I've noted that sometimes convenient features are too far from users' expectations or habits. For example, I've noted that http://small-improvements.com uses a form without a save button and they explain it like this:

Once you've entered your names, just leave this page, no need to save anything. Your manager will then approve your selection. You may modify your the list until either the time to pick users is up, or until your manager has approved your list.

With no Save button on the form, it took me several minutes to figure out what to do next. Usually I don't read all text on a page (especially at the bottom).

What do you think is better - to exceed users' expectations and provide new features or stick to what users already know?


I don't think it's an either/or question. By all means introduce new features, but give the user feedback about what's happening. For instance, I recently designed a survey where the user just had to select their answer, and it would advance to the next question. The question advancing gave the user feedback that their answer had been accepted, and they could return to that question by clicking the previous questions number (shown in a tab). So the problem with the Small Improvements site is a lack of feedback, not that trying to do something new is a bad idea.

  • +1 for "the problem with the Small Improvements site is a lack of feedback, not that trying to do something new is a bad idea." May 13 '14 at 23:07

Definitely stick to the users knowledge. The brain makes connections when we learn something. There is sometimes no need to reinventing the wheel. If it's a simple feature, stick the basic principles. Humans learn from repeating actions a lot or making connections from something they know. People use schemata to store information, if you change the schemata, the user will take longer to use your site and the abandoned rate will be higher.

Additionally, even though the application saves the content without any interaction of the user, we are used to this 'reassurance' or confirmation step that says :Thank you for sending this form'. If users don't get feedback, it's possible that they feel really lost. Specially the non- savvy ones.

In some cases it's good to provide new ways to do flows, but as long as the user understand quickly what needs to be done - intuitive. If a 300 word explanation is required, then the feature doesn't work.

  • Yes, but you don't need to revert to the "known" way of doing things either. In this particular case bringing up a notification when stuff has been saved would be enough not to have to read the 300 word explanation. Jun 1 '13 at 16:11

One way to attempt to measure this is to consider the return on investment. If a new system is more efficient, but the increase in efficiency is outweighed by the extra time needed to learn the new system, the time spent learning it can be a net loss.

For instance, if a sign up form is designed that, once learned, is more efficient than a more conventional sign up form, the investment in learning the new system will not be returned because the sign up form is only used once. But with something used many times, like an email system, the time spent learning a more efficient one could very well have a favorable payoff.

These efficiency vs. convention choices have been around a long time, e.g. the dvorak vs. qwerty keyboard choice.


I am a firm believer in trying something new and somehow break the trends others are following. However, when it comes to user experience of a website we’ll have to stick to the basics and find the scope of improvement gradually. For instance, forms with ‘Autosave’ feature will be a good start. I mean there’s no harm in trying something new, till it’s working fine!

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.