With the recent launch of the Quora Public Data Program we are excited to share an introduction into the type of content and interaction you’ll find from its 300+ million monthly visitors. In this post, we’ll take a look at ways of approaching your analysis of public Quora data and highlight what makes it a unique source of insight.
Quora is the leading knowledge sharing platform with questions ranging from the philosophical to practical. The focus is on shared information and experiences. There are a few Quora specific elements that are helpful to understand:
- Questions are tagged with relevant topic labels. This helps to organize conversations together and makes discovery of more relevant content easier.
- Answers are upvoted by the community pushing high quality responses to the top.
- Users are able to build public profiles complete with their areas of interest and expertise.
The recent Preparing For the Normal report on Quora highlighted several topic areas of growth during this unprecedented time. As fan of bread and baking, I was interested to see the jump in interest in Sourdough. Such a great example of the breadth of topics on Quora!
If I want to analyze this increase in sourdough related content further there are a few ways to compile a relevant dataset:
- Keyword Specific: look for mentions of “sourdough” across all questions, answers, and topics
- Utilize topics to find additional relevant questions and answers. In this example, beyond the Sourdough Topic I might want to expand and also look at:
- Identify experts in the area of Sourdough using the “knows about” section of author profiles
- Search for brands related to Sourdough: Whole Foods, King Arthur Flour
Once I’ve identified my dataset, the Quora specific metadata helps round out the picture. Examples include:
| Analysis Question | Useful Metadata |
| Which questions are getting the most traction? | number of views, follows, answers; can use these to create a customized ranking of questions |
| Which answers are most resonating with the Quora community? | number of views, shares, comments, upvotes; can use these to create a customized ranking of questions |
| Who are the authors influencing this conversation? | author profile info including number of answers by topic, upvoted answers, content views; for example, can identify who are the Sourdough influencers in the eyes of the community |
| What is the context of the sourdough questions? | topic tags; for example, seeing an increase in the sourdough topic tag alongside the working from home tag |
| What brands are being mentioned? | topic tags; for example, comparisons of flour brands |
| How has the conversation and interaction level changed over time? | Initial posting date as well as any updates to questions and answers; for example, questions have long life spans and increase/decrease in activity as the larger conversation shifts |
These same approaches for identifying and analyzing a dataset are applicable across the wide variety of Quora content. Please reach out if you’d like to see the specifics for your industry or brand!
