Evaboot
From podcast with Val: «With your QA phase, I can almost upload a new build to Chrome store blindly» - Jean-Baptiste Jézéquel, Co-Founder. A browser extension
Date
From podcast with Val: «With your QA phase, I can almost upload a new build to Chrome store blindly» - Jean-Baptiste Jézéquel, Co-Founder.
A browser extension for parsing LinkedIn
Mission
The purpose of this extension is to scrape LinkedIn. Specifically extract, clean, and find emails from LinkedIn Sales Navigator in one click. Its goal is to save time for sales managers and lead generators.
Challenges
The main challenge was a huge variety in LinkedIn page structures and differences that we gradually discovered, including changes in the layout that come with changing search filters.
Solution & Value
After the client has agreed to an initial code review which is vital for projects like that, we have fixed the critical bugs and made the extension work. Next, our team refactored the code, optimized the algorithm for new emerging types of LinkedIn pages, and made it independent of the set of filters on the page.
After a month, we measured the impact of our enhancements on key product metrics:
Feature adoption rate: 5.9% increase in weekly active extension users. Churn reduction: 7% decrease in drop-off rate (percentage of users who stop using the extension after a set period). Retention rate: 3.3% increase in retention (users with the extension vs. without), showing that those who install the extension stay engaged longer.
Handling API Disruptions & Browser Expansion
The extension heavily relies on LinkedIn's internal API, but when LinkedIn changed its API behavior, requests began returning null or empty responses, causing a surge in user-reported issues. After confirming the extension's core functionality was intact, we integrated Sentry for error monitoring to prioritize the most critical problems.
We discovered that Chrome was blocking API requests locally on the client side. Since browser behavior couldn’t be altered, we proposed a backend solution: the client’s server would fetch the LinkedIn data and pass it to the extension frontend. This successfully resolved the issue, restoring stability for users.
Initially, the extension was meant for Chrome, but later we extended it to a couple of other browsers requested by the client's userbase - Brave and Sidekick. While both browsers are Chromium-based, Sidekick introduced a challenge with its built-in ad blocker, which blocked the critical LinkedIn API request.
Thanks to our backend solution, not only was the problem fixed for Chrome, but the extension also began working seamlessly in Sidekick and Brave. This extension of compatibility met user demand and broadened the product’s reach.
🎥 Here is how the Co-Founder at Evaboot described his experience with us on podcast with Val:
