Inner Bloom
A smart journal that acts as a digital mirror
and emotional companion
OVERVIEW
Designing a sustainable journaling habit through emotional resonance and trust
Inner Bloom began as an exploration of how journalling apps can sustain long-term engagement. Most rely on prompts, streaks, and reminders to drive return. Those tools can boost activation, but they can also create pressure, especially in something as personal as reflection.
Inner Bloom explores a different approach. Instead of structuring what to write or rewarding daily consistency, it focuses on ownership, continuity, and trust. The idea is that people return to reflection not because they’re told to, but because the space feels meaningful enough to revisit.
I conceptualized, designed, and built innerbloom.garden (still evolving!)
METHODS
Core interaction model design
Rapid interactive prototyping
AI confidence gating
Visual design
PROBLEM
Journaling is easy to start and hard to sustain
It requires emotional effort, offers delayed reward, and is often framed as a daily obligation. Many apps address this with prompts, streaks, and reminders, which increases activation but also introduces pressure.
I wanted to explore a different question: How might a journaling product support return behavior without relying on guilt, streak loss, or constant prompting? The goal was not daily compliance. It was sustainable re-entry and emotional relevance over time.
User goal: Make reflection feel personally meaningful enough to return to. Users should feel seen.
SCOPE
Choosing the hardest version of the habit
I focused on freeform, blank-page journaling because it’s the highest-friction version of the habit. Unlike prompt journalling, it does not guide the user toward a predefined direction. It depends on someone already feeling the need to reflect.
Freeform also preserves ownership. When someone decides what to write, the entry feels owned. I believe that cultivating a sense of ownership would increase emotional relevance, and that relevance would sustain return more effectively than structured prompting alone.
DESIGN
Your inner garden
Each journal entry becomes a persistent plant in a shared garden, which can accumulate either spatially or chronologically depending on preference. This reframes journaling from “logging” to “observing.” By framing journaling as a living garden with social presence, the experience shifts from a task to a relationship, giving users a reason to return beyond discipline alone. Users return not just to write, but also to see.
Early versions included counts and summary overlays. These pulled users into evaluation mode. Removing them increased time spent exploring the space itself, a quieter form of re-engagement
AI THRESHOLD GATING
AI as a lever for meaning, connection, and trust
AI reflection was one of the riskiest parts of the system. If the reflection isn’t personalized or diverse enough, users might feel cheated that their entry data was being used without meaningful reward.Journaling is a sacred act that requires a lot of trust for continued investment. The goal is to let users feel seen, so a connection can be made.
This is where I had to balance control and automation, by setting conditions:
-
Users choose what the model can see (mood-only, mood + journal text, or none)
-
Certain types of reflections only appear when there’s sufficient signal (>95% confidence)
-
If there isn’t enough confidence for a thoughtful reflection, settle for a general journal prompt or a fun mantra
-
AI generated output is capped and cached to avoid repetition
In practice, this meant many sessions had no AI response at all in the beginning. However, it increased trust and encouraged writers to write more thoughtfully to unlock insights. Early users had reported that the AI summary felt supportive when it appeared.
DESIGNING FOR RE-ENTRY
Engagement without pressure
Most habit systems rely on loss aversion, something I wanted to avoid because I wanted users to feel s. Instead of focusing on streaks or reminders:
-
Inactivity is reflected visually through weeds appearing, that you can pull out
-
There are no resets or penalties for returning late
-
Streaks are understated, but if you hit a streak you get surprises in your garden
LEARNINGS
Ownership comes first, personalized prompts later
Bloom Journal reframed growth around trust and recognition rather than pressure.
In a small early cohort of 3 users, two reported returning after a lapse and described the re-entry mechanic (pulling weeds) as playful rather than punitive. Another reported deriving relevant insight from reviewing her own entries. While not statistically significant (yet), this aligned with the product’s core bet: that environmental feedback and ownership could support safe return and reflection for something as personal as a journal.
Learnings so far:
-
AI output should earn its presence, personalized prompts can come later
-
Retention improves when re-entry feels forgiving
View next project →
A social platform that connects city locals with newcomers