Overview
The last 30 days at a glance
Daily active people
Unique people active in the app each day, last 30 days
iPhone vs Android
Share of active people, last 30 days
App health · 7 days
Most used sections
Tab visits, last 30 days
Top features
Most tapped, last 30 days
Latest activity
Newest events in the cache
Section Usage
App Opens Over Time
Activity Heatmap
Session Analysis
Per-User App Opens
Section Views by User
Recent Events
Tokens by Feature
Daily Token Usage
Feature Distribution
Feature Usage Over Time
Cost Projection
Feature Funnel
Prompt / Completion Ratio
AI Usage by Hour
Token Usage by User
Recent AI Calls
💰 Cost Intelligence
SKU: 37B1-4623-6F54
✅ Action Checklist
💰 Remote Config Pricing Tiers
| Daily Fetches | Rate | Status |
|---|---|---|
| 0 – 100,000 | Free | CURRENT |
| 100,001 – 10,000,000 | $0.06 / 10,000 fetches | Not reached |
| Above 10,000,000 | $0.01 / 10,000 fetches | Not reached |
🧮 Daily Fetch Cost Calculator
📅 Key Dates Timeline
Questions Over Time
Manual vs Preloaded Split
Word Count Distribution
Engagement Summary
Question Patterns
All Questions
Weekly Retention Cohorts
Churned Users
Events
🎯 Goals & Benchmarks
Set targets: DAU is today, the others are this month so far (follows the platform toggle)🧱 Stuck screens
Ranked by a stuck score: how often people leave without doing anything, weighted by how long they stayed first. A screen people leave instantly is a wrong turn. A screen they sit on for a minute and then leave untouched is one they could not work out.
🪜 Flow drop-off — the exact step
Every multi-step task, broken down by where it died. This is the table that turns "40% abandon onboarding" into "38% of abandons happen on page 14", which is a thing you can actually go and fix.
💥 Errors, by where the user actually was
Grouped by screen and message. The screen is the one the user was on when it fired, not the one the failing code lives in - those are often different, and the first is where the investigation should start.
😠 Rage taps
The user tapped one control four times in two seconds. They are telling you it looks broken, whether or not it is - a slow control and a dead control feel identical from the outside.
🚪 Dead ends
The last screen before the session ended. Some of these are healthy - people finish and leave. A screen that ends sessions but is not a natural finishing point is where people give up.
🐌 Slow operations
Median and worst-case duration. The median says whether it is usable; the p95 says what the unlucky fifth of users actually live with, and that is the number people abandon over.
🔍 Session replay
One session, step by step, in order. Pick a session to read exactly what happened - the fastest way to understand a single confusing report from a tester, and the only view here that shows causes rather than aggregates.
Language adoption
Which languages users actually switch to, and whether it was already cached or needed a download. A language with many downloads and few repeat switches is one people tried and left.
Undo — the trust signal
Every switch reversed within the 12-second window. A high rate here means people are landing somewhere they cannot read, which adoption counts alone will never show: a switch that gets undone still counts as a switch.
Preview → switch conversion
Whether looking at sample phrases changes the decision. Low conversion means the preview is doing its job (people are backing out of the wrong language); zero previews means nobody found the eye icon.
Bulk download runs
Started, finished, stopped, and how long it really took against the estimate shown to the user. The gap between the two is the bug.
Translation preferences
Address form, regional variant, native digits, Wi-Fi-only, and Ara's language override. Each of these was built on an assumption about what people want; this is where the assumption gets checked.
Storage actions
Deletes and download-without-switching. Both are paths that are easy to ship broken and hard to notice: a delete that does not free space, or a download button that silently switches the language instead.
Problems
Failed downloads, languages deferred for Wi-Fi, and reported wording. A language failing for many users is a catalog problem; failing for one is a network.
