Cost Analysis at E&C: From invoice validation to budgeting and forecasting

A single invoice, checked once. That's what it took to find a $180 000 discrepancy for a US automotive client. Most companies never find out whether they have an error like this, because checking them properly would mean rebuilding every invoice from raw consumption data, contract terms and published tariffs, work almost nobody does. The result? Budgets are built on numbers that may already be wrong. Tariffs run to dozens of pages, consumption is measured at intervals most finance teams never see, and rates change every year, sometimes every month. Getting a true picture of cost, and then planning around it with any confidence, depends on checking the numbers from source, rather than trusting what arrives on an invoice.

This is what cost analysis means at E&C. It starts with validating every invoice, and extends through budgeting, forecasting and accruals, so a client has not just an accurate read of what they have spent, but a reliable basis for what they are about to spend.

What are commodity and non-commodity energy costs?

An electricity invoice looks like one number, the total due, but it's built from two very different kinds of cost.

The first is commodity: the energy itself, priced according to the supply contract and the volume metered. This is the part a client has negotiated directly with their supplier.

Everything else is non-commodity, and it's typically the larger, more complex share of the bill. Network charges cover getting electricity from generation to the client's meter, split into transmission (long-distance, high-voltage) and distribution (the final, local delivery). Capacity charges reflect the cost of keeping enough generation and grid infrastructure available to meet peak demand. Taxes, renewable levies, and supplier fees make up the rest, each set by a different body: a tax authority, a grid operator, a regulator, a supplier, and each updated on its own schedule.

None of these components share a calculation method. A commodity charge is price × volume. A network charge might depend on peak demand in one specific hour of the month. A tax might be flat, a percentage, or tiered by consumption band. Checking an invoice properly means recreating every one of these calculations independently, not confirming the total looks reasonable.

How do we validate an energy invoice?

Energy tariffs are not single numbers. Germany alone can carry fifteen separate cost components on a single invoice, each with its own calculation.

To verify a bill, you need the same inputs the supplier used: the contract terms, the consumption data at the right interval (most clients are billed on fifteen-minute readings), and the official published tariffs for that location and that period. From those, you rebuild the invoice independently. This is what we call a 'shadow invoice'. It is not a sense check against the total. It is a full recalculation, line by line, against source data, compared back to what the supplier billed.

We do this for every invoice, not a sample. The reasoning is straightforward: a discrepancy doesn't announce itself. The only way to know a bill is wrong is to have already checked it.

What scale means here

Most external advisors price and resource this work by contract. A client with one contract covering twelve connection points and a client with one contract covering two hundred connection points look, on paper, like the same engagement.

That is not how the work behaves. Each connection point carries its own consumption profile, its own tariff parameters, its own line items to recalculate. A contract covering two hundred connection points means two hundred separate validations, each with its own consumption profile and tariff parameters. A contract covering twelve means twelve. We price and staff the work on that basis, which is also why the errors we find tend to track the number of connection points checked, rather than the number of contracts signed.

What can energy invoice validation uncover?

Most invoices are correct. Across the invoices we validate, around 2–5% contain an error, which is why a sample would miss most of them and why we check every one. The errors that do appear can be large: for one US client in the automotive sector, a single invoice carried a discrepancy of $180 000. A client receives many invoices, so most see at least some discrepancy over time, and the totals compound: a multi-country agricultural and food production client recovered €208 000 through invoice validation alone, on top of separate savings from portfolio optimization and contract negotiation.

Errors run in both directions. Suppliers occasionally overcharge, and they occasionally undercharge. Validation doesn't guarantee finding money. It guarantees knowing: every invoice is checked against what it should say, and any gap, in either direction, is reported.

Once a discrepancy is confirmed, nothing happens with the supplier without the client's authorization. The choice of whether to raise it, and how, sits with the client.

How does invoice validation improve energy budgeting?

Once you have a full, validated picture of historical cost (commodity, network charges, taxes, broken down by component), projecting it into next year is a matter of applying contracted or hedged prices to expected volumes, and updating the non-commodity components against the latest published tariffs.

This is also why we are candid about how far that projection should reasonably go. Clients sometimes ask for a five-year budget. We can build one, and have. But the further out the projection runs, the less it tells you. Non-commodity costs, in particular, become genuinely unpredictable past a two-year horizon: distribution tariffs, taxes, and regulatory charges shift in ways no one can forecast with confidence five years out. We will build whichever timeframe a client wants, but we would rather say plainly where the value drops off than sell a longer document because it looks more thorough.

Budgets are then checked monthly against the actuals they were built to anticipate, with quarterly or monthly reforecasting available where conditions move enough to warrant it.

That same breakdown does more than inform a budget. Knowing exactly how much of a cost sits in the commodity price, and how much is network charges, taxes, and other non-commodity components, tells you where a procurement strategy can change the outcome. Commodity costs are what a supplier negotiation, a hedge, or a change in contract structure can move. Non-commodity costs, in most markets, can't be negotiated away, only planned for accurately. A strategy built on a validated split between the two puts negotiation effort where it can work, rather than spending it on a component no supplier controls.

Read our blog on Setting up an energy budget: 8 step approach to avert a crisis

How do clients track energy costs and invoice data?

All of this – the invoice-level detail, the variance findings, the budget-to-actual tracking – lives on ePoint by E&C, our energy data management platform. A client looking at their cost analysis dashboard sees the full split: regulated against unregulated charges, then broken down further into commodity, network charges by type, and taxes and contributions, refreshed as new data comes in rather than batched into a periodic report.

Cost analysis overview

The budgeting view works the same way: forecast price against budget price by month, with the variance flagged in absolute and percentage terms, alongside the volume comparison that shows whether actual consumption is tracking what was planned.

None of this replaces the judgement of the people doing the calculation. It is the record of what they found, kept current and visible, rather than delivered as a static report once a quarter.

What comes next

Validation and budgeting are not one-off exercises. Once a budget is set, two further steps keep it useful through the rest of the year.

Forecasting takes the original budget and updates it as conditions change: actual consumption from validated invoices feeds back in, distribution tariffs and regulatory charges are refreshed against the latest published rates, and forward-looking cost estimates move with current market prices and any new hedges executed since the budget was first built. This runs monthly or quarterly, depending on what a client needs, so the budget stays a live working document rather than a fixed number revisited once a year.

Accruals handle the gap between when energy is used and when the invoice for it arrives. Each month, the cost of invoices not yet received is estimated from market prices, the hedges already in place, expected volumes, and the latest grid fees and tariffs, agreed in advance with the client so the numbers feed directly into their own financial reporting cycle.

Together, these mean a client is never working from a static snapshot. The same validated data that catches a discrepancy on one invoice is also what keeps next month's forecast accurate and next year's budget realistic.

What this means in practice

If you are responsible for an energy budget, the question worth asking is not whether your invoices are probably correct. It is whether anyone has rebuilt the calculation behind them, recently, against the rates that applied at the time. Cost analysis is not about finding the occasional billing mistake. It's about ensuring every financial decision, from invoice approval to forecasting, budgeting and procurement strategy, is built on numbers that have been verified. Without that foundation, every decision carries unnecessary uncertainty.

That is the gap this work is built to close.

FAQs:

See it in practice
Request a demo of ePoint by E&C to see the cost analysis and budgeting dashboards, and how invoice validation feeds into them.