I want analytics for analytics

You’ve created a fancy dashboard or a cool report. It tracks the latest and greatest vital information straight from a real-time API source updated every three minutes. You see the data refresh and gladly share it with your organization/client. And then you wait. Crickets.

The analytics and dashboard solutions I have worked with don’t actually tell me if anyone is using them. They do tell me what is going on at whatever I’m monitoring, but services like Google Analytics or Klipfolio don’t show any statistics on how often they have been viewed.

To be honest, this seems sort of a strange problem to begin with, kind of a “the shoemaker’s children go barefoot”-type of a situation. Fingers crossed someone who can make analytics for analytics happen reads this.

PS. This is my first post published with AMP enabled. I hope this makes reading it through social media easier.

PPS. Turns out AMP doesn’t work like I thought it would. I’ll look into this later.

Disclamer: There might or might be a possibility to actually monitor report/dashboard usage in the services I mentioned, but none I could find at the time I was working with them the last time.

Cloud or code?

Cloud services aimed at businesses offer a cost-effective and a hassle-free way to create a digital service which would otherwise need a team of developers, a lot of money and expensive hosting. Most of them are mouse driven and require little to no technical knowledge after the initial setup. But when is using a cloud based service not the route you should take? Consider these questions first:

  • Do you need custom features?
  • Should the data be hosted at a certain location (country, continent) due to legal reasons?
  • Do you know exactly what your requirements are or should there be room for improvising new features?
  • How well does the cloud service scale feature and price wise?
  • Have you reviewed all of the use cases for the service and compared them to its features?
  • Are you prepared to accept an arbitrary limitation that might pop up and prevent something from working the way you thought it would?

I’m a big fan of using cloud services, but they’re not a magical one-size-fits-all solution. The better your own specification and planning for what you want out of a cloud service, the easier it is to figure out if it fits your business’ needs.

Centralized data

Never ever keep the same data in two separate places. Meaning, if you somehow have two or more different databases where you keep a table of your customers / offices / equipment, make absolutely sure there is automated synchronization between the two. I’ve worked with data sources where there are upto five different locations (cloud services, different SQL databases)  for the “same” data. In the end nobody really knows which is the most accurate source, nor which one is truly up to date. KISS works here the best: One source of data per dimension.