A complete restart
The version 1 of the app was built by an agency. By the time I joined, a 20-person customer success team was handling complaints the agency couldn't fix. Not because of slow turnaround, but because the architecture made it impossible. The app also carried none of the waterdrop® brand. It looked like a different product.
I joined in August 2021. No designs, no backlog, no foundation. The decision was already made: start from scratch. V2 launched in January 2022.
Left: what we inherited. Right: what we shipped five months later.
Research-driven decisions
Every major product decision went through the same filter: surveys, interviews, and feature importance exercises across 6 markets and 8,864 respondents. The customer success team gave us direct access to users from day one. It was the fastest feedback loop I've had on any product.
Research numbers at a glance
Total survey respondents
8,864 customers
V1 feedback sample
580 customers
V2 surveys
945 customers
LUCY surveys
562 customers
Qualitative Interviews
14 customers
Markets Covered
6 markets
Key insights
Faulty Streaks
The competitions tab was the second least favourite feature in V1, not because users didn't want gamification but because losing streaks was an engineering limitation that couldn't be fixed. Users rejected a broken implementation, not the concept. Trophies ranked joint second in the V2 feature prioritisation exercise with 22 votes
Drinking Tracker
The drinking tracker was the most loved feature in every survey it appeared in. 49.5% favourite in V1 (August 2021), 58.2% most loved in V2 (March 2022). Across 9,000+ respondents and two separate surveys, users kept telling us the same thing.
Reports requests
Reports were the single most requested feature in the March 2022 user research, ranking first in a prioritisation exercise with 14 users and 24 votes. Users wanted the app to show them their own behaviour.
Brand content requests
The newsfeed was being used as a brand content channel, not a hydration tool. 40.6% of users in the April 2022 dedicated survey opened it to learn about waterdrop products. 55.3% of their content requests were commerce-related: discounts, limited editions, and product launches.
In-app Shop
The in-app shop was the least favourite feature in V1, rated below every other feature in the September 2021 survey. Its removal for V2 had user data behind it before the rebuild even started.
Reworked app architecture for V2.
The app as a focused product
V2 wasn't just a rebuild. It was the first time we had the space to define what the app actually was.
The job was clear: help people drink more water. 88.3% of users in the V2 survey said the app did exactly that. 82% opened it more than three times a day.
Making it stick meant building features that reinforced the core job rather than pulled attention away from it. The drinking tracker was the most loved feature at 60.6%. Users wanted to track, to improve, to see their own progress over time. That shaped everything that followed.
Gamification
In V1, losing streaks was an engineering limitation, not a design problem. We rebuilt the system from scratch for V2. Trophies ranked joint second in user research with 22 votes. The V2 survey showed 17.3% named them favourite and 22.3% least favourite. Still a work in progress.
Reports & Statistics
61% of users who activate reminders drink when prompted. Around half of all users turned them on. The top reason the app failed the other 11.7% was not having reminders activated. The opt-in flow mattered as much as the feature itself.
Drinking reminders
61% of users who activated reminders drink when prompted. It is a simple mechanic, but it sits at the heart of the app's job. We built it to interrupt the right moment, not just to send notifications.
Visual identity
V1 bore no resemblance to Waterdrop. The colour palette, typography, and icon system matched nothing users knew from the brand. V2 was the first time the app looked like it belonged to the same company. Qualitative feedback from the April 2022 survey listed design among the things users actively liked.
Building the wrong feature
The CRM team wanted a news section in the app, a channel to push sales content directly to users. We argued against it. Leadership disagreed.
So we built it.
A dedicated survey in April 2022 showed us what was actually happening: 40.6% of users opened the newsfeed to learn about waterdrop products, not to track hydration. Their top requests were discounts, limited editions, and product launches. The CRM team couldn't sustain it either. One post a week was the reality.
The March 2022 V2 survey made the case clearly. 4.1% of 930 respondents named the newsfeed their favourite feature. 45.5% of 862 respondents named it their least favourite. We took the data to the CPO. It was removed on October 6, 2022.
Offline First
LUCY tracked hydration automatically via Bluetooth. In batch testing, most users assumed it worked straight out of the box with any bottle - it didn't. Calibration was required, and the connection wasn't always reliable. The reflective bottle interior meant tracking accuracy couldn't be guaranteed.
The engineering limitation couldn't be designed away. So the app had to work perfectly without LUCY. Every feature was built as a standalone first. LUCY layered on top, automatic logging when connected, manual entry when not. Users who paired earned double points.
The LUCY story in full
The hydration app had to work without LUCY. But when it did connect, it became something different. Designing the pairing experience, the onboarding, and the physical interface was a separate case study in itself.
Two apps, not one
In 2023, the Head of Ecommerce proposed turning the hydration app into a sales channel.
We pushed back. Adding payments, checkout flows, and a full product catalogue to an app maintained by two developers wasn't viable. And users loved the hydration app for what it was — embedding a shop would have cheapened it. No comparable hydration app had done this. We proposed a standalone product instead.
The shop app launched in 2024. It generates approximately €1M per month.
waterdrop® Hydration app
launched 2021
waterdrop® Shop App
launched 2024
Impact & Metrics
The rebuild produced a product users returned to daily, across seven markets. Retention, satisfaction, and recommendation scores all tracked upwards over the life of V2. App users showed 19.3% higher average order value than non-app users. 91.6% of LUCY customers paired their cap with the app.
iOS and Android
400k downloads
Average app store ratings
4.7 stars
Opening the app more than 3 times a day
82% customers
7-day retention
35.9% retention rate
App helps users drink more water
88.3% customers
Shop app monthly revenue
€1M
Reflections
Every time we considered expanding what the app did, the data pushed back. The newsfeed diluted it. The shop-in-app would have diluted it further. The features users asked for most — reports, trophies — were the ones that deepened the core job, not expanded it.
The app's strength was its clarity of purpose. That wasn't a brief we were given. It was something the research kept telling us, decision by decision.
©2026 Teodora Gheorghe