One thing I've been wondering about: We spend a lot of time getting Salesforce right โ CPQ, approvals, dashboards. Then we connect DocuSign, ERP, billing โ and call it done. But in most projects, the real pain starts after. Prices don't match between Salesforce and ERP. Bundles that work in CPQ break in provisioning. Deals close, but accounting has to manually fix the data. It's rarely a technical failure. It's that the business logic wasn't aligned across systems from the start. Do you see this actually slowing down the business? Or have you found ways to catch these gaps before they hit production?
this is the failure mode i run into most, and it never shows up in testing because each system is individually correct. salesforce is right about the price it was told, the erp is right about the price it was told, and nobody owns the sentence that says which one is true. what's held up for me is picking a system of record per field, not per object. price lives in one place, quote structure lives in another, everything else reads. the moment two systems can both write the same field you have signed up for permanent reconciliation, and reconciliation is a person, not a process. for catching it before production, test the seam rather than the systems. take five real historical deals, the ugly ones with bundles, discounts and a mid-term change, and push them end to end before go live. the gaps surface in an afternoon. they never surface on a clean test record, which is exactly why they land on accounting instead.
This is spot on, Jonatan. "Reconciliation is a person, not a process" โ that's the line that sums it up perfectly. We've seen the same: clean test records never break. The ugly ones with mid-term changes and tangled bundles always do. Testing the seam is the only way to catch it. One thing we'd add: even with a system of record per field, the real friction often comes from mapping processes, not just fields. Sales, finance, and delivery teams usually have different definitions of "done" โ and that's where the seam turns into a wall.
agreed, and i'd go one step further on it. the different definitions of done aren't a communication failure, they're all correct. sales is done at signature, finance is done when it's invoiceable, delivery is done when it's provisioned and working. each team is right for the thing it's measured on, which is why the meeting where everyone agrees on one definition of done never survives contact with the quarter. what's worked better for me is refusing to converge them and writing them down as separate states instead. closed won, billable, live. three states, three owners, and a named owner for each transition between them. once that exists the argument stops being about what done means and starts being about who is late, and that one has an answer. which is the same governance point as the field one really. the seam only turns into a wall when nobody owns the crossing.
Really appreciate this perspective, Jonatan. This is one of the clearest frameworks I've seen for handling cross-system alignment. I'd love to pick your brain a bit more on this โ mind if I DM you?
