When Product Images Beat a Configurator

This guide argues against a purchase we make, which is worth saying at the start so the reasoning can be judged on its merits rather than taken on trust.

A large share of brands who ask for a configurator are better served by a set of product images. Not adequately served, better served, on the measure that matters, which is whether customers understand what they are buying and complete the purchase.

The reason is structural. A configurator is a system that has to be operated, maintained and kept true to a catalogue that changes. A set of images is finished the day it is delivered and never lies afterwards, because it never changes.

Count the combinations first

The single decisive number is how many valid configurations actually exist, and it is usually smaller than the marketing description implies.

A product with three finishes, two sizes and two hardware options has twelve combinations. Twelve images cover every one of them, exactly, forever, with no rules engine, no platform licence and no maintenance obligation.

The arithmetic stays favourable further than people expect. Even fifty or eighty combinations can be pre rendered once, and the cost of doing so is frequently below the first year of a configurator project.

The case for a configurator strengthens when the count runs into hundreds or thousands, which is genuinely common in furniture and industrial products and genuinely rare elsewhere.

Ask whether any combination is invalid

The expensive half of a configurator is the rules engine, and it is only doing work if there is something to prevent.

If every option combines with every other, there are no rules. The configurator is a viewer with menus, and a set of images with the same menus does the same job at a fraction of the cost and none of the maintenance.

If some combinations cannot be manufactured, the rules engine is preventing customers from ordering things that do not exist, which is a real and valuable function that images cannot perform.

That single question separates the two purchases more reliably than any other, and most brands can answer it in a minute.

Ask who is guiding the customer

Configurators assume self service, and a substantial share of products are not sold that way.

In a consultative sale a person is already interpreting the product for the customer, answering questions, steering toward what fits. That person is far more effective with a strong image set they can navigate quickly than with a configurator the customer drives.

Worse, a configurator in a consultative sale frequently goes unused, because the salesperson has a faster way to get to the same answer and reaches for it.

The test is whether anybody is in the conversation. If somebody is, images and their expertise usually beat self service configuration.

Ask how fast the catalogue changes

This is the question that decides whether a configurator stays an asset or becomes a liability, and it is rarely asked before purchase.

A stable product line changes slowly and a configurator built for it stays accurate with modest attention.

A line that turns over seasonally, discontinues finishes regularly and revises pricing frequently will drift out of date, and a configurator that shows a discontinued option lets a customer configure something they cannot buy, which is worse than showing nothing.

Images have the same staleness problem and a much lower cost of correction, because updating an image set is a production task rather than a system change.

The maintenance nobody budgets

A configurator is not delivered, it is operated, and that distinction is the largest hidden difference between these two purchases.

Somebody inside the business owns keeping it true. Platform updates arrive. Browsers change. New products appear and old ones vanish. None of that is optional and all of it recurs.

A set of images has none of those obligations. It sits on a server, works in every browser that will ever exist, needs no licence and requires attention only when the product changes.

For a brand without the internal capacity to own a system, that difference is not a preference. It is the whole decision.

The cost comparison, stated plainly

It is worth putting numbers of the right shape next to each other, because the two purchases are usually compared on the wrong axis.

An image set is a production cost with a definite end. It is quoted, produced, delivered and finished, and the only recurring expense is producing more when the product changes.

A configurator is a production cost plus a licence plus an integration plus a permanent internal role. The first invoice is the smallest part of it, and the parts that follow arrive every year for as long as the system runs.

Comparing the first invoices makes them look closer than they are. Comparing three years of total cost of ownership makes the difference obvious, and that is the comparison a business case should contain.

What happens to each when the product changes

Product change is certain, and the two purchases respond to it very differently, which is worth thinking through before choosing.

When a finish is discontinued, an image set becomes partly out of date and continues working. Somebody removes or replaces the affected images when they get to it, and nothing breaks in the meantime.

When the same finish is discontinued in a configurator, a customer can still select it and configure a product that cannot be manufactured, which produces an order the business has to unwind and a customer who was told they could have something they cannot.

That asymmetry is the strongest practical argument for images at any brand where nobody has clear ownership of keeping systems current, because the failure mode of a stale image set is mild and the failure mode of a stale configurator is not.

What images do better than configurators

Atmosphere and desire, which is what most consumer purchases actually turn on. A lifestyle image of a product in a beautiful room does something no configurator does, because the configurator shows the object and the image shows the life.

Speed, because an image loads instantly and a configurator loads eventually, and a meaningful share of visitors leave during the difference.

Reliability, because an image cannot break, cannot produce an invalid state and cannot show a wrong price.

And reach, because images travel into advertising, email, catalogues, print, marketplaces and retail partners, while a configurator lives on one site.

What configurators do better than images

Handling combination counts that cannot be pre rendered, which is the honest core of the case and the only one that is unarguable.

Preventing invalid orders, which for a manufacturer is a real cost saving rather than a customer experience improvement.

