Aftermarket Potential per Machine: Why the Revenue Gap Can Only Be Explained with Service Data

Management wants to know how much service revenue is generated by the installed base. Six weeks later, a presentation is on the table: a customer list, average figures by machine type, and a heat map based on machine age. Six months later, someone asks what happened to those opportunities. No one can say, because the list never made it into the system used by Sales and Service. Measuring the aftermarket potential of an installed base is fairly straightforward today. More difficult are the two questions that follow: Why is revenue missing for this specific machine? And who will handle this starting tomorrow, and in which system? Here’s a quick overview of what matters:

• Comparison group instead of overall average. Potential only becomes apparent when machines of the same type, similar age, and comparable market value are placed side by side.
• The revenue gap needs an explanation. Whether a customer buys parts elsewhere, there is no contract, or the machine has changed operators is only revealed through service calls, field visits, and contract status.
• The result belongs to the machine. The potential and rationale are listed as values in the machine file, where anyone who deals with the customer can find them.
• The value becomes an action. Opportunity, task, or campaign—initially a suggestion that a person approves.

Why One-Time Potential Assessments Fall Flat

The data describing a machine in the aftermarket is almost never stored in a single location. Spare parts invoices are stored in the ERP system, posted to the customer and rarely to the serial number. The operator and their contacts are stored in the CRM. Service calls and malfunctions are documented by the service software, while the delivery configuration is stored in the PLM. Each system has its own number ranges. A capacity analysis compiles this data all at once, usually involving a great deal of manual work. The result is a snapshot of the situation as of a specific date. After that, new orders come in, machines are sold or relocated, and contracts expire. After three months, the list is no longer accurate, and updating it would require the same manual work all over again. Yet the importance of this business is undisputed. In the machinery and equipment sector, service accounts for more than a quarter of revenue, and while product revenue may plummet by about 30 percent, service revenue falls by only 5 to 8 percent (Roland Berger / KVD 2026). Nevertheless, the installed base remains a blind spot in many companies: known in aggregate, but unknown on an individual machine basis. That is why even the best one-time analysis is wasted if its results are not permanently attached to the machine.

Measuring Aftermarket Potential: Comparison Groups by Machine Type, Age, and Market

Taking an average across the entire installed base is of little help. A three-year-old machine has different needs than a twelve-year-old one, and a packaging line in Poland has different needs than the same line in Switzerland. The comparison only becomes meaningful when made within a comparable group: same machine type, same age range, comparable market. Within this group, revenue per machine is compared by category, such as spare parts, service calls, maintenance contracts, upgrades, and rebuilds. The difference from the group average represents the calculated potential. A typical pattern often emerges: Newer machines generate revenue primarily through spare parts and service, while older machines increasingly generate revenue through rebuilds, upgrades, and modernization. The calculated potential represents an upper limit. Some operators have their own maintenance teams with manufacturer training; some machines operate on a single shift and wear out more slowly than the group average. Therefore, every discrepancy requires a plausibility check before it is turned into a work order. This check is performed by the service data in the next step. This process requires a fleet. With 2,000 machines of a given model in the field, the comparison groups are valid; with twelve individually built systems, they are not. For mass-production manufacturers with a large installed base, this is therefore the obvious starting point.

What data is required for this

Gaps should be clearly identified. Machines with an unknown year of manufacture should be grouped separately, rather than using estimated values that skew the average. Often, this group itself is indicative of a problem: When the year of manufacture is missing, contact with the current operator is usually missing as well.

Explaining the Revenue Gap: What Service Data Reveals

The sales data show that a machine is underperforming compared to its peer group. However, the data do not reveal the reason. That reason can be found in the service cases, the operational report, the contract status, and sometimes in who last reported on the machine. Three patterns repeatedly emerge in practice.

