Google Analytics is not only an amazing Tracking & Analytics platform. It could be used also as powerful database that allows you to store and retrieve amazing data, that since yesterday would be too expensive to manage. Let’s see how, looking at my direct experience.
Intro:
As CTO I’m the responsible of the whole tech infrastructure of Punctis. Punctis is a loyalty platform through which websites are able to reward (with virtual points) visitors who like their pages, clicking the well-knows “Like” button. So, websites which wants to integrate Punctis on their pages, has to substitute the standard Facebook Like button with our button (which is exactly the same in terms of visual, exactly like AddThis, ShareThis, Gigya, do).
In practice, Punctis is placed (in hidden way) between the user who wants to like a content, and Facebook. This is a huge opportunity for Punctis: being able to track in real time what users like around the web (on pages which use the Punctis Like button). Today we have partnership with some of the biggest Italian publisher, and our buttons are clicked hundreds and hundreds time a-month! This means a huge amount of valuable data, and I do not want to let them go away!
Need:
I want to be able to store these informations: which pages have been liked? When? How many time? Who liked each url (browser, os, etc)? And then I want to be able to explore this data, subsetting it. Example: which is the partner who is generating more like on a monthly basis? What is the browser mainly used by the user who share content about Sport? What is the percentage of like made by mobile device instead of by desktop? What is the trend about “like” on mobile device? And so on. Collecting this data adds value to Punctis (AddThis and ShareThis became a big data company exactly for this reason: they track a huge amount of like and sell these informations to brand and research institutes).
Problem:
Each “like” is a new row in the table database, containing URL, data, browser, os, etc. This is stressful for the server, and doing queries on this table will soon become impossible due to the too large amount of data. I would have need to invest in more complex database, hire an expert of this kind of databases, etc etc. Possible, but not the best idea for a startup, not at this stage.
Solution:
We placed an invisible Google Analytics tracking code inside the Punctis like button: each time the button is clicked a javascript script triggers an Event on Google Analytics, saving the liked URL, the id of the partner, and other informations.
Wassup? Using the Google Analytics user interface I’m now able to know what I needed, and even more!
This for example is the list of most liked URL between 20 October and 19 November, for one of our partner:
And I can use all the data that Google Analytics collects, like browser, operating system, etc to go deeper into this data. For example here I want to see which are the most used operating system by people who liked a specific web content (a blog post), in a definite date range:
Wanna know where the people who like a specific page come from? Simple:
And so on!
This huge amount of data is now reachable through a great user interface (thanks Google!), without any resource spending by Punctis because the servers who stores all these data are Google’s servers (thanks Google!).
More:
One of the value propositions of Punctis is to give to partners insights about the audience who like its contents. So I have to find the way to show all these data under the punctis.com domain, in a custom dashboard (I cannot give access to my Google Analytics Account to each Partner, it would be not good). What to do? Simple, we implemented Google Analytics API to retrieve the data and the collections of data we need, and show them to the partner when he request them through its dashboard, and this is the result (without any query on our database):
So, let’s start thinking to Google Analytics not only as a Tracking&Analytics tool: it would help you to save resources and have better results!
