processing automation, our operations team led the charge because they understood the pain points of manually processing hundreds of warehouse applications monthly. IT provided the technical guardrails and security requirements, but operations defined what success looked like. That partnership is what made it work. The biggest mistake I see is treating a CoE as a cost center that slows things down. Done right, it's your accelerator.
What I've seen is that automation works best when IT and the business each own the part they're built for. If one side tries to run the whole thing, it usually stalls. IT should lead on architecture, security, integrations, and making sure the automation is stable. But the business teams need to lead on the workflow itself, because they're the ones who understand the pain points, the edge cases, and the real cost of errors. When business teams define the rules and IT enables the execution, the automation fits how people actually work. That balance is what turns a pilot into something that scales reliably across the organization.