Less Data Noise,
More User Retention

HydroPoint is an award-winning leader in smart water management solutions, dedicated to developing innovative and environmentally sustainable technologies.

Description

Key problem

The insight

Additional improvements

Design system

Client

Hydropoint is US SaaS platform

Industry

SaaS

Team

Product specialist, data engeeneers

Users are Operations Managers at large retail and commercial chains — like Lowe's with 939 locations. Their job is to manage water costs across a huge portfolio without visiting each site.

My role: worked directly with the Director of UX, high level of autonomy, 5-month contract.

Big problem = money: users did not return after the first session, losing their subscription.

No design system → every screen from scratch, developers are slowing down.

Client problems I was facing

I. Registration Stage
and Its Variations

Raising user retention
Updating design system
Refining the UI

Users weren't coming back

Big problem = money: users did not return after the first session, losing their subscription.
A stakeholder wanted to cram a large amount of data onto a single screen.

I pushed back with my own design solution, and to validate it, I went and talked to the team that interfaces directly with users. My hypothesis was confirmed — the overloaded screen would indeed have made it hard for users to navigate, which backed up my decision.

Chat Interface Instead
of a Form: An Alternative Approach to Registration

A stakeholder proposed the idea of testing a chat‑based interface instead of a step‑by‑step questionnaire. The hypothesis was that a chat experience would feel more familiar to users and reduce cognitive load during data entry.

User need this section to monitor multiple water consumption metrics across various locations — including availability, cost, and usage. All data types were displayed together, making it hard to focus on specific insights or compare between regions.

The insight: think like Google Maps

I introduced a layer-based map interface, where users can toggle between distinct data types, such as Water Bills, Site Surveys, Climate Data, Asset Mapping.

The main decision is to use layers instead of everything at once. Analogy: like in Google Maps where you switch between traffic, satellite, transport. Here: Water Bills / Site Surveys / Climate / Asset Mapping.

Additional impovements to get it right

ITo help users read the data more easily, I also refined several other details. For example, I addressed how the number of objects clustered at a single point was displayed, improved the infographics, and introduced color-coding for key elements. I also proposed displaying the important data for each layer directly on the map, within a collapsible panel — so users could access relevant information without cluttering the main view.

Complex data - finally readable

No design system → every screen from scratch, developers are slowing down.    

One of the most important tasks in refining the UI and before adding new layouts was to streamline and improve the design system. There were a huge number of unused colors in the color palette, it was important to create a basic palette, auxiliary colors, and accent colors.

After improvements

The number of returning users grew significantly.

III. First Portfolio
+ Get‑started
landing pages

The interface became more convenient, development was streamlined thanks to the new design system, and the overall UI quality improved.

Less Data Noise,
More User Retention

Client problems I was facing