BlaBlaCar implements pipelines in half a day and saves over 64 hours on data ingestion with Rivery.

Background

BlaBlaCar is the world’s leading long-distance carpooling platform – a global, trusted community of over 100 million drivers and passengers in 22 countries. The platform connects people looking to travel long distances with drivers heading the same way, so they can travel together and share the cost.

Rivery is like a magic box. Once you open it everyone else wants to use it. There are so many use cases. 
Antoine Lefebvre, Data Engineering Product Owner, at BlaBlaCar.

Challenge

BlaBlaCar was seeking to ramp up their marketing data ingestion process. Their main goal was to ingest Ads Campaigns Reporting data from Google, Facebook and Bing ads – at scale. With limited access to data sources and the right tools, their developers spent hours on repetitive work – manually coding with Python for data schema changes and ongoing maintenance.

A lot of the products we tried out were limited so we quickly discarded them due to the limited number of connectors. They were also too expensive or with inappropriate pricing models

Solution

BlaBlaCar searched for other API ingestion tools when Rivery came up in their Google search results. They tested out Rivery along with another solution.  They quickly realized the benefits of having access to more marketing connectors, fields, REST API and pricing were too good to ignore. 

Google APIs change fields so we absolutely wanted to contain and auto-manage these schema changes. We also needed REST API and BigQuery support. Rivery’s ease of setup, wide range of marketing connectors, field-mapping options, and instant ability to detect was a turning point for us.

Results

Before Rivery, BlaBlaCar was stuck for days on end trying to connect APIs, and sometimes weeks trying to ingest data for analytics and reporting. It now takes them half a day to connect APIs and get their Tableau data analysis-ready. 

Migrating Facebook ads only took us a week to implement. Before that, we had to update the APIs every three months and our data analysts would be blocked for days. For one use case with bus GPS data we used two different APIs – and in half a day our pipelines were up and running.

Rivery is now BlaBlaCar’s default starting point for anything data ingestion-related. They’re now looking to try out reverse ETL and Webhooks and other cases.