Aftermarket parts, yes; genuine parts, no

A typical scenario: The machine requires regular service calls and maintenance visits, but spare parts sales are well below the group average. So the machine is running and needs parts, but those parts are coming from elsewhere. Possible causes include aftermarket parts, the operator’s own inventory, or orders placed through another company. The logical next step is to talk to the customer, often followed by a spare parts package with a availability guarantee.

Old machine, incidents are piling up, no contract

The machine is more than ten years old; there have been several incidents involving the same assembly in recent months, and there is no maintenance contract in place. A rebuild or upgrade offer would be appropriate here, or alternatively, a maintenance contract. The rationale is already documented in the case history, making the offer clear to the customer.

The serial number appears when a new operator takes over

A serial number is associated with an inquiry or order from a company that is not listed as the owner in the file. The machine has been resold. The new operator often has no documentation and no relationship with the manufacturer. For the aftermarket, this represents a new customer in its portfolio, with a need for inspections, training, and spare parts. What all three patterns have in common is that they only become apparent when revenue, service cases, and contracts are linked to the same machine. With revenue lines alone, one can only conclude that something is missing.

From Opportunity to Transaction in the Service Process

To ensure that a finding doesn’t just end up as a slide, it is saved in the machine’s record: the potential per category and the rationale. In the Digital Machine File (IOTAM), each machine has a data record that includes the model series, configuration, operator, contracts, and service history. Because the machine file is stored in Salesforce, opportunities, contracts, service cases, and service calls are already linked to the same customer and, in most cases, to the same machine. This means the potential ends up right where sales and service teams are already working. Based on this value, specific actions are generated through rules. If a machine is significantly below its group’s average and does not have a contract, the service sales team is assigned a task. A clear candidate for a rebuild becomes an opportunity. Many similar cases in a market result in a segment for a service sales campaign. The person in charge is the Service Sales representative who already handles the customer. They identify the potential when the account is opened, along with any open cases and active contracts. A central office maintains the rules and reviews them once a quarter to determine which ones are effective and which ones merely create extra work. Service Decision Intelligence (SDI) provides the explanation. SDI links a machine’s revenue gap to its case, service call, and contract history and supports each recommendation with a source citation—that is, with the specific data records on which it is based. SDI runs on the customer’s infrastructure, is EU-compliant, and is independent of the language model. The same questions can also be asked via Claude, Agentforce, or other AI interfaces without copying the data. Automation is automated in stages. Initially, the system makes recommendations, and a human in the service sales department approves them. Only once it becomes clear which rules consistently apply do individual processes run without approval. For quotes sent to customers, approval is generally still required. Because every process is tied to a machine, it’s possible to measure the impact: identified potential, processed orders, and new business won each quarter. This also makes it possible to answer the question of what has changed after one year. The effort is worth it: Manufacturers with a proactive, contract-driven service model generate an average of 48 percent of their EBIT contribution from service, while those with a reactive approach generate about 15 percent (Roland Berger / KVD 2026). More on these figures can be found in the article “Service in Machinery and Equipment: The Underestimated Business.”

Where to Start

The process follows these four steps: Digitize → Connect → Decide → Automate. Digitization means that every machine is assigned a clean data set. Connecting means that revenue, service cases, and contracts are linked to this data set. Decide means that a comparison group and an explanation show where and why revenue is missing. Automate means that proven rules give rise to processes that require no manual intervention. The greatest effort is required in the first two stages, and that’s where it pays to start small: a product line with many machines in the field and a solid data foundation. That’s where the comparison groups come into play, and the first patterns emerge quickly. The article on the installed base as a growth driver describes how this fits into an ongoing service organization. Two pragmatic approaches:

• Installed Base Assessment – when machines, operators, and revenue are currently scattered across ERP, CRM, and Excel. The assessment shows which data is sufficient for comparison and where the mapping breaks down.
• introductory call – once the data is in place and you want to determine which series and rules the pilot will use to get started.

Start with the model series that has the most machines in the field. That’s where the potential is greatest and the comparison is most reliable.