Configure-to-Order, CPQ
Stefan Jungen
27. July 2026
"The right product is not always the 'right' product.”
Anyone selling complex, highly configurable products is ultimately also selling a promise: that what the customer receives is actually what they meant to buy. The key question, therefore, is not simply how to generate a quote as quickly as possible, but what determines whether a customer trusts a supplier in the first place. The central idea is this: trust does not begin when the machine is delivered, but much earlier — at the moment when configuration and quoting take place. What the customer sees is what they get.
This article explains why CPQ systems alone do not fully solve the trust issue, what role 3D visualization plays in it, and how companies can use process consistency to prevent technically correct quotes from still missing the customer’s actual needs.
CPQ stands for Configure, Price, Quote. Without such a system, the sales process for complex products typically runs through many manual loops: capturing the request, creating a technical concept, calculating the price, preparing documents, presenting the offer — and with every follow-up question, the cycle starts again. With CPQ, this process becomes automated or partially automated: configuration follows predefined rules, pricing comes directly from the system, and the quote is generated at the push of a button. The result is rule-compliant and manufacturable — the configurator does not even allow technically incompatible combinations.
CPQ is therefore used in two classic forms: as an internal quoting system, with which inside or field sales teams can quickly generate a reliable offer during a customer meeting or quote preparation, and as a self-service configuration system, in which customers or partners assemble what they need themselves. At first glance, these look like separate worlds, but in reality they operate on the same logic and the same data basis.
"The right product is not always the 'right' product.”
But this is precisely where there is a gap that sounds counterintuitive at first: The right product isn’t necessarily the right product. A configurator delivers a result that is technically completely correct—all rules are met, fully constructible, error-free. Still, it may not be what the customer actually wanted. “Valid” doesn’t necessarily mean equal to intended.
Complex products must meet hard requirements: a machine has to fit into the available installation space, integrate into the existing system, and match the specific application. The subtle problem arises when the configurator does everything right — based on the selected choices, the product is flawless. But an option may have been misunderstood, a requirement not translated properly, or a question answered incorrectly. The result is a technically perfect machine that is still the wrong one for the customer’s actual application. Because everything in the configurator looked correct, the error is often only noticed at delivery.
The customer ends up receiving something they did not really intend to order. Industry analysts describe this phenomenon as “buyer regret,” and it is costly: on the supplier side, it leads to rework and delays, and acceptance may be withheld. On the customer side, it leaves disappointment; in the worst case, the customer turns to a competitor on the next project even though everything was technically correct. The damage therefore affects precisely the relationship that the sales process was supposed to build — a risk that a structured quoting configurator can reduce from the outset through clear, rule-based guidance.
CPQ already reduces configuration errors significantly because embedded rules prevent technically unsuitable combinations from being created in the first place. But a gap still remains: a complex configurator asks dozens of questions about dimensions, options, performance values, interfaces, and dependencies. Customers or sales teams work through the process field by field — often seeing only options, parameters, and technical codes, perhaps a generic catalog image. As long as the specific configured product does not become visible, the final certainty remains open: is what is valid also what was intended?
This is exactly where a rethink is needed. Trust should not arise late in the sales process — the earlier, the better. With highly configurable products, the process begins with configuration, and that is exactly where trust must also begin so it can carry through into the quote and all the way to delivery.
In this context, visualization is not decoration, but trust infrastructure — it works live, at the very moment when decisions are made. This means what a person sees and recognizes as their own product: depending on the use case, the 3D model, the dimension drawing, the circuit diagram, or the rendering. When customers see their configured product, commitment arises exactly where it must arise — not weeks later at delivery. According to the text, this aligns with market observation: Gartner (2025) now treats 3D configuration for complex physical products as a baseline expectation rather than a nice-to-have.
For such an image to create trust at all, it must be real — meaning generated from the configuration rather than maintained separately. With highly configurable products, this only works in an automated way, because otherwise each variant would require its own manually maintained image — an impossible task when thousands of variants are involved. Only a parametric configuration mechanism can automatically generate the appropriate image for each variant. In this context, “real” does not necessarily mean maximum detail, but consistency with the real product: the live model is deliberately lean, but performant and correct where it matters.
The image the customer sees does not stand alone. The same configuration from which it is generated is also the source for price, quote, bill of materials, 3D model, CAD exports, drawings, and handoff to downstream systems. Because all outputs come from one source, they no longer drift apart unnoticed: a rule is maintained once, and all downstream outputs follow — without redrawing, without manual re-entry. That does not mean deviations become impossible; if a rule or interface is set up incorrectly, one output can still be wrong. What disappears with such an architecture, however, is the silent drift of separate copies that diverge over time. That is what consistency means in practice: outputs that evolve together instead of separately.
The switchgear cabinet configurator from Siemens shows how this principle works in practice: components are dragged and dropped into the interior layout and automatically snap to predefined anchor points; the 3D model updates with each component, and the cabinet can be viewed from all sides. From this one configuration, everything else then follows — bill of materials, pricing, quote request at the push of a button, and CAD exports in common formats such as STEP, DXF, and DWG.
Two developments will shape this trust question even further in the coming years. The first is self-service: B2B buyers are used to handling things online in their private lives, and a new generation of buyers increasingly expects the same in business. The obvious concern — that self-service makes sales obsolete — is too simplistic. Pure self-service tends to lead to costly wrong decisions for complex investment goods. Self-service with real 3D, by contrast, makes sales more valuable, because when customers have already configured and seen their product themselves, they come into the conversation with a concrete request rather than a vague idea. According to Gartner data cited in the text, most B2B buyers want both paths combined — self-service and personal advice.
The second development is artificial intelligence. AI will increasingly take over parts of the configuration process: parsing requirements, identifying the right product, proposing a configuration, and even optimizing price and quote in seconds. This speeds up the process considerably, but trust in AI-generated results is still developing. In B2B, a significant share of buyers still want AI results to be confirmed by a human before acting on them. The more AI takes over configuration, the louder the question becomes: is this actually correct? Real visualization provides exactly this confirmation, because the customer can immediately see whether the AI suggestion actually fits. AI will increasingly take over the doing, while visualization remains what creates trust.
Four terms summarize what matters: valid, intended, seen, received. Trust in B2B sales does not arise only in the quote or at delivery, but already during configuration — namely when the customer sees their product and recognizes it as what they actually need. A CPQ system alone guarantees validity, but not yet commitment; only continuous visualization generated directly from the configuration closes the gap between what is technically correct and what the customer actually wanted. The decisive factor is consistency: what the customer sees must remain the same product all the way through to manufacturing. Companies that want to move in this direction do not need to transform their entire portfolio at once — starting with a single product that delivers fast, visible value is enough before rolling the approach out step by step.
A CPQ system ensures that a configuration is technically correct and manufacturable — in other words, valid. But that does not rule out that a requirement was misunderstood or that an option was selected ambiguously. The result can therefore be technically flawless and still not what the customer actually wanted. Only a real visualization of the configured product closes this gap.
No — and for highly configurable products that would not be practical anyway. A parametric configuration mechanism generates the appropriate image automatically for each variant as soon as a selection is made, without every combination having to be modeled or maintained manually.
No. Pure self-service can lead to wrong decisions for complex, consultative investment goods. Self-service with real 3D visualization, by contrast, makes personal sales more valuable, because the customer enters the conversation with a concrete, visually validated request instead of starting with only a vague idea.
LUCA NAGEL
Project & Product Manager
This might also interest you
