Need help with problems like this?

Let's talk about your next steps.

Published

November 10, 2026

Should-cost analysis for chemicals and plastics: what your quote should cost

How to build a should-cost analysis for chemicals and plastics: map the route, price each input with an index, and check a supplier’s quote against it.

A supplier quotes €2,688 a tonne for next quarter, 12% more than last time. You can accept it, push back, or ask for a breakdown. What is hard to do on the spot is say what the material should cost. Without that number, the negotiation starts on the supplier’s terms.

Should-cost analysis estimates what a product should cost to make and deliver, built up from its inputs, conversion costs, logistics and a reasonable margin, so you can set it against the price you are quoted. For chemicals and plastics, most of that cost follows raw material and energy prices you can track in public indices.

Most guides to should-cost modelling are written for machined parts and bills of materials. This one shows how to build a should-cost model for a material, where to find the numbers, and how to use it when a quote arrives.

In short:

  • Build it from the production route: what the material is made from and with, not what it weighs.
  • Price each input with a dated index, so you can see what moved and when.
  • Compare the change since the last agreed price. It is more reliable than the absolute level.
  • Treat conversion costs and margin as ranges, not single numbers.
  • The gap between the should-cost and the quote is what the supplier needs to explain, and what you can counter.

Holding a quote you’d like to check? We’ll go through what its inputs support with you.

What is should-cost analysis?

Should-cost analysis builds an estimate of what a product should cost from the bottom up: raw materials, energy, conversion, logistics and margin. It gives the buyer an independent number to set against a quote. It is not a demand to pay cost price; it shows which parts of a price are explained by costs and which are not.

It differs from price analysis, which compares a quote with other quotes, past prices or a market index. Price analysis tells you whether a price is in line with the market. Should-cost analysis tells you why the price is what it is, and which part of it a supplier would have to justify. Our guide to price benchmarking for raw materials covers how the two fit together.

The UK government treats it as standard practice. Its Sourcing Playbook requires a should-cost model estimate for complex outsourcing projects and adds: “Outside of central government, production of a Should Cost Model is considered best practice.”

Why is should-cost different for chemicals and plastics?

For a machined part, most of the cost sits in machine time, labour and the parts list. For a chemical or resin, most of it usually sits in feedstock and energy, which are traded and published as indices. That makes a material should-cost easier to keep current, but more sensitive to which inputs, which region and which dates you choose.

The production route decides which inputs matter. PVC is made from ethylene and chlorine, and chlorine is made by electrolysis, so electricity is one of its main cost drivers; we explain how in inside the price of PVC. European ammonia is made from natural gas, while most Chinese ammonia starts from coal, so the same molecule can have very different costs depending on where it is made.

Input prices also move every month, sometimes every day. A material should-cost is therefore never finished: it is a model you rerun when prices move or a quote arrives. That is also its strength. When a supplier cites higher costs, you can check which inputs moved, by how much and since when.

What goes into a should-cost model for a material?

Five blocks: raw materials, energy, conversion, logistics and packaging, and margin. Each needs a quantity per tonne of product and a price, and each price should come from a source you can name and date. The blocks that move with markets need indices; the others can be estimated and updated less often.

BlockWhat it coversHow to price itHow often it moves
Raw materialsFeedstocks and intermediates per tonne of product, including yield lossesA public index or price report for each inputMonthly or faster
EnergyGas, power and steam per tonneGas and power benchmarks, or producer price indices for energyMonthly or faster
ConversionLabour, maintenance, depreciation and plant overheadsAn estimate for the plant's scale and region, updated with an inflation or wage indexYearly
Logistics and packagingFreight, duties and packaging to your siteFreight quotes, freight indices and published duty ratesQuarterly, or at once when a duty changes
MarginThe producer's returnA range, wider in tight markets and narrower in weak onesWith the market cycle

Quantities come from the production route and the producer’s yields, not from the product’s composition by weight. An input that makes up half of a product’s weight can make up much more or much less of its cost, depending on its price, so weight shares are no substitute for cost shares.

How do you build a should-cost model, step by step?

Map the production route, put a quantity and a dated price on each input, add conversion, logistics and a margin range, then compare the result with the quote. For a negotiation, run the model at the date of the last agreed price and at today’s date: the change between the two is the most defensible number you will have.

