A graph-based calculation engine, a sector-typed UI, an AI Copilot. Under the hood of every ABC model built on Performa, in ten minutes of reading.
Internally, Performa sees only Nodes and Edges. A node can be a building, a machine, an employee, an activity, a product. An edge is an allocation rule with a driver. The engine propagates costs across the graph and returns the unit cost and margin of every cost object.
On top of the engine, the UI renders industry-specific screens: workshop, hotel, professional practice — each with the custom fields that sector actually needs. The generic graph underneath keeps the system flexible without sacrificing sector-fit UX.
Not a checkbox list. Each of these solves a concrete problem anyone who has built an ABC model in Excel knows well.
Employee × activity matrix. Each cell ("Smith: 80% Machining, 20% Assembly") becomes an allocation rule without you writing it. Updating the timesheet updates the model. No side-Excels, no formulas replicated thirty times.
Not just constant drivers. The engine resolves aggregated drivers at calculation time: Sum (a field summed on a filter), Count (nodes matching a filter), SumProduct (two fields multiplied and summed), GroupSum (aggregated by group), Residual (group saturation via capacity − Σ others). Complex models stay maintainable.
Two-hop allocation templates (Resource Pool → Activity → Cost Object), consistent by construction — the engine guarantees that in every generated pair, the driver of the first hop equals the sum of weights in the second. Reusable across scenarios and clients.
For every activity with production capacity, the engine calculates how much has been absorbed by cost objects and how much stays idle. The cost of unused capacity appears as a separate P&L line ("under-utilization cost") rather than inflating product cost.
Cost-to-Serve for a single product, service or client. P&L by product, area, channel, business unit, cross-dimension. Every number is drill-down-able: click a cost line → see the rules that generated it → open the general-ledger entries at its origin.
Every scenario is independent: editing the baseline doesn't touch the what-if. Deep-copy to duplicate a scenario in one click. Side-by-side comparison between scenarios with absolute and percentage deltas on every cost object.
Not a generic chatbot. The Copilot knows Performa's schema, your active model, your calculations. It suggests drivers, explains why a node has a given cost, helps map cost items to activities. Useful while building the model and when interpreting results.
A user belongs to N organizations. Each organization is an isolated tenant: data, calculations, scenarios, users — all separated. The management accounting advisor runs the entire client portfolio from a single login. Roles per organization (Admin/Editor/Viewer) let you share the model with the client in a controlled way.
€ 307M+
costs already allocated in production
38
active client organizations
369
nodes in the largest live model
< 400ms
calc time on typical models (200 nodes, 500 edges)
Numbers as of July 10, 2026, measured on the production database. Architectural capacity supports up to 500 nodes per scenario, 5,000 cost items, 2,000 allocation rules — comfortable headroom beyond the largest models currently live.
If you have an Excel allocation file or an ABC model you want to modernize, send it over. In 20 minutes we'll tell you whether Performa is the right fit.
Write to us