Habit-Tracking App

Redesigning task
creation to feel
effortless

Role Product Designer
Focus UX Optimization
Platform iOS
Year 2026
Grit app screens
The problem

Users found creating a task to be complex.

Initial configuration is the core activation moment, and this is even more true for a habit tracking app like Grit.

Yet, user feedback analysis revealed that creating a habit felt complex and time-consuming for many users, even though the system technically allowed quick creation. Meaning that the issue wasn't actual complexity. It was perceived complexity.

68% of new users dropped before creating their first habit
40s+ average task creation time
Old Design

UX/UI Analysis

Users were confronted with a list of 13 settings options all before a single action was done. While core users appreciated the level of customization, new users felt overwhelmed by the number of options. The app's UI was satisfying and was not considered a priority.

  • Form anxiety Too many visible options created pressure before action.
  • Decision fatigue It's hard for the user to tell which action is useful and which isn't.
  • Lost in progression The lack of interaction makes it hard to say what changes have been made.
Key Insight

Users don't want to configure a tool.
They want the tool to adapt to their needs.

Design principles

Four decisions that shape the experience

01 Quick win

Limit the number of steps before achieving the task to increase satisfaction.

02 Focus on simplicity

Creating a simple task should be simple.

03 Reveal complexity progressively

Show more only when needed.

04 Adapt to user's behaviour

The system should anticipate user's next actions based on their actions.

The solution

Limit manual actions through a smart text field that recognizes and pre-configures key settings.

New flow:

Typing Smart recognition Manual customization

Constraints & design implications

  • Minimum redesign The current UI should be preserved. Changes are additive. No screen gets rebuilt from scratch.
  • Apple Intelligence alignment The recognition field must fit within Apple Intelligence's current system conventions: visual language, interaction patterns, and on-device processing expectations.
Step 1

Instant creation

Creation is reduced to a single text input. The previous flow required 7+ steps before a habit was saved. Drop-off data showed most users abandoned during configuration, not because they didn't want the feature, they just didn't want the work.

  • Users type the habit name and press enter.
  • They can save. The new habit is created.
Name your habit…
Step 2

Smart recognition

The system pre-fills common values based on the name. Pre-filling avoids the blank-form anxiety without removing control. Users override only when needed, which happens less often than expected.

  • An icon is automatically suggested based on the habit's name.
  • By default, the "Habit type" is set on "good" (green color). If the negative form ("Stop", "Do not"…) is used in the name, the habit transforms into a "bad" one (red).
  • Similarly, a habit is "daily" by default. If another frequency is mentioned in the name, it will automatically update the setting.
Take a hot bath every week
Good habit Green Bath icon Weekly
Do not smoke
Bad habit Red No cigarette icon Daily
Step 3

Make the experience lively and dynamic

The changes are immediately visible in the preview. A live preview closes the feedback loop between input and output. Without it, users have to mentally simulate the result, which increases abandonment. Seeing it in real time also confirms the system understood their intent.

  • Users see the impact of their settings right away.
  • In this way, they directly identify how the system works.
Live Preview
🛁
Take a hot bath Every week
Step 4

Interconnected options

Advanced options are displayed based on their relevance in the user's context. Showing everything at once overwhelms casual users; hiding everything frustrates power users. Context-sensitive revealing serves both without compromising either.

  • By default, the frequency is set to "Daily." If the user changes this setting, they are offered to set a start date.
  • If the user adds a start date, then we offer them to set an "End date", etc.
🔄 Repeat
Every week
📅 Starts on
Today2026
Step 5

Reveal complexity progressively

All advanced options are moved into a secondary screen. Early testing showed that removing options entirely triggered pushback from power users. Moving them behind "Show more options" preserves full control while keeping the entry point clean. Nothing removed, everything deferred.

  • Not used options are hidden by default in a "More options" section.
  • When enabled, they are automatically displayed so that the user can have a clear overview of their actions.
  • Nothing is removed. Everything is just delayed.
Editing
✏️ Take a hot bath every week
Type
Good
Icon
🛁
🔄 Repeat
Every week
📅 Starts on
1 Apr2026
📅 Ends
Show more options
New design

One field to rule them all

With a single action, users can set up their core settings in a smooth, progressive and natural way.
By minimizing friction, the new design allows users to focus on what matters most: improving their habits.

  • Single text input for instant habit creation
  • Smart recognition auto-fills type, icon and frequency
  • Live preview reflects every setting change in real time
  • Advanced options revealed progressively, nothing removed
Impact

Results

Tracked in Amplitude · 30 days post-launch vs. 30 days prior
+25% First completion rate

New users completing their first habit creation rose from 32% to 40%, driven by reduced friction in the redesigned task flow.

+27% Task Creation Speed

Median time to create a habit dropped from 49s to 36s, as the new flow eliminated redundant steps and surfaced key options earlier.

+9% FLTV

Projected increase in long-term value, modelled from improved retention and task completion rates post-launch.

Future enhancements

What I'd improve next

The redesign improved first task completion and habit retention, but the recognition system still relies on rule-based matching. Here's where I'd invest next.

30 min run Matched to Running 30mn · Daily Confidence
Improving smart recognition accuracy The current habit-matching system uses rule-based pattern recognition, which works for common inputs but struggles with phrasing variations. Training a lightweight model on real user inputs would reduce mismatches and shorten the time between typing a habit and having it correctly categorised.
30 min... 30 min run Running 30 min yoga Wellness 30 min walk Exercise
Autocomplete in the habit input Users who drop off at the blank input don't need a static list of suggestions. They need help completing what they're already typing. An autocomplete surfacing matches as the user types stays invisible until needed: it nudges undecided users without adding noise for those who already know what they want.
Final note

Complexity isn't necessarily a problem.
As long as it is appropriate for the context.

Next project

PepTalk

Turning a generic chatbot into a content-driven coaching experience