Google AMA Video
A new search enhancement that invites noted individuals to record video answers to the most-asked questions about themselves, their projects, and topics they are considered experts in.
Sample
Videos live in a carousel on the search-results pages of the person who answered and the topic the video is tagged for. The video will also pop up in a one box if the exact question is searched and audio of the video is played when the question is asked on Google Home. Go ahead. Either search for or ask your Home, “How did MC Hammer get his name?”
Project
Launch product app and CMS. Work with engineers to set product up for scale and automation. Also, leverage personal and professional contacts to help partnerships team recruit users.
Role: Content Strategist/Specialist
Process
Set an editorial voice that was at once conversational and neutral.
Work with policy to create guidelines for both questions and answers. Also provide a deep dive into the topics and types of questions that should be block-listed.
Identify top 1000 occupations and create database of questions for each. Having a knowledge panel is the only criteria required to participate in this product. And, anybody with a knowledge panel is tagged with at least one occupation on the backend. By preloading the bank with questions according to occupation, we were guaranteeing that anybody who signed up for the product would automatically get a set of semi-relevant questions—even if we didn’t have a robust question data set for that person.
Offer white-glove treatment to early adaptors—conducting research on each person and writing personalized questions.
Create bank of fill-in-the-blank questions that the machine could easily customize with little error.
Work with engineers to build CMS for product.
Strategize product scalability.
Lead partnership team’s chef recruitment and work with the Google Talks team to coordinate pitching to speakers.
Identify which Googled questions were best answerable by humans in this medium or best left for another type of search result. We found that the most-asked topic questions roughly fell into two buckets: how-to and eCommerce. We also knew that these videos worked best under a minute in length. By thinking about user intent, we could speculate that the person Googling “How to make a chocolate lava cake?” would be more satisfied with a YouTube video or recipe. In a similar line of thinking, in terms of answer quality, we wondered if the owner of a cosmetics line should answer “What is the best mascara?” Perhaps it was best to eliminate questions that could lead to product placements. This was a continuing process with rules that were constantly updated.
As smart as the machine is, human eyeballs are needed to audit. We trained a team of vendors to do just this based on our guidelines and learnings.
