Check out the Github Repo
Google's Great, But...
A majority of their 'Top Hit' websites are bought - not earned. And despite their advances in Machine Learning and Artificial Intelligence, Google's recommendations are bland and don't predict me - or my friends - well at all.
For the purpose of this piece, I just 'Googled' places to eat in Princeton, NJ, my home-town. 5 out of the top 6 suggestions are meant for people ages 60+ and with exquisite, refined tastes; definitely not what I'm going for and certainly not what my search histories have indicated. Google's 6th recommendation was closed - and had been closed - for the past few months.
Regardless, none of those top recommendations are 'popular' in Princeton. From my experience as a local, there are no lines or long-waits on Thursday/Friday night, or large family brunches on Saturday/Sunday morning, or people running around ranting and Instagramming their favorite meal there.
As someone who travels often, that's worrying; how much have I missed out on by blindly trusting one corporation's recommendations? Haven't we all heard the old adage "never put all your eggs in one basket"?
If only I had thought of using social media (public FB/Insta posts) to find out the really popular places the locals' loved, and visited there.
But wait, social media isn't the only indicator of the popularity of a place. The number of credit card transactions or the number of ATMs in a region (Capital One's API) can also hint toward highly-trafficed places, and combine that with social media and articles tailored for locals and neighbors (Comcast's EveryBlock), and we have the data to make our own-informed decisions.
That's what I set about doing this weekend: combining big-data APIs and services to create a more fool-proof and reliable - crowd-sourced - way of finding out information about everything: from the best places to eat, to the new stores on the block, to even the community service options available to me.