In every company, there is a great day when the real time tool provided by Google Analytics is discovered.
Amazed by this tool, everyone get around a screen and is able to see what users are doing at the company’s webpage at that exact moment. Ultimately people will become used to this, so there will be some companies that will display this tool with a great deal of pride in giant screens across the office.
When that much attention is focused in the real time tool, everyone assumes that now the office will behave as if it was a NASA command control room. This tool provides critical information for decision making and it helps to understand how users behave. Something like this:
Nonetheless, as the days go by, the screen ends up becoming a giant cinema screen display that everyone watches baffled, but no one really know what to do with it:
And in every office there is the guy that starts doing the math and calculates:
Is there any way to reduce this pattern? Can we make any consolidated conclusions with this information? The answer as always is: of course.
Key Elements: understanding information, what it is and for whom.
“The data in analytics will be segmented or it won’t be data at all”
Google Analytics has an API for real time reporting, from which the information can be extracted, put into different filtersand processed, providing specific information for different work teams.
Team number 1: system update and maintenance.
This graphic collects “real time” information every 15 minutes, even though it can be configured to collect information with up to 1 minute intervals. What happened in the dot that’s marked with the red arrow? The most probable answer is that there had been a system failure. Actually, if we check the time that’s displayed on the graphic, we can see that at that moment, there had been an increase in production changes that made the site inaccessible for a few minutes.
The ideal solution to this problem came from a developer who wrote a script that would automatically alert maintenance of a system failure if there was an x% amount of time happening between visits in the site.
The same way we can create a solution related to users activity, we can create solutions for other interactions in the site that we cannot loose track of.
This graph shows the amount of events happening at an e-commerce site when adding a product to the shopping cart. Is the site working? NO. Every drop marked with a red arroz shows system failures, a graph like this could help gather information in a matter of seconds in order to fix the problems and increase effectiveness.
Team number 2: Moderating
Social media, forums, newspages with open posts, any type of sites with direct interaction between users has to have a moderator that controls and assures that the site policies are being applied. That job can be really complex; the turmoil of information sometimes can create confusion about what becomes urgent and what doesn’t require full attention. We developed a solution for this problem, in this case for the site Taringa, the second largest social network in Latin America.
The mechanism is as follows: Using the real time API, we can detect which posts are being visited the most at that exact moment. Then, the program will crosscheck the posts ID’s with a listing of posts that have already been checked. Then the program will report only a top 20 of the posts that haven’t been monitored yet.
In the column “Estado”, we can mark when we checked that specific URL and it will automatically be added to the “Checked URL” list.
The idea of this simple examples and solutions is to highlight something really important, The analytic reports that we design have to be:
- Easy to check and access
- Easy to read and understand
- Provide concrete information that will help us make decisions.
We can seriously take advantage of the potential of real-time data collection if we can understand what we are looking for and what we want to see.