I’m often asked how LaborVoices intends to compile Yelp-like reputations for employers. The short answer is that we’re gathering that information from current workers, themselves. The eventual goal is a search-able, multilingual database of worker opinions on employers, placement agents, and others. To get there, we’re taking several smaller steps.
The pilot phase of LaborVoices will be completely voice-based, in local languages (Hindi, Kannada, Gujarati, etc.) in a Q & A format. Potential migrants ask questions by leaving voicemail-like queries, and current workers at the destination answer those questions. This creates a kind of voicemail forum, where every time someone answers your thread, you get a missed-call that lets you know to check for new answers.
Later stages will begin by cleaning up and harmonizing the data, itself. This process will involve first tagging the queries and interchanges (Delhi vs. Bangalore, construction vs. transportation industries), then translation and transcription into an English-language backbone database.
Then the magic happens. The real coup will happen in the last stage, where worker opinions are rolled-up into reputation information for sectors and location (Delhi construction workers think X about their employers) and then broken down by smaller regions and individual employers. Eventually, we want workers to be able to safely identify their employers upfront; this part may need to be delayed until the number of workers involved is big enough to ensure anonymity.
Are there better ways to approach this? We’re all ears.