Nudge your customers toward better choices

When car rental agencies include insurance unless you specifically decline it, or software vendors recommend that you click “next” for a quick install, they’re choosing default options for you — covertly or overtly guiding your choices. Well-designed product or service defaults benefit both company and consumer by simplifying decision making, enhancing customer satisfaction, reducing risk and driving profitable purchases. Ill-conceived defaults can leave money on the table, fuel consumer backlashes, put customers at risk and trigger lawsuits — costing companies dearly.

We’ve published academic articles, advised governments and served as legal experts on defaults. On the basis of this work, we’ve developed a primer for managers on which defaults companies commonly must consider and how to avoid the pitfalls of ill-considered defaults.


In all too many companies, management leaves the determination of default settings for a new offering to a forms designer or a programmer. But setting defaults is complex, requiring companies to balance an array of interests, including customers’ wishes and the company’s desire to maximize profits and minimize risk. Because of their powerful effect on customer behavior, default policies ought to be examined at the highest levels in the organization.

Let’s take a close look at the taxonomy of defaults, which we place into two broad categories: mass and personalized.


Mass defaults apply to all customers of a product or service, without taking customers’ individual characteristics or preferences into account. A common example would be an online retailer’s using standard shipping unless the customer actively chooses rush delivery.

BENIGN DEFAULTS. These represent a company’s best guess — absent preference information — about which product or service configurations would be most acceptable to most customers, and would pose the least risk to the firm and the customers. In seeking a good default for all customers, firms will find that no simple formula applies to all cases.

Companies can design mass defaults in ways to maximize profit without being deceitful or jeopardizing customer satisfaction. Many auto manufacturers, for example, set their online order-taking software to present customers with the least expensive, stripped-down models of their cars. Customers then upgrade features one by one. In our research experiment with a leading European carmaker’s customers, we found that when we changed this default so that customers were initially offered cars loaded with features — options they could then decline — they chose vehicles with more features, raising the sales price by over $1,500 without decreasing customer satisfaction.

RANDOM DEFAULTS. When customers are assigned arbitrarily to one of several default configurations, the selection is called a random default. As experimental tools, random defaults can help companies to uncover customers’ preferences, allowing them to migrate from using mass defaults to creating personalized ones.

For example, when online marketers send e-mail, they must choose HTML or plain text as the default setting. In the absence of information about customers’ preferences, many companies opt for a best-guess benign default. To make a more informed default choice, a firm could randomly send half its e-mails as text and half as HTML, with links that allow recipients to switch. By monitoring how many people switch from each default, and determining which browsers and operating systems they are using, the company can make informed guesses about which default new prospects are likely to want.

HIDDEN OPTIONS. When a company uses hidden options, the default is presented as a customer’s only choice, although hard-to-find alternatives exist. Dell sells computers with either Windows or Linux operating systems, though the Linux option does not appear in the main product configurator, where most customers select the features they want; it’s accessed only through an obscure link on the site.

Often hidden options are a simple expedient for companies and cause no harm to them or their customers. But when hidden options appear to intentionally obscure choices that customers may value, companies risk a customer backlash. Dell’s customers haven’t responded angrily to the hidden Linux option, but as of September 2008, nearly 30,000 Dell customers had cast votes on the company’s Web site, urging the firm to add Linux to the main list of operating system options.


Personalized defaults, as the label implies, reflect individual differences and can be tailored to better meet customers’ needs. Let’s examine three of the most common types.

SMART DEFAULTS. These defaults apply what is known about an individual customer or segment to customize settings in a way that is likely to be ideal for the customer and the company — or at least is a better fit than a mass default would be. The data that smart defaults use include demographic or geographic variables, or even measurements taken by a product itself.

For example, smart defaults can fill in the country code on an online order form based on the customer’s IP address. What makes such configurations defaults, and not simply customized settings, is that customers can opt out — for instance, by rejecting smart default investment allocations in favor of a different mix.

PERSISTENT DEFAULTS. When an airline automatically assigns aisle seats to customers who have previously chosen them, it is using a persistent default. The assumption is that a customer’s past choices are the best predictor of future preferences.

Using persistent defaults whenever possible is an easy way to enhance customer satisfaction and loyalty. The principal downside is that customer preferences can change, and the persistent default can become an annoyance. A sensible work-around is to make opting out simple and transparent, or to use reverting defaults, which automatically change back to a mass default setting after some period of time.

ADAPTIVE DEFAULTS. Adaptive defaults are dynamic: They update themselves based on current decisions that a customer has made. This is particularly helpful in online environments, where a customer makes a sequence of choices. Adaptive defaults can serve as advisers, helping people identify sets of features that they will probably want, based on the preferences of other consumers in the company’s database.

Personalized defaults are a good choice when customer information is readily available, as they’re more likely to satisfy customer needs. But companies need to be aware that they, too, can sometimes backfire. For example, personal video recorders such as the popular TiVo system make use of the owner’s history of recorded TV shows to automatically record shows it deems the owner might like. While pleasing most users, this personalized default has also led to customer complaints when television recorders have mistakenly selected content some owners found objectionable.

It takes careful research and experimentation to align defaults with both a company’s and its customers’ long-term interests. Firms that manage defaults strategically and ethically can expect to be paid back with loyalty and trust.

(Daniel G. Goldstein is an assistant professor at London Business School. Eric J. Johnson is a professor and director of the Center for Excellence in E-Business at the Columbia Business School in New York. Andreas Herrmann is a professor and director of the Center for Business Metrics at the University of St. Gallen in Switzerland. Mark Heitmann is a professor at Christian-Albrechts University in Kiel, Germany.)



Sometimes companies require would-be customers to make active choices or be denied use of the product. We call that a forced-choice alternative to defaults. For example, software customers often have to accept a user agreement before proceeding with an installation. Such contracts are also common in many recreational settings: Vacationers who want to go white-water rafting or horseback riding, for example, typically must sign a liability waiver in advance — or forgo the activity.

Forced choice is appropriate when it is wiser for the company to deny access to a product or service than to accept the potential costs that customers who fail to agree to the terms of use might generate. Forced choice may also make sense as an alternative when the best choice of default isn’t obvious. For example, some carmakers, such as BMW, require customers to actively pick an engine or exterior color without any guidance from defaults.

But forced choice has a downside: A customer may make ill-informed choices when first engaging with a product. (Who actually reads the text of a license agreement?) The customer might even be scared off by the requirement that he make a choice at all, or because he doesn’t understand the options.

Whether forced choice or a default setting is the better strategy depends on the company’s goals. Forced choice may increase the risk that customers will forgo or regret a purchase (research shows that the act of choosing raises the likelihood of post-purchase regret), whereas defaults can alienate some customers by making a product seem very different from what they had in mind.