Driving 25% engagement through predictive insights

Driving 25% engagement through predictive insights

Designing predictive data visualisation to solve EV charging uncertainty and reduce user anxiety This project is under NDA hence, most of the user research is not included in the case study.

Designing predictive data visualisation to solve EV charging uncertainty and reduce user anxiety This project is under NDA hence, most of the user research is not included in the case study.

My role - User interviews, product strategy, UX & UI design, interaction design

Feature

Live

Live

Team

Product Manager

2 Developers

Data Analyst

QA

Background & project goals

The app was experiencing a lot of drop-offs by EV drivers due to inactivity.

The ChargeNow app is used by EV drivers to find chargers, charge their car and pay through the app. The product team wanted to increase user activity on the app, by adding a new feature that would separate us from competitors.

KPIs

We came up with a hypothesis that with the implementation of this feature, the user activity would increase by 25%

The feature itself might not directly generate revenue, but creating key performance indicators, the performance of the feature will be measured in these aspects to understand the patterns of user activity and the impact on revenue generation.

Track active users of the forecast feature vs. total app users

Track active users of the forecast feature vs. total app users

Measure charging sessions of the charger initiated via the forecast

Measure charging sessions of the charger initiated via the forecast

Analyse usage patterns of day or time to understand user habits

Analyse usage patterns of day or time to understand user habits

Problem

Existing research showed that drivers had a problem trying to find an available charger especially during peak hours

The customer experience team had gathered feedback that the EV drivers often faced uncertainities finding an unoccupied charger, especially during peak hours or when in unfamiliar locations.

Solution and features

I designed a feature from 0 to 1, that would help EV drivers plan their charging trips efficiently to skip wait times

Popular times is an interactive forecast feature that displays real-time availability and weekly forecasts of a charger.

Popular times section

  • Interactive graph with 24 clickable bars, each representing an hour

  • Real-time animated bar

  • Collapsible section that reduces cognitive load

Time and day selection

  • Help texts linked to status bar of the charger

  • Help text provides explanation on the bar height

  • Tabs show availability forecast of a selected day

Info pop

  • Provides transparency to users about data gathered.

  • Gives an overview of how the data can be interpreted

Research insights

I spoke with EV drivers to understand the major pain points

I interviewed 5 users who ranged from a rookie to an experienced EV driver to understand their problems and expectations.

Distance

Driving to a preferred charger is far and can result in energy loss

Plug compatibility

Driver might not get a plug that fits with the vehicle's charging port

Wait times

Most chargers in the city are occupied and hence wait times can be longer

Information accessibility

There is uncertainty in the info for the user to know when a charger will free up

Competitor analysis

Looking into our direct competitors, I found that none of them showed availability of a charger in real-time.

I benchmarked 6 competitors, to find commonalities and uniqueness. Most of them showed availability forecast of an entire charging station with multiple, which is not helpful for a driver who is looking for a specific charger.

Collaborating with the product team

Before jumping to design, I defined user stories with the PM to help myself understand context, needs and desired outcomes

As an EV driver...


  1. I need to know when is the best time to charge my vehicle during peak hours so that I can avoid waiting.


  1. I am planning a long drive and I want to know if I can find an available charger, without worrying about running out of battery.

I collaborated with the PM & developers to understand tech constraints, APIs and possibilities

The solution was designed based on these possibilities


  1. API available to fetch live status of each charger at a charging station.

  2. Forecast data is based on last two weeks of charger availability.

  3. The charger availability forecast data is related to the plug type selected

Design process

Based on competitor analysis, I wanted to create a feature with better accessibility standards

I focused on these principles that helped me think through my designs-

Accessibility

Visualise data that is understandable for interpreting

Visualise data that is understandable for interpreting

Visual hierarchy

Focus on most important info which users want to see

Focus on most important info which users want to see

Relatability

Reduce ambiguity and guide user to interpret correctly

Reduce ambiguity and guide user to interpret correctly

A. Figuring out the placement

After an A/B test with users, I found that version B was preferred over A due to the info relatability and visibility of the section.

Forecast information changes based on the plug type selected. Users in the interview mentioned that pricing is also important while choosing a charger.

Version A

  • Feature will appear hidden from the viewport.

  • Users cannot relate the feature to the plug selected on the top.

Version B

  • Section is closer to plug selection; hence it is relatable.

  • Pricing display is not hidden from view port.

Collapsible section

  • Section is collapsible to reduce scrolling and cognitive load.

  • Provides more visibility for pricing display in the viewport.

B. Designing the section

The section will be collapsible with labels to accommodate the pricing information below it

I designed an overview for showing the status of the charger in the collapsible view, as it's easier for user to scan before opening the section to analyse the data. Iterations are under NDA and hence only mock-ups that did not make it to the project are shown here.

This is an idea that I thought of to show colour coded labels that reflect the live status of the charger.

Expanded view with status label. This idea did not make it through as you can see why, below.

The content had to be translated into 8 European languages. The characters count increased and the design did not prove to be scalable because of spacing and readability issues.

This is a coded PoC that shows the character count increasing when translated.

This is the final design "Version 1" that went live. During the implementation phase, there were tech constraints that prevented us from implementing advanced features (it will be implemented when version 2 will go live.)

C. Building design components

I used nested components method in Figma to build the feature from scratch. It resembles a cell, where small components (atoms) are put together to form larger components (molecules and organisms).

D. Designing edge cases

Talking to developers during design sprints was super helpful to identify edge cases.

Sometimes, there are user scenarios where the data might not be available at a charging station or it takes a lot of loading time to display.

Loading state

Data not available

Key learnings

The feature evolved based on users' feedback on readability, comprehensiveness, and usage patterns.

While many ideas have been gathered through user feedback, it can be enhanced further. Initially, it will be observed in its Version 1 stage to gather user activity that will help in the implementation of Version 2.

Break- even phase of the business had an impact on the feature's implementation.

During the company’s transition, limited developer capacity led to advance features being put on hold to prioritise other tasks. This taught me that compromises are sometimes necessary to meet business needs while ensuring the solution remains as usable as possible.

Are we a match?

Let's work together!

© 2025 Apoorva Narayan

Proudly built on Framer.

Are we a match?

Let's work together!

© 2025 Apoorva Narayan

Proudly built on Framer.

Are we a match?

Let's work together!

© 2025 Apoorva Narayan

Proudly built on Framer.