Writing
Long-form pieces on ERP architecture, finance systems, and the decisions that determine whether an implementation works.
Companies spend months evaluating ERP software and a few weeks evaluating the firm that will implement it. That is backwards.
Boards are demanding an AI strategy. Finance teams are buying tools that sit unused. The sequence that works is unglamorous and it is not optional.
Integrations don't break on the software. They break on master data and on the order you load it in.
Decide what has to hit the GL, when, and at what level of detail. Then architect every upstream system to deliver exactly that.
Platform expertise is table stakes. The question is whether the person designing your revenue recognition has ever survived a month-end close.
Three signals tell me whether a company will actually finish an ERP implementation. Missing one is the reason most projects stall, not the software.
Most ERP damage happens after go-live, from the inside — over-provisioned access and ungoverned configuration changes. The fix is a permissions model and a change process.
Best-of-breed beat the all-in-one platform. But the win only holds if you enforce one rule: enter data once, then let it flow downstream.