A specialized industrial product earns its place when it solves a customer's process problem repeatedly, not when it merely adds another unit to a catalog.
The customer buys a process that works
A manufacturer may need a weld to hold under vibration, an adhesive to cure at a particular speed, a food line to clean between batches, or a packaging machine to seal thousands of units consistently. The purchased item is only one part of that result. The customer also needs the right specification, application advice, availability, training, and confidence that a replacement will fit the process already in place.
Illinois Tool Works (ITW) operates across seven industrial segments and describes its businesses as serving distinct end markets with specialized products. Its company overview reports operations in 49 countries and approximately 43,000 employees, but the scale alone does not explain why an industrial customer would stay. The relevant unit is a problem solved inside a customer's production system.
80/20 began as a factory discipline
ITW's 2025 Form 10-K says its 80/20 Front-to-Back Process began as a manufacturing-efficiency tool in the 1980s and was later expanded into a company-wide business-management process. The filing describes focusing on the largest and best opportunities while reducing cost, complexity, and distractions associated with less profitable opportunities. That is a resource-allocation rule, not a claim that every small customer is unimportant.
In practice, a division can examine which customers generate the most useful combination of volume, margin, technical challenge, repeat demand, and relationship depth. It can simplify a product range, alter sourcing, standardize a process, or stop serving a low-return configuration. The intended result is not fewer products for its own sake. It is more attention available for the applications where the division can provide something difficult to replace.
The rule also creates a boundary. A low-volume customer may depend on a product that is critical to a larger system, or a niche may contain the next technical opportunity before its revenue is visible. A financial ranking can identify where money is earned; it cannot by itself establish where a process would fail if support disappeared.
Innovation starts at the customer's machine
ITW describes customer-back innovation as beginning with difficult customer problems rather than with an internal research list. The company says its divisions partner with key customers to solve technical challenges and reports approximately 21,800 granted and pending patents through 2025. The source supports the method and the patent count; it does not establish that every patent creates a durable advantage.
Working from the customer back changes what must be learned. An adhesive must be tested on the customer's substrate and line speed. A welding solution must work with the operator's equipment, material, and safety requirements. A food-equipment improvement must survive cleaning, maintenance, and production scheduling. The design is not complete when the laboratory sample works; it is complete enough when the customer's process can use it repeatedly.
That history can become a form of protection. A rival may copy the visible product but still lack the application data, qualification work, field service, and trust that let the customer change a live process without unacceptable downtime. The protection is real but conditional: it has to be renewed as the customer's materials, regulations, equipment, and production targets change.
Decentralization keeps knowledge close
ITW's Business Model has a third element: a decentralized, entrepreneurial culture. Divisions can customize how they apply the common process to their customers and end markets. ITW calls this flexibility within a framework. The framework provides common language and expectations; local teams retain the knowledge of a particular machine, plant, or customer.
That arrangement helps when the useful decision is local. A division can change a product, service response, or sourcing route without waiting for a central office to understand every application. It can also preserve relationships that are not visible in a corporate average. But decentralization makes common learning harder. A successful solution in one division may not travel easily to another, and inconsistent systems can make it difficult to compare quality, service, or product economics.
Pruning changes what the company can still do
Portfolio decisions are physical decisions because they alter which tooling, suppliers, engineers, service people, and production lines remain available. If ITW exits a low-return product, it may free capacity and money for a more differentiated one. The customer who loses that product must then find another qualified source, redesign a process, or carry more inventory.
This is where the 80/20 method encounters its hardest test. A product can be small in ITW's accounts but large in a customer's failure consequences. Conversely, a high-revenue product can be technically replaceable and demand little specialized knowledge. The method can reveal concentration; it cannot replace engineering judgment about the function being preserved.
Organic growth requires money before the result
ITW's enterprise strategy began in 2012 with goals around growth, margins, and returns. The company says its Next Phase, launched in 2024, makes organic growth a priority by 2030. That shift asks the operating model to create new customer demand rather than only improve the existing portfolio.
Money must arrive before that result. A new application can require engineering time, customer trials, tooling, validation, inventory, and a sales team that understands the problem before revenue is certain. A division under pressure to protect short-term margin may postpone those steps even when the customer need is real. A customer may also be unwilling to pay for a trial until the product is proven. The 80/20 process can direct investment toward a promising opportunity, but it cannot remove the time between technical proof and commercial adoption.
Records measure the business, not the whole process
ITW can record revenue by segment, product margins, patents, delivery performance, and customer orders. Those records help decide where to allocate resources. They do not by themselves show whether a weld was applied correctly, whether an operator had the right training, or whether a substitute component will behave the same way after years of use.
The missing information often appears first at the customer's machine. A field engineer may see a recurring failure; a customer may change a material without telling the supplier; a division may discover that a cost reduction altered installation time. The correction requires a path back through sales, engineering, manufacturing, sourcing, and management. Decentralization can shorten that path when authority is clear, but a local observation still needs money and permission to become a product or process change.
The model compounds capability and trade-offs
ITW's durable pattern is not simply that it owns many industrial brands. It repeatedly narrows attention to important problems, develops solutions with customers, and lets divisions execute within a common operating method. That can turn small technical advantages into relationships that are difficult to displace.
The same pattern can also narrow the future. A customer that falls outside the profitable 80 may lose a source before another is qualified. A division that optimizes its own result may preserve local performance while making shared learning harder. The model remains useful when financial selection stays connected to the physical function customers need and when observations from the field can still reach the people able to change the next design, process, or support decision.
Inside CompanyGraph
View this perspective inside CompanyGraph.
CompanyGraph