Which options should a product configurator preselect?

Choose the starting configuration deliberately. Explain what is already selected and what an alternative changes. A preselected option is part of the selling decision, even when the customer can replace it.

Herrmann and colleagues (2011) used a simulated racing-bike configurator. The primary abstract reports that moving a feature default toward a higher-priced level shifted selections, including toward options near the default. That supports treating the starting option as influential rather than neutral.

The accessible publisher material does not provide a verified effect size, full comparison table or uncertainty interval. It cannot tell you how many extra paid orders an upgraded wheelset would produce, or which default is best for your customers.

A starting option gives shoppers a point from which to compare alternatives. It may also look like the seller’s recommendation. Those are explanations consistent with the study; they do not establish that an expensive option is suitable.

The full results were not recovered after checking the publisher, an author’s publication list and institutional records. This module is therefore conditional. The simulated bike setting does not establish a paid-sales gain, and its reported sales-maximizing recommendation is not a universal rule for customer fit.

When testing a default in an early step, follow every later selection and the final configuration. A result on the defaulted component alone can miss a change elsewhere.

Donkers and colleagues (2020), Experiment 3, analyzed 1,566 adults making an imagined sofa configuration. Participants chose a frame and then fabric; one group saw the higher-price frame preselected, while another made that same first choice without a default.

The later higher-price fabric was chosen by 68.9% with the frame default versus 77.1% without it, a statistically clear difference in the reported analysis (p = .003). Subtracting the rates gives 8.2 percentage points: about eight fewer people per 100 chose that fabric in this imagined task. It does not mean an 8.2% fall in store sales. The first frame choice itself showed no statistically clear difference.

Accepting a preselection may involve less deliberation than choosing the same option actively. The authors’ explanation is that this changes what people carry into later choices. The useful operating lesson is to measure the sequence.

Participants did not buy these sofas. The result cannot predict paid orders or profit, and a result without a clear difference on the frame does not prove identical preferences. Your products, customers and sequence may produce a different pattern.

If a default is intended to build understanding or repeat preference, measure a later decision without that default. Acceptance today does not establish what the customer will choose actively tomorrow.

Kuhn, Ihmels and Kutzner (2021), Experiment 1, studied 90 mostly student participants choosing organic or conventional groceries. Each shop contained ten item choices; the first used preselection and the second required active choice. Participants could win a selected basket and unspent money in a raffle.

In the first shop, an organic rather than conventional default increased the adjusted organic share by 21.7 percentage points (standard error 4.77; p < .001). Applied to a ten-item basket, that is about 2.2 additional organic items on average, an arithmetic explanation of the estimate. In the later active-choice shop the estimate was 3.35 percentage points, with no statistically clear difference (p = .53). That uncertain result does not prove zero carryover.

The interface can steer a choice without building an enduring preference. Separate an immediate response to a setting from what a customer understands or chooses when the setting is removed.

These were incentivized experimental baskets rather than ordinary paid orders, mostly among students. A short sequence of shops cannot establish long-term habits. The finding does not tell a retailer whether a default teaches product knowledge; test understanding directly if that is your objective.

Treat an optional paid extra separately from the components needed to make a usable product. Test a clear active choice that shows the extra charge; assess understanding and reversals alongside journey speed.

A 2024 UK government-commissioned study by Alma Economics, Research into Potentially Harmful Online Choice Architecture, randomized 5,889 participants across simulated shipping and protection-plan journeys. This is government research, not a peer-reviewed journal paper. Participants received neither the chosen products nor an outcome-dependent payment.

Appendix Table A5.6 reports that preselected paid extras reduced journey time by approximately 0.24 minutes versus no default in each task after accounting for the product. Multiplying 0.24 by 60 gives about 14 seconds less. That is time saved in a simulated journey, not proof of more paid orders, better customer fit or informed agreement.

A shopper can move quickly because a decision was made for them. That makes speed and understanding different questions. Checking the final selection and the customer’s explanation helps distinguish convenience from an overlooked charge.

The task involved no real product purchase, and the time estimates have sampling uncertainty. The report’s choice-effect headline and its statistical tables use inconsistent percentage language, so no add-on selection uplift is quoted here. This study cannot establish your legal obligations or an optimal paid-extra policy.