Honest Defaults for Buyers

Published October 23, 2026 · Last reviewed October 23, 2026

Durable practice (not new). This is established communication practice, not a discovery or a trend.

What it fixes

A default that benefits you at the buyer’s expense and works because nobody notices it: the ticked insurance box, the renewal with no reminder, the pre-selected top tier. The buyer ends up with something they did not choose and learns about it from a statement.

Tactics that fail the disclosure test, and the version that passes:

• Pre-ticked paid add-ons. Passing version: add-ons that cost extra start unticked. A free, recommended setting may be preselected if you say so.
• The most expensive tier preselected. Passing version: preselect the tier you would recommend for the need the buyer has told you about.
• A trial that turns into a paid plan without a reminder. Passing version: state the date and the amount at sign-up, and email before it happens.
• Cancelling takes more steps than joining did. Passing version: the same number of steps, in the same place.

How to run it

1. List every preselected choice in your checkout, sign-up, renewal and notification settings: ticked boxes, default plan, default billing period, auto-renewal.
2. For each one, decide which option you would recommend if this buyer asked you in person, and set that as the default.
3. Say what the default is in plain words beside it, and give the reason you chose it.
4. Make changing it one clear step, in the same place the default appears, with no extra page, form or phone call.
5. Send a reminder before anything renews or charges, and make cancelling take no more steps than signing up did.

“I have set this to annual billing with a reminder email two weeks before it renews, because that is what I would recommend for a team your size. Switch to monthly here, in one click.”

Say it out loud

The default announced with its reason:

“This is preselected because I think it suits you. You can change it right here.”

It still works when the buyer knows exactly what you did, because a stated default carries your recommendation, and the meta-analysis below found defaults to be more effective when they are read that way. A default that only works while the buyer overlooks it is the pre-ticked box, and it does not belong here.

Profile translations

• Profile D At checkout, leave paid add-ons unticked and preselect the plan that matches the need the buyer entered. In the renewal email, show the date, the amount and the cancel link in the first three lines.
• Profile B Proposing a decision to your team: “Unless someone objects by Thursday noon, we go with option A. To object, reply with one sentence.” The default is stated, dated and cheap to decline.
• Profile C Newsletter or community sign-up: leave the marketing-email box unticked. If you set a default frequency, name it: “Weekly digest by default; change it here.”
• Profile E Petition or sign-up forms: leave “share my details with partner groups” unticked, and default to the option that shares least.

How to tell it worked (within 24 hours)

Action

Ask three recent customers which options were preselected on their order.

Why

A default the buyer can name was a default the buyer saw.

Verification

They name the preselected options correctly without opening the page.

Action

Time a cancellation from the account home page, using a colleague who has never done it.

Why

Opting out should cost no more effort than opting in did.

Verification

They finish in one pass and in no more steps than signing up took. Profile D

Action

Tally this week’s support messages that say the sender did not realise about a charge or a renewal.

Why

Those messages are the buyer reporting a default they did not see.

Verification

The tally is lower than last week, and each reply you send names the default and how to change it.

Why it works, and how strong the evidence is

Jachimowicz, Duncan, Weber and Johnson (2019, Behavioural Public Policy) pooled studies in which one option was preselected. Defaults had a considerable influence on average, but most studies found positive effects, several found none, and two found negative effects. Two factors partly explained the spread: defaults were more effective in consumer settings than in environmental ones, and more effective when they operated through endorsement, meaning people read them as what the choice designer thinks they should do, or through endowment, meaning they reflected the status quo.

That second factor is the reason for steps 2 and 3. A default you can state as a recommendation has a mechanism that survives disclosure. One that depends on being overlooked is not what the endorsement route describes.

Johnson and Goldstein (2003, Science) compared effective consent rates for organ donation across European countries with opt-in and opt-out rules and found large differences. Their follow-up (2004, Transplantation) reports a smaller difference in actual donation numbers after controlling for other national differences, and points out that every choice of default has costs on both sides. The comparison across countries is observational, so it cannot show that the default alone caused the gap.

A default changes what people pick. Whether the pick serves them depends on the default you chose, which is why step 2 starts from a recommendation. For a catalogue of how defaults are misused in shops, see Mathur and colleagues (2019): in a crawl of about 11,000 shopping sites they recorded instances of items added to carts through preselected checkboxes, of subscriptions charged under the pretense of a one-time fee or free trial, and of pre-selected costlier variations.

Worked example

An illustration, with invented figures.

The after version starts the paid add-on unticked, tells the buyer the date, the amount and the exit before the money moves, and puts cancelling in the same place as joining. Watch for drift: a later redesign that moves the cancel link or reorders the boxes reintroduces the first version without anyone deciding to. Re-run the three-customer check after each checkout change.