Built from raw tables
Each line of the P&L is a node in a graph with its own quality gate. Rows that fail are counted, never silently dropped, and every node carries a chained hash of its sources.
The P&L Twin rebuilds your P&L line by line from raw sell-in, sell-out, price, cost and freight tables. Every number carries where it came from, what was discarded on the way, and whether it was measured or estimated. Each decision gets its own copy of the twin, so its effect is priced before anyone approves it.
Each line of the P&L is a node in a graph with its own quality gate. Rows that fail are counted, never silently dropped, and every node carries a chained hash of its sources.
Cost with no dimension to land on stays marked as not allocated. Flows, capital and decompositions are kept apart, because adding them together is the most common way a P&L analysis inflates.
Each decision gets a base, a scenario and a delta, with ranges. When the underlying data changes, the scenario is flagged stale and recomputed.
Every assumption is tagged measured, derived, benchmark or pending. Estimated lines carry their formula, their range and an owner who can change them.
Net revenue, cost of goods and freight, by every dimension your data supports.
Built from classified raw tables, node by node.
Where did this number come from, and what was left out?
Chained hash per node and counted discards.
Which numbers are measured and which are not yet?
Origins: measured, derived, benchmark, pending.
What does this decision do to the P&L before we take it?
Base, scenario and delta, with ranges.
What does an agent need to know about this business?
A read-only context feed from the twin.
A potential is not a forecast. A missing measurement is treated as unknown, never as zero.
A Value Tree maps your P&L in 14 days and shows where the decisions that move it sit.