Priced on cost, in a business built on value

Priced on cost, in a business built on value

An eight week sprint. One senior analyst and two juniors. Thirteen stakeholder interviews across four regions, ten product lines, two interview rounds each, and a workshop with the regional pricing managers at the end.

A speciality chemicals producer sells products whose whole reason to exist is that they do more with less. A polymer that lets a paper mill run at a third of the previous dosage is not a commodity. It is a saving on the customer's line, measured in tonnes of chemical not bought and hours of machine time not lost.

The company priced it as a commodity. In every one of its four regions, the minimum price guidance a product line manager was given began with variable cost and added a markup. The markup band differed by region. The logic did not.

Nobody in the company thought this was correct. One regional director put it plainly during the fieldwork: "We cannot be cost plus on everything. It doesn't reflect the value that certain products and services bring to the customer." The question the sprint was asked was not whether to move to value based pricing. It was why a company that already believed in it had not managed to do it.

Cost plus is not a decision anyone makes, it is what happens when nobody can see the alternative

Cost based pricing survives for a reason that has nothing to do with belief. It is the only method that can be executed with data the company definitely has. Variable cost is known to the decimal. Customer value is not known at all unless somebody goes and measures it, customer by customer, product by product.

So the failure is not in the pricing logic. It is upstream of it, in whether the organisation can see what its products are worth, and downstream of it, in whether anyone finds out what happened to the price after it left the room.

That is the shape this study kept returning to. The pricing conversation looks like it is about margin policy. It is actually about evidence: who collects it, where it goes, and whether the person setting guidance ever sees the outcome.

Thirteen people, four regions, two passes

1. Map the process as it is actually run

   Structured interviews with the global pricing manager, the regional director and the product line managers, mapping the real sequence from cost collection to price setting to monitoring, and the tools used at each step.

2. Go back a second time and validate

   A second round with the same people, confirming the flows, going deeper on how the tools are used in practice rather than how they are documented, and testing which factors actually move a price.

3. Count the bottlenecks rather than rank them

   Every problem raised was tagged to a process step and a region, so the output was a frequency rather than an opinion. Eight distinct bottlenecks emerged across six process steps.

4. Build the pricing method, then test it on the people who would use it

   A four step value based pricing framework, a worked quantification on a real product, and a workshop with regional pricing managers to find out whether it survived contact with the people expected to run it.

The friction is at the end of the process, not the beginning

Ask where pricing goes wrong and most organisations will point at the front of the process: we do not know our costs well enough, we cannot see the competition. This company said something different, and it said it almost unanimously.

The most cited problem, raised by nearly everyone and in every region, was that the deal system was not kept up to date. Lost bids in particular went unrecorded, or were recorded without a reason. The second most cited, again in every region, was that the product line manager who set the guidance was not in the room when the final price was agreed, and often was not told what it had been.

Put those two together and the mechanism becomes clear. Price guidance went out. Prices came back. The connection between the two was never closed, so nobody could learn from it. A pricing function that cannot see its own outcomes cannot improve them, whatever method it uses.

The ranking is the finding. Cost transparency, the thing pricing projects usually attack first, was the least cited problem in the company. The two problems that were cited by everyone both sat after the guidance was issued.

One company, four different answers to the same question

The formal pricing process was global. What happened inside it was not. The degree to which the person who sets the guidance is involved when the guidance is broken varied so widely between regions that they were, in practice, running four different pricing models under one policy.

The regional differences ran further than governance. One region was already adjusting its guidance on market signals through a standing commercial meeting and had built genuine premiums on two differentiated product lines. Another understood value based pricing perfectly well but had no way to capture what a product actually delivered during a trial, so it had turned its attention to standardising competitor data instead. A third had real wins, but reached them by running the trial first and setting the price afterwards, once the customer's own reaction had revealed what they thought it was worth.

That third approach is not a workaround. It is value based pricing discovered by accident, and it worked.

What the value looks like when somebody finally counts it

Halfway through the sprint the study stopped arguing about method and quantified one product, on one customer, using the numbers the customer's own trial had produced.

The product replaces a multi component treatment system with a single one. In the trial that was used as the reference case, the dosage needed per tonne of the customer's output fell from around three kilograms to around one. On the usage assumed in the model, that is roughly two hundred tonnes of chemical a year that the customer no longer buys, worth on the order of four hundred thousand euro at the average price the study used. A second, smaller lever followed from the same change: less cleaning, fewer blocked pumps, simpler dosing, put at about thirty thousand euro a year on an industry estimate.