  1. Map the route: list the main inputs for your grade and region. Your supplier’s technical data, your own R&D team and published process descriptions are the usual starting points.
  2. Set quantities: put a quantity per tonne of product on each input, allowing for yield losses.
  3. Price each input: pick one index or price report per input, note its date and currency, and use the same basis every time you rerun the model.
  4. Add the rest: estimate conversion, logistics and packaging, and set the margin as a low and a high case.
  5. Compare at two dates: run the model as of the last agreed price and as of today, and set the change against the change the supplier is asking for.

Timing matters as much as the inputs. Contract prices often follow their inputs with a delay, set by the review dates and index periods in the contract. We describe how that delay works in how an energy shock reaches your invoice.

What does a should-cost check look like in practice?

Take the last agreed price, split it into its cost blocks, apply what each block’s index has done since, and add up the change. In the example below, the inputs support an increase of about 7%. The supplier asks for 12%, so about five points of the increase need another explanation.

A worked example, with hypothetical numbers. The last agreed price for a resin grade was €2,400 a tonne. The supplier now asks for 12%, or €2,688.

Cost blockShare of last price€ per tonneIndex change sinceChange, € per tonne
Main feedstock55%1,320+8%+105.60
Energy10%240+20%+48.00
Other raw materials10%240+2%+4.80
Conversion, logistics, packaging15%360+3%+10.80
Margin10%240Held in euros0
Total100%2,400+169.20 (+7.1%)

The inputs support a price of about €2,569 a tonne. The quote is €2,688, so about €119 a tonne, roughly five points of the twelve, is not explained by what the inputs did.

That gap is not automatically unjustified. It may be a margin the supplier is rebuilding after a weak year, a cost the model misses, or a tighter market. But it is the part the supplier has to explain, and the part you can counter. The same method, applied to a single price increase letter, is in what to check before you accept a supplier’s price increase.

Where do you find the numbers?

Public producer price indices cover most chemicals and energy by product group, monthly and free. Price reporting agencies cover individual chemicals and polymers in more detail, for a fee. Your own invoices, contracts and supplier breakdowns fill the gaps. Note the source, the date and the time lag for every number you use.

  • EU: Eurostat’s producer prices in industry, monthly, for the EU and each member state, by product group such as plastics in primary forms or basic organic chemicals.
  • Germany: the Federal Statistical Office’s producer price index for industrial products, published monthly with a press release and detailed tables by product.
  • United States: the Bureau of Labor Statistics’ Producer Price Index, which “measures the average change over time in the selling prices received by domestic producers for their output.”
  • Energy: gas and power benchmarks for the producer’s region, such as the Dutch TTF gas price in Europe.
  • Individual chemicals and polymers: price reporting agencies such as ICIS, S&P Global Commodity Insights and Argus, by subscription.
  • Currency: the ECB’s euro reference exchange rates, when inputs are priced in dollars and your contract in euros.

Indices show how far apart the steps of the chain can move. In Germany in September 2026, according to the Federal Statistical Office, naphtha cost 53.8% more than a year earlier, energy 14.9% more and basic chemicals 11.3% more. A should-cost model makes that difference visible for the material you buy.

Keep in mind what an index can and can’t tell you. It averages many producers and grades, so it shows how far and in which direction an input moved, which is what you need for the change. It is a weaker guide to the exact level for your grade.

How accurate does a should-cost model need to be?

Accurate enough to show which parts of a price are explained by costs and which are not. The absolute level of a should-cost model carries real uncertainty; the change since the last agreed price is more reliable, because many errors in the level, such as a misjudged conversion cost, appear at both dates and cancel out.

Three habits make a model more trustworthy. State the result as a range, with a low and a high case for margin and conversion. Test it against history: if the model would have explained the supplier’s past price moves, it deserves more weight now. And rerun it when the indices are published, so the number you take into a negotiation is current.

A model that can’t be updated is the most common reason should-cost work stops. Keep the inputs few, the sources named and the dates visible, so anyone on the team can rerun it.

What mistakes make a should-cost model misleading?

Most errors come from picking the wrong inputs or the wrong dates: weight shares used as cost shares, an index from the wrong region, prices from after the contract’s index period, or a surcharge counted twice. Each one makes the model look precise while pointing in the wrong direction.

