Would you rather access this page in English?

Prefere acessar essa página em português?

¿Preferirías acceder a esta página en español?

Blip Desk

Improving a mobile experience of a chat app focused on sales and customer service with chatbots and human interation.

Blip is an intelligent contact development platform. It allows the creation of chatbots for all types of businesses and integrates with various communication channels such as WhatsApp, Instagram, Messenger, Email, SMS, Blip’s native chat, and others. For cases requiring human interaction, the platform includes a human support tool called Blip Desk.

Today, Blip Desk has approximately 33,000 users providing support across diverse contexts. Message exchanges are organized via tickets with different statuses to help agents manage workflows. The mobile app currently has over 10,000 users.

challenges with the previous app

  • The app used webview technology displaying Blip Desk in a responsive format, leading to performance issues
  • Multiple accessibility problems
  • Interactions were not designed for mobile experience
  • The app did not follow the brand’s design system
  • Agents struggled to find important existing features
  • Long activation time
  • We wanted to improve satisfaction rates

accesibility study

We conducted an accessibility study with volunteer designers to analyze how the platform aligned with accessibility criteria, advancing this discussion within the company.

We selected Web Content Accessibility Guidelines (WCAG) criteria to experiment with, adapted Blip Desk’s interface, and tested with two groups: one with low-vision users and another without.

development of a new app

Due to the app’s issues, we decided to rebuild it using different technology. With the freedom to redesign the experience from scratch, we created a minimum viable product (MVP) of the new Blip Desk for a beta program.

We prioritized features by analyzing usage data and user behavior from app tracking and incorporated feedback from NPS and CSAT surveys. Interviews with users who provided feedback and companies interested in entering a beta program helped us understand expectations.

key activities

Feature prioritization
New journey mapping

Designing new component
Interface prototyping
Interaction documentation

Usability testing
Product Usability Scale evaluation

Development handoff and oversight
Beta program planning and recruitment

mobile design system and new components

I led the adaptation of Blip’s design system for mobile products. This involved revising interactions, proposing mobile-specific components, and applying accessibility learnings to improve colors, typography, and interactions. These updates were later extended to all company products, including those for desktop.

new beta app

To test the app in real-world scenarios, we launched a beta program with select companies. The features prioritized for the beta program were part of an MVP aligned with company goals and product metrics.

feature showcase before login

We highlighted key functionalities Blip wanted to promote directly on the login screen via a carousel with brief explanations. This improved discovery of underused features.

onboarding

We accelerated product activation with an onboarding flow explaining the app’s core sections, helping users learn faster and encouraging them to test new features.

chat screen

Legibility and interactions of message components were improved. Microinteractions like writing messages, selecting quick replies, and audio recording were refined to feel natural and use familiar interaction patterns.

file and media upload

We streamlined image, video, and document uploads with a preview carousel, allowing users to easily edit or add files.

beta program results

  • 18 companies actively using the beta app
  • Product satisfaction rose from 64% to 82%, according to survey data
  • Company activation time dropped from 6.7 hours to 1.6 hours
  • 30-day retention increased from 8,5% to 26%
  • Percentage of users performing support tasks grew from 45% to 94%

Software used: