23 March, 2016

Minimize Errors in Mobile App Forms Using Interaction Design Patterns

Majority of errors on mobile apps occur on input forms. Users can input information with utmost ease and flexibility, if only input forms were designed better. In an old post ‘Contextual Keypad – Engaging Users via Mobile Inputs’, I highlighted the need for displaying context-sensitive keypad based on the input field type – alphabetic, numeric, email, password and so forth. A little after that, I found Luke Wroblewski’s work on design patterns that helps reduce form errors significantly. In this post, I will highlight Luke's three interaction design patterns that help reduce errors form fields:
Inline Validation
Inline Validation provides real time feedback as users enter values into corresponding fields. Let’s consider an example. On Dashlane iPhone app, if a first time user creates an account, the password field has a placeholder text, ‘Enter a strong password’. As user types the input, the app provides real time instructions on what rules the password field should comply with. This helps users to enter password without forgetting to enter a specific value or making a mistake.
Password Hints on Dashlane mobile app
Luke Wroblewski’s team ran a study by comparing forms with inline validation with those which do not have any inline validation. The forms with inline validation had 42% completion rate and a 22% decrease in errors. This shows how a small step towards good user experience helps apps succeed in the real world.
Input Types
Specifying input types helps reduce errors by many folds. One of my old posts ‘Contextual Keypad – Engaging Users via Mobile Inputs’ addresses exactly this topic. Consider ‘Reset Password’ feature on Ctrip mobile app.
Reset Password behavior on Ctrip mobile app
If user taps on ‘Email’ input field, three things can happen:
  1. The keypad displayed is not email-friendly keypad. For user to search and locate ‘@’ symbol, he has to change the keypad where ‘@’ symbol is present (note that the keypad complexity increases further if many non-English languages are involved)
  2. User might end up typing his first name or user name or the email id before ‘@’ symbol (i.e., if the email id is Parimala@gmail.com, user might enter just ‘Parimala’)
  3. Since there is no real-time validation, user assumes that the given value is correct and taps on ‘Reset password’ button. Boom! An error, ‘Invalid Email: Please provide correct email address’ is displayed.
As you see, changing the input field type to display ‘email’ keypad could avoid user to perform 1st step mentioned above. It might be good to place place holder text 'user@example.com' inside the email field, so user becomes aware of the format. One might ask, ‘what kind of a user would enter invalid values despite knowing it’s an email field.’ It is a valid question. Yet, time and again, it has been proven that many users would do that. As Don Norman says, ‘An error that can be made will be made.’ Hence, we need to validate checkpoints in right fashion, to reduce such errors, if it’s in our control.
Input Masks
Consider a phone number field. Depending on your country, country code, changes. In such a case, phone number field would have 2 components: Country Code>Phone Number>. In below screenshot, can you guess if user needs to enter phone number with country code or not? It’s hard to tell. This could lead to errors.
If formats for input fields are such a big deal, why aren’t we taking them seriously? In ideal scenarios, it is advisable to insist users to enter input values in specific formats by implementing input masks.
Input mask is a mechanism by which one specifies the format in which the input will be accepted. For e.g. Phone number can have 2 input fields: Country code and phone number, as displayed below:
Whatsapp has clear demarcation of Country Code and Phone number
Similarly, date fields, credit card fields, social security number, bank account number and others can use input masks. 
Input masks not only help reduce errors, but also guide the user to enter input in the right format.
Input masks can be taken a step further by revealing the input pattern at the beginning itself. Consider the screenshot below. User would enter a 16 digit card number and keep checking constantly at your credit card and the field just to ensure you have not missed any numbers (Note that the credit card will have space in between every 4 digits). Instead if the below card number field could show 'XXXX-XXXX-XXXX-XXXX', it would be helpful to the user.
To summarize, reduce errors by:
  1. Using Inline Validation
  2. Specifying Input Types and
  3. Using Input Masks for formatting and accuracy
Have you thought about reducing errors on your mobile app forms lately?