Producing a structured specification that reaches a quoting system, which eliminates re typing and the errors that come with it.

And giving a customer agency, which for certain products and audiences genuinely increases the likelihood of purchase because they built it themselves.

The middle option almost nobody considers

The debate is usually framed as images against a configurator, and there is a third product sitting between them that suits a surprising number of brands.

It is a guided selector: a small set of questions that narrows a customer toward the right product, ending on a pre rendered image of the specific configuration they arrived at. No 3D, no rules engine, no live rendering. Just a decision path and a library of images at the end of it.

It gives the customer the sense of agency that makes configurators persuasive, it handles a combination count in the dozens or low hundreds comfortably, and it carries none of the platform, integration or maintenance burden of a real configurator.

It is also considerably cheaper to build and to change, because updating it means replacing images rather than reopening a system, and it can be built on the same image library the brand needed anyway.

What the analytics tell you afterwards

Both purchases produce data and the data answers different questions, which is worth knowing before choosing rather than after.

An image set tells you which variants people looked at, which is useful and shallow. It reveals popularity and nothing about intent.

A configurator tells you what people assembled, abandoned, returned to and nearly bought. That is genuinely valuable product intelligence: which combinations customers want, which options nobody selects, and where in the decision they give up.

For a manufacturer deciding what to produce next, that behavioural data is sometimes worth more than the sales the configurator generates, and it is an argument for configuration that has nothing to do with conversion. It is also the argument least often made, because it benefits product rather than marketing and marketing usually owns the budget.

The hybrid that suits most brands

The framing that resolves this best is not one or the other, it is images for the many and configuration for the few.

Most brands have a small number of genuinely combinatorial products and a long tail of simple ones. The simple ones are covered completely by images. The complex ones justify a configurator.

Building the component library for the complex products also produces the images for them, since the library renders both, which means the two are not competing budgets so much as sequential outputs of the same investment.

That sequence, library first, images immediately, configuration when justified, is the arrangement that wastes the least.

A test that costs almost nothing

Before committing to a configurator, publish an image set covering the main combinations and watch what customers do.

If they engage with the variants and convert, images were sufficient and the configurator would have solved a problem that did not exist.

If they consistently ask for combinations that are not shown, the combination count is genuinely too large for images and the configurator case is made with evidence.

That test costs a production run and answers a question that otherwise gets settled by conviction, which is the most expensive way to settle anything in this category.

The internal politics of this decision

Worth naming because it affects the outcome more than the analysis does.

A configurator is a visible, modern, impressive project that somebody can present internally. An image set is unglamorous and reads as maintenance rather than as innovation.

That asymmetry pushes decisions toward the configurator regardless of what the product needs, and the people who would push back frequently lack the standing to.

The way through it is not to argue against ambition but to insist on the combination count and the maintenance owner, both of which are factual questions with answers that settle the debate without anybody having to lose an argument.

When to revisit the decision

When the catalogue grows enough that the combination count crosses out of what images can cover.

When customers start asking, repeatedly, for combinations that are not shown.

When the sales process moves from consultative to self service, which changes who is interpreting the product.

And when somebody internal takes ownership of the catalogue as an actual role, because that removes the constraint that made a configurator inadvisable.

The honest summary

Count the valid combinations. Ask whether any combination is invalid. Ask who guides the customer. Ask how fast the catalogue changes. Ask who will own it.

If the combinations are few, everything combines with everything, a person is guiding the sale, the catalogue turns over quickly, or nobody will own it, buy images and spend the difference on making them excellent.

If the combinations run into the hundreds, some are genuinely unmanufacturable, the sale is self service, the line is stable and somebody will own it, a configurator earns its cost and images will not substitute.

Most brands sit in the first group and buy from the second, and correcting that is worth more than any negotiation on either purchase.

Frequently asked questions

How many combinations justify a configurator?

Below a few dozen, images cover every one exactly, forever, with no licence or maintenance. The arithmetic stays favourable even at fifty or eighty. The case strengthens when the count runs into hundreds or thousands, which is common in furniture and rare elsewhere.

What is the single fastest way to decide?

Ask whether any combination is invalid. If everything combines with everything, the rules engine has nothing to prevent and the configurator is a viewer with menus, which images plus menus deliver for a fraction of the cost and none of the maintenance.

Do configurators work in consultative sales?

Usually poorly. When a person is already interpreting the product for the customer, that person works faster with a strong image set they can navigate than with a configurator the customer drives, and they quietly stop using the configurator.

What do images do better?

Atmosphere and desire, which is what most consumer purchases turn on. Speed, since they load instantly. Reliability, since they cannot break or show a wrong price. And reach, since they travel into advertising, email, print and marketplaces while a configurator lives on one site.

Is there a way to have both?

Yes, and it is the efficient path. Build the component library, which renders both the images and the configuration. Cover the simple long tail with images, and add configuration for the genuinely combinatorial products once the combination count justifies it.