Daily Active Users
Section Popularity
Top Screens
Event Type Breakdown
AI Usage Comparison
Daily AI Tokens by Platform
Error Rate Comparison
Permission Grant Rate
💰 AI Cost by Feature — iOS vs Android
Feature Adoption by Platform
📱 Top Screens
Most visited screens by view count
👆 Top Feature Taps
Most tapped UI elements across all screens (iPhone only: Android does not record feature taps yet)
📈 Daily Interactions
screen_view · feature_tap · action_tap per day
🔢 Top Actions
Named action_tap events, most frequent first
Top Screens by Views
All screen_view events ranked by count
| # | Screen | Views | Unique Users | % of Total |
|---|---|---|---|---|
| Load data to see results. | ||||
Top Feature Taps
feature_tap events, what users are tapping most (iPhone only: Android does not record these yet)
| # | Screen | Feature | Detail | Taps | Unique Users |
|---|---|---|---|---|---|
| Load data to see results. | |||||
Data Entries Logged
data_entry events, what users are tracking (iPhone only: Android does not record these yet)
| # | Data Type | Screen | Sample Values | Count | Unique Users |
|---|---|---|---|---|---|
| Load data to see results. | |||||
Settings Changes
settings_change events — how users configure the app
| # | Setting | Value | Count | Unique Users |
|---|---|---|---|---|
| Load data to see results. | ||||
✨ Recently Added Features
New capabilities shipped across iOS (Prāana) and Android (Novara) during the current development sprint — May 2026
- 🗺️ Muscle Fatigue Map — 7-day rolling fatigue heatmap per muscle group
- 🏆 Post-Session Summary Sheet — auto-shown stats, PRs and share
- 📊 Volume Landmarks — MEV / MAV / MRV per muscle group
- 🎯 Readiness Banner — HRV + sleep + fatigue score
- 📅 Deload Suggestion — triggers after 4+ sessions/week
- 🔥 TDEE Card — Maintenance / Cut / Bulk calorie targets
- 🥗 Macro Split Card — Protein / Carbs / Fat breakdown
- 📏 Ideal Weight Card — Devine / Robinson / Hamwi / Miller formulas
- 📐 Body Fat Category Gauge — Visual zone indicator
- 🎉 Milestone Celebration — Overlay when goal weight reached
- 🍎 HealthKit Write-back — Weight & body fat synced to Apple Health
- 🩸 Blood Glucose Write-back — Saves to HealthKit from Blood Report entry
- 💓 Blood Pressure Write-back — Systolic + diastolic saved to HealthKit
- 🔧 HealthKit Auth Fix — Resolved NSException threading crash on launch
- 🎨 Theme Picker — Accent colour + Bright Mode toggle
- 🤖 Ara Personality Picker — Friendly / Coach / Clinical tone
- 🗑️ Multi-step Data Reset — Step-by-step progress flow
- 📓 Journal Settings — Prompts, biometric lock, export
- 🏋️ Goal Planner & Injury Log — Moved into Workout screen toolbar (matches iOS)
🚀 Feature Overview
Fan-menu feature opens — screen_view events for Workout, Body Comp, Food, Fitness, Blood, Meds, Health News, LunaFlow & Meal Plan
📊 Feature Opens (Total)
How many times each fan-menu feature was opened
📅 Daily Feature Opens
Opens per day for each feature over the selected range
Feature Detail
Per-feature opens, unique users, and last opened timestamp
| # | Feature | Opens | Unique Users | % of Feature Views | Last Opened |
|---|---|---|---|---|---|
| Load data to see results. | |||||
| Screen | Action | Count |
|---|
| Screen | Feature | Detail | Count |
|---|
| Time | User | Setting | Value |
|---|
📥 Export Filters
Data Preview
🔴 Errors Over Time
Daily app_error count
🗂 Error Type Breakdown
Distribution by errorType
⚡ Cold Launch Time (ms)
Daily average — target < 800 ms
🔑 Permission Outcomes
Granted vs denied per permission (health and notifications)
Recent Errors
Latest 100 app_error events, newest first
Performance Log
Latest 100 performance readings, newest first
🚀 Onboarding Funnel — Page Reach Rate
% of users who started that reached each page (returning users who sign in skip pages 13 to 19)
⏱ Average Time Per Step
Mean seconds spent on each onboarding page (stays over 30 min left out)
Step-by-Step Breakdown
Reach count, drop-off rate, and average dwell time per page
Recent Onboarding Events
Raw onboarding flow log, newest first (latest 100)
🐛 Reports by Category
Breakdown of report types
📅 Reports Over Time
Daily count in selected date range
All Feedback Reports
Newest first — click a row to expand the full message
📢 Announcements
Write a short message here and it appears as a one-time popup for every user the next time they open the app — no build or release required.
✏️ Publish a New Announcement
🟢 Currently Live
What every user sees on their next app open, until they dismiss it once.
📜 History
Every announcement ever published, newest first. Publishing a new one does not remove old rows here.
🔒 Date of Birth Correction Requests
Newest first · 🟠 Orange border = ID verification required before approving
📈 Dashboard DB Usage
Reads & writes made by this dashboard today (12:00 AM – 11:59 PM). Tracked locally in your browser — resets at midnight.
📊 Quota Usage
🕐 Reads by Hour (Today)
How many Firestore reads this dashboard made each hour
📅 Last 7 Days
Daily read totals from this dashboard
🔍 Read Operations Log (Today)
Every Firestore fetch this dashboard made today
Crashes per day
Counted on the day the crash happened, not the day it was sent
Crash groups
The same bug on many phones shows up once. Most common first.
Sleep & Calm
Sleep & Calm opens per day
Every open of the hub or a tool, last 30 days
Tools: opened and used
Opened = the tool was shown. Used = someone finished it (or saved it). A big gap means people look and leave.
Diary, course and nightmare rewrite
How far people get with the longer tools.
How people got there
Hub opened from Quick Add, or a link straight into a tool.
Screen tips
Each screen's tips
Skipped at = the average step people were on when they closed it. Skipping at step 1 usually means the tip came at a bad moment.
Contact us
What people pressed
Buttons on the Contact us screen. The three chips are the ready-made questions.
Why the helper failed
offline / network = the phone had no connection. http_429 = too many requests. http_5xx = the AI service had a problem.
Find a person
Paste a user ID, install ID or ticket reporter ID. Part of an ID works too.
People who hit problems recently
Crashes, errors and failed AI requests in the last 7 days.
💶 Money
Everything you pay for, in euros. Some lines fill themselves in; for the rest, type the amount from the bill once and it carries over each month until you change it.