Why product and B2 stay separate
Products identify the item sold. B2 identifies the invoiced buyer. Combining them before either is resolved would smuggle assumptions into both dimensions.
Product categorization answers what was sold. B2 buyer segmentation answers who was invoiced. They are independently established at their proper grain, then joined for reporting.
Need the exact B2 rule order? Open the B2 classification decision sequence →
Read left-to-right within each lane, then downward into the reporting join. The three B2 levels are evidence-development layers; final assignment still follows the stated precedence rules.
Products identify the item sold. B2 identifies the invoiced buyer. Combining them before either is resolved would smuggle assumptions into both dimensions.
Sales channel, personal name, tax status, salesperson, ship-to address, or an unapproved product cue. Those signals can prompt review, not automatic attribution.
Buyer segmentation is a revenue outcome dimension. Lead/source attribution would be a separate, later join using a deterministic order, contact, or campaign key.
Audit rule: preserve the B2 run/version, rule, and evidence source with every refresh. A later inference should never overwrite the historical basis of an invoiced result.