Thirteen providers, three kinds of rival

The chemical is bought once. The control layer is bought every day.

A paper machine is a long, hot, wet argument between speed and quality. Pulp goes in at one end, paper comes off the other, and along the way a dozen chemistries are dosed into the stock to hold the fibres together, keep the water clean, stop the microbes, control the foam and keep the sheet from breaking. Get the dosing wrong and the machine slows, the sheet tears, or the mill pays for chemistry it did not need.

The ask was not a market size. It was a map of who else sells this, what they promise, how they charge for it, and who they had to partner with to build it. The client was deciding where to invest, where to buy and where to collaborate, and needed the outside view before the next planning cycle.

Our client, a global speciality chemicals company with revenue of around €3 billion, had been selling those chemistries into pulp and paper mills for decades. It had also built a digital service layer on top of them: sensors, online monitoring, automated dosing, alarms, reports. Five application areas, seventeen named services, covering the process from pulping through to the finished sheet.

The commercial logic was straightforward. A mill that buys monitoring tends to buy control next, and a mill running your control loop does not switch chemical supplier casually. The digital layer was a retention mechanism as much as a product.

What nobody inside the company could answer was whether that layer was ahead of the market or behind it. Competitors were visibly announcing digital products. Machinery suppliers were selling control systems into the same mills. Software companies were arriving from outside the industry with analytics platforms and no chemistry at all. The client had an internal roadmap and an acquisition already made, and no external view against which to judge either.

Digital services are priced inside a chemical contract

There is no market report for this. Digital services for pulp and paper are not a reported category, and for a simple reason: for most of the companies selling them, they are not a separate business. They are an attachment to something else.

The offering is public because it is marketing. The economics are private because they are contract terms.

A chemical supplier bundles monitoring into a supply agreement. A machinery supplier sells advanced process control as an option on a line it is already installing. A software company licenses a platform that happens to have a paper mill on it. None of these appear as a line in an annual report, and none of them carry a published price.

That shapes what a study like this can and cannot do. Service catalogues, technical descriptions, application areas and customer references are all published, because vendors want mills to read them. Prices, attach rates, revenue and margin are not, and no amount of desk research turns them up.

There is a second difficulty, which is that the competitor set is not one set. A chemical rival, a machine builder and an analytics platform all sell into the same mill, all use the word digital, and are not comparable on any single axis. Ranking them on one list produces a ranking that means nothing. They had to be grouped before they could be compared, and the grouping is itself a finding.

Three kinds of rival, and only one of them is a rival

The client had been thinking about this market as a single competitive field. It is not. Thirteen providers sell digital services into pulp and paper mills, and they fall into three groups that behave completely differently.

Direct chemical competitors sell what the client sells: chemistry with a control layer on top. They share its customers, its application areas and its commercial logic. There are four of them, and they are the only true rivals in the set.

Automation and machinery providers sell the machine, and increasingly the software that runs it. Six of them. They are wider than any chemical company, because their platforms serve several process industries, but they do not sell chemistry and they cannot advise on it.

Technology providers sell analytics with no industry attached. Three of them. They are strong where the others are weak, at machine learning and data infrastructure, and weak where the others are strong, at knowing what a paper machine actually does.

Two of these groups are competitors. One is a supplier of capabilities the client cannot build alone. Treating all three as threats would have been the expensive mistake.

Plotting them this way makes the strategic question legible. Relevance and portfolio width pull in opposite directions. The providers most like the client have the narrowest offerings. The providers with the widest offerings are the least like it. Nobody occupies the top right, which is the interesting part: there is no player in this market who is both close to the client's business and broad across the mill.

Only two of the four chemical competitors are actually peers

Four chemical companies sell digital services into these mills. Catalogued service by service, they do not look alike.

Where the counting stops. This grid records what each provider publishes as a named service. It does not record how many mills run it, what it costs, or whether it works. Attach rates and prices are contract terms and are not disclosed by anyone in this market.