Roughly four hundred and thirty thousand euro a year, on one product, at one site, from the two levers that can be checked. That is the number a salesperson can put in front of a purchasing manager. Before this sprint, nobody in the company had it, for any product, at any customer.

The price and the value were being set by two different logics that never met

Once the value is a number, the pricing question changes shape. It stops being "what is our cost plus our markup" and becomes "what share of the value we create should we ask for".

The study built that as a simple, deliberately transparent tool. Price equals the share of the created value that is captured, plus the variable cost of making the product. Move the capture rate and every downstream figure moves with it, so a pricing manager can see what a given ambition is worth before committing to it.

Two things came out of running it. The first is that the current cost plus price sits near the bottom of the range, which is what cost plus does when the product is genuinely differentiated. The second is more useful commercially: at every capture rate that was tested the margins were strong, which means there is room to price below the maximum for customers who are price sensitive and still be far ahead of where cost plus lands. Value based pricing here is not a demand for a higher price on every deal. It is the ability to know what you are giving away when you discount.

What is deliberately not published. The capture rate table, the product's price, its variable cost and its margins are the client's commercial position and are not shown. The method is what transfers.

Four steps, and a decision about which products are even candidates

The recommendation was not "adopt value based pricing". Applied to the whole portfolio that instruction fails, because most of a chemicals portfolio is genuinely commodity and pricing it on value is a way to lose it. The recommendation was a sequence, with a filter at the front.

1. Segment customers on whether they perceive value at all

   A decision tree rather than a matrix, starting from whether the customer sees high value in the product, then testing willingness to pay a premium for operational savings, for reduced risk, for customisation, for sustainability, and for service. Customers who will not accept a value led conversation are identified before anyone tries to have one.

2. Select products on value against cost to serve

   Two axes. Customer value, meaning the benefits and cost efficiencies that support a premium. End to end cost to the company, meaning production, customisation, service and compliance. Four quadrants, four different pricing answers, only one of which is value based pricing.

3. Quantify the value in the customer's own units

   List the levers, split them into cost saving and revenue generating, put an annual number on each, and normalise to a value per unit of product used, so that the figure can be compared with a price.

4. Communicate it, and make the evidence travel

   Equip sales with a product specific value narrative backed by an operational outcome, and circulate successful trial cases between regions, which is the step that turns one mill's result into an argument the whole company can use.

Alongside the method sat a less comfortable recommendation about the role itself. The product line manager in this company was a connector, sitting between sales, research, sourcing and operations, and setting guidance from cost. For value based pricing to work the role has to become the owner of the pricing argument, consulting and sometimes challenging sales rather than supplying them with cost inputs. That is a change in standing, not in process, and it is the part no framework can deliver.

What this study could not settle

The value quantification rests on one customer trial, reported second hand through an interview rather than measured independently. The study says so itself, and it is the right caveat. The company's own regional assessment made the same point more sharply: performance benefits vary significantly by site, which is precisely why a single value per unit of product is a starting point for a conversation and not a price list.

Two of the four value levers in the original model were revenue levers, and they were stated as the customer's revenue rather than the customer's margin, with the customer's own cost of producing that extra output not deducted. They also carried the larger share of the modelled total. They are not reported here for that reason. A value model that counts a customer's gross revenue as value created will overstate every case it is applied to.

Nothing here was measured after implementation. The sprint ended with a framework, a worked example and a trained group of regional pricing managers. Whether prices moved is not something an eight week engagement can know.

The workshop feedback contains its own honest signal. Participants rated the usefulness of the content highest and their confidence in actually applying value based pricing lowest, of the three things they were asked. Content transfers faster than capability, and the gap between the two is the real implementation risk.

Sources and method

Thirteen structured stakeholder interviews across four regions, two rounds each, covering ten product lines. A workshop with regional pricing managers to test the framework. A review of the tools used at each step of the pricing process, from cost collection through guidance setting to compliance monitoring.

Bottleneck frequencies are the share of the thirteen interviewees who raised each issue unprompted, tagged to a process step and a region. They are a measure of what people said, not an audit of what happens.

All monetary figures on this page are illustrative, built from a single customer trial reported during an interview, and are shown only where they reconcile from their own stated inputs. They are not measured client outcomes and should be rebuilt against real customer data before being used in a commercial conversation.

The client is described and not named. Product names, customer sites, internal systems, regional margin policy and the economics of individual products have been removed. Regions are lettered rather than named where the finding is about governance, because the combination of region and practice would identify the organisation.