Recommend – A Social Network for your Recommendations
Startups Helping Startups
As a startup, one of the best things we get to do is work with fellow startups to help them grow and expand their businesses. Over the past several months we have had the great pleasure to work closely with the team from Recommend, a French startup that is changing the way people get, use and share recommendations.
A New Way to Recommend
Recommend have created a new type of social network powered by trust and based on word of mouth, that allows users to share recommendations with their friends and get recommendations from people they trust. You bypass anonymous reviews and social media noise by following people, publications or blogs you trust to give you just the right advice. You can browse the latest recommendations on your homepage (a bit like the Facebook Newsfeed) or search for exactly what you need. Using Recommend you can find everything from an amazing restaurant to museum exhibitions, babysitters or even a good-value plumber!
It All Started with Some Data
To get started, the data team at Recommend needed to collect high-quality content from selected publications and blogs. The aim was to seed their network with great recommendations from trusted sources and allow users to curate lists of cool “recos” to share with their friends. As most of the sites they were interested in did not have APIs, the team at Recommend came to import·io as a solution to their data acquisition problem. The nature of the data they needed presented two interesting challenges:
Firstly, the data Recommend needed to build out their extensive network was spread across dozens of sources. Although they only wanted a thin layer of data from each site, in order to get it they had to build an API to each site individually. Quality, not quantity was their repeated mantra. Using import·io they were able to reduce the time it took to create each API – to only a few minutes. Very quickly the data team was able to create APIs to dozens of sources including Timeout, Telemara, and LeFooding.com; bringing in 1,000s of high-quality recommendations that users can now view, save and recommend to their friends!
The second challenge Recommend faced was that they needed to get at data which changes frequently. New recommendations are added all the time, and in order to give people accurate information, the recommendations had to be pulled live from the sites. All the data Recommend collected via import·io was available over the API in real-time allowing them to pull fresh data into their app on a daily basis.
Give it a Try
Recommend has just launched worldwide in English and French as both a web and mobile app. The team at Recommend have got big plans to expand to other languages and cultures and we continue to work closely with them to add more recommendation sources.