Two of the four span the client's whole territory: pulp production, papermaking, water treatment and microbe control. Those are the peers that matter. One of them offers eight distinct digital services, the widest of any chemical company in the set, including a wet end stabiliser and an entrained air control service that nobody else sells.

The other two are concentrated in water treatment and microbe control. They compete hard there, with legionella and biofilm services the broad peers do not offer, but they are absent from the papermaking side of the mill entirely.

That distinction changes what the client should do about each. Against the two broad peers, the contest is portfolio against portfolio, and the answer is depth: which services actually close the loop rather than merely reporting. Against the two water specialists, the contest is local, in one application area, and the answer is whether to match them there or concede it and defend the rest.

Every machinery provider sells the same control layer

The machinery group is where the client expected to find the real threat, because these companies are already inside the mill with a control system and a service contract. The catalogue tells a more specific story.

Fifteen distinct services across the six providers, and exactly one of them is offered by all six: advanced process control. Three more appear in half the group: distributed control systems, quality management and wet end control. Everything else is a single provider's speciality.

The spread is wide. The broadest provider in the group sells seven of the fifteen services, two more sell six each, and the narrowest sells two. So this is not a uniformly capable group. It is one or two genuinely broad platforms and a tail of specialists who have automated the part of the mill they build machines for.

The strategic reading is that the machinery group has the control layer and not the chemistry. They can hold a set point. They cannot tell a mill which chemistry to dose to hit it. That gap is the client's, and it is the reason the same six providers appear again in the next finding as partners rather than rivals.

The technology providers cannot sell into this market alone

Three technology companies were profiled: two industry agnostic analytics platforms and one pulp and paper specialist. Between them they offer ten services, and the split is stark.

The two generalists bring machine learning, digital twins, predictive maintenance and equipment effectiveness monitoring. One of them has been deployed by several pulp and paper companies and clearly works. Neither has any service specific to the process: no liquor analysis, no wet end, no chemistry.

The specialist is the mirror image. It knows the process well enough to sell online measurement and black, white and green liquor analysis, and it sells mostly hardware. Its analytics offering is thin.

Neither shape is a competitor to a chemical company with domain expertise and an installed base of dosing equipment. Both are candidate partners, and for opposite reasons: one supplies the technology the client lacks, the other supplies measurement the client could dose against.

Everyone promises the same six things

Reading thirty-seven published deployments and every value proposition in the set produces an uncomfortable result: the promises converge. Whatever the provider sells and whichever group it belongs to, the customer pain points it claims to solve reduce to six themes.

When the promises are identical, credibility is the product.

This matters more than it looks. If every provider promises the same six outcomes, then the promise is not a differentiator and marketing spent on it is wasted. What differs is the evidence behind the promise and the position from which it is made. A chemical company saying it will reduce your chemical usage is making a claim against its own revenue, which is a far stronger signal than a software company making the same claim.

Five ways to charge, and the industry is moving towards two

How a digital service is priced determines what kind of business it becomes. Five models were identified across the thirteen providers.

One-time capital expenditure is the oldest and the least attractive: it turns a service into a project, with no recurring revenue and no reason for the customer to stay. Service contracts, where the digital layer is bundled invisibly into a supply agreement, are common among chemical suppliers and are effectively how the client already operated. They work, but they hide the value: nothing on the invoice tells the mill what the software was worth.

The two models the industry is moving towards are subscription, where the service is sold and renewed on its own terms, and upgrade on the physical product, where the software is attached to equipment already being sold. For a company that sells chemistry and dosing equipment, both are available, and the second is the easier first step.

Co-development sits in the middle. It is not really a pricing model so much as a way of sharing the cost of building something neither party can build alone, and it recurs constantly in the partnership record.

What this rests on

Every count on this page comes from the service catalogues assembled during the sprint and can be recounted from them. Where a position on a chart is a judgement rather than a measurement, the chart says so on its own source line.

On anonymity. The client is not named, and neither is any competitor, partner or technology provider. That second decision is deliberate and goes further than usual. In a field of thirteen companies, naming twelve of them identifies the thirteenth, and the thirteenth is the client. The findings are reproduced in full; only the names are withheld.