Search code examples
databasearchitecturesocial-media

recommended ways to mine social media data and save it for analysis


I apologize if this question is a bit vague. But I am working with a friend of mine on a media company and we now have non-trivial customer base. The insights facebook and instagram provide are useful but not perfect and so I have started using the facebook API to save the data locally so that I can do some more analysis and combine it with other sources. I don't have much experience in this area and I am unsure whether there already some products that make it easier to amalgamate different kinds of social media data. In this vain I was looking for some suggestions in which software to use, what databases to use etc. At the moment I download csv files and then uplaod them to jupyter which isn't very efficient. We would like to better understand the groups of customers that exist and how we can better tailor to these specific groups.


Solution

  • You should probably start with your business question: What do you want to get out of it? I.e., When you say analysis, what actions are you expected as a result of your analysis?

    There are many (from freemium to really expensive enterprise-grade) tools to help with sentimental analysis. You can probably scrape / make API calls and do text analysis with many tools such as rapidminer, Python, and R, etc.

    For data visualization, Tableau is one of the more popular choices, and Google Data Studio is getting better and better by the day.

    If you have the budget, you can look at Simply Measured, Hootsuite Impact, etc. as well.

    Disclaimer: I have no association with any tools or products mentioned above

    Added based on OP's comment: If you are thinking of paid advertisement say on Facebook, you can take advantage of the look-a-like audience of the ad platform (i.e., Facebook's Help Page) Most advertising platforms let you place a conversion pixel on a conversion page (i.e., thank you page of an order form, etc.), and the platform will use its algorithm to find more of these types of people from its network.

    The downside, of course, is that the ad platform algorithm is a black box, and you have to rely on it for targeting.