MistakeWhy it misleadsWhat to do instead
Weight shares used as cost sharesA cheap input can be heavy and an expensive one lightWork out cost shares from quantities and prices
An index from the wrong region or routeA gas-based European producer and a coal-based importer face different costsMatch the supplier's route and region
Ignoring timingToday's spot price set against a price fixed on last month's inputsUse the dates and index periods in the contract
Counting a surcharge twiceAn energy surcharge on top of a base price already indexed to energyAsk what each line of the price covers
A fixed marginMargins swing with the cycleUse a range, and say which case you are in
One number, no rangeFalse precision invites a dispute about the number instead of the gapShow a low, central and high case

Some costs sit outside the five blocks. Carbon is one: if your supplier pays for emissions, part of that cost is already in the price, and part may come as a separate line. We look at how to separate the two in why should-cost models need a carbon layer.

Questions to ask your supplier

A should-cost model is stronger when the supplier fills some of the gaps. These questions work whether or not you share your own numbers:

  1. Which inputs drive the price of this grade, and in what quantities per tonne?
  2. Which index or price report do you use for each input, and for which dates?
  3. What share of the price does each input make up at today’s prices?
  4. What has changed since our last agreed price, input by input?
  5. Is any of the increase already covered by a surcharge or an index clause?
  6. Which part of the increase is margin, and why now?
  7. What happens to the price if the main input index falls back?

From a should-cost number to a decision

A should-cost model tells you where a price stands today. The decision also depends on where the inputs are heading. An increase the inputs support may still be worth countering if those inputs are turning down, and one that looks high may be worth locking if they keep rising; we look at that trade-off in is a fixed-price offer worth the premium?

Sybilion helps industrial buyers decide what to pay for the materials they purchase, and when to commit: accept, counter, lock or wait. For the materials it covers, it shows which cost drivers sit behind each price and how far they have moved, with a forward view stated as a range, so a quote can be checked against what its inputs support before you sign. It works alongside the systems you already use.

Check your next quote before you sign.

Frequently asked questions

What is a should-cost model?

A spreadsheet or tool that estimates what a product should cost from its inputs, conversion costs, logistics and margin. For a chemical or plastic, it combines quantities per tonne from the production route with dated prices for each input, so it can be rerun whenever input prices move or a quote arrives.

What is the difference between should-cost and will-cost?

Will-cost is what something is expected to cost if things carry on as they are. Should-cost is what it ought to cost if it is made efficiently at today’s input prices, with a reasonable margin. The gap between the two is where the savings conversation sits.

Is should-cost analysis only for manufactured parts?

No. Most should-costing tools were built for machined and moulded parts, but the method works for chemicals and plastics too. There, most of the cost follows feedstock and energy prices, which are published as indices, so the model is easier to keep current.

Can you do should-cost analysis in Excel?

Yes, for a handful of materials. A spreadsheet with one row per input, a quantity, a price source and a date is enough to start. It gets harder to keep current across many materials, suppliers and regions, which is where most should-cost work stalls.

How often should you update a should-cost model?

Update the raw material and energy blocks when their indices are published, usually monthly, and conversion and logistics once a year or when something changes, such as a new duty. Rerun the whole model whenever a quote or a price increase letter arrives.

Explore more customer  stories

Frequently Asked Questions

What data do you use?

We use only the verified data from official institutions, market research companies, and other reliable sources vetted by us.

Each data source has to pass an extensive verification process before it is used in our analysis.
Each data source has to pass an extensive verification process before it is used in our analysis.Each data source has to pass an extensive verification process before it is used in our analysis.

How accurate are your trends?

We only provide forecasts that bring significant improvements (30%-70% relative error reduction) in comparison to established baselines.

What security measures do you use?

We use the latest and highest security standards in cloud architecture and access policies.

All data we used is anonymized and doesn’t contain any reference to customers or otherwise.

What do you mean by explainable?

Explainability means understanding why trends may unfold in a certain way and what external market factors influence them. Sybilion provides context and transparency to help you understand these factors.

Can I confidently share my data with you?

Yes. Our AI does not require data, that is significantly more sensitive than what you would anyway share in your annual reports.

We handle data with care and apply the latest security and hosting standards.