How to Smell a Rat with So-Called “Self-Service Analytics”

April 18, 2019
by: Matthew Halliday
Blog

When I assumed the role of chief evangelist for Incorta, I imagined I’d spend my days helping really smart people apply our analytics platform in new and creative ways to address their business challenges, and pontificating about what the next wave of modern analytics will look like. I do these things, of course, but I also spend a great deal of my time educating these same really smart people about what self-service analytics is—and what it is not.

Thanks to a myriad of analytics vendors blowing the “self-service” horn—most of whom only offer visualization or dashboarding capabilities that still require a great deal of support from technical resources—people are confused about what “self-service analytics” really means. And I don’t blame them.

But it’s time to clear up the confusion. It’s time to understand what true self-service analytics looks like, so you can spot it when you see it—or smell a rat when you’re pitched technology that doesn’t measure up.

Here it goes.

An analytics tool is NOT really “self-service” if:

  • Users don’t trust the insights it delivers.
  • Users can’t access data in near real-time.
  • Users can’t create and format reports on their own without technical help.
  • Users can’t develop new insights on the fly without technical help.
  • Users can’t get answers to subsequent questions without technical help.
  • Users can’t immediately access the transactional details that comprise topline or aggregated insights.
  • Users have to wait weeks or months for help from IT. 
  • Users have to attend long or ongoing training classes to understand the tool’s basic use.
  • It only does visualization and/or dashboarding.
  • It delivers incomplete, inaccurate or only aggregated insights.
  • Its query speeds are slow.
  • It loses data fidelity during analysis.
  • It returns different results depending on who runs the data and how they do it.
  • It doesn’t inherit full, existing security settings from source systems.
  • It needs legacy techniques like star schemas, data modeling or Extract Transform Load (ETL) in order to work.
  • Your technical team has to try to predict all of the questions users will need answers to before they can fully deploy the tool—an impossible task.

Some of these statements might surprise you. But self-service analytics isn’t just the capability of “Can I use a tool to get something?”

True self-service analytics is “Do I have data that’s meaningful? Do people engage with data that’s meaningful?” After all, people can make an informed decision that’s wrong if they based it on inaccurate or incomplete data. And the ability for a user to build their own dashboard isn’t that useful if the dashboard presents info that’s an abridged version of the truth—or downright inaccurate.

That’s why we at Incorta want to help our customers make “smart” decisions, not just “informed” decisions. And that’s why we developed a revolutionary new platform that delivers true self-service analytics—one that gives you all of the self-service criteria in the list above that the others do not.

We’d love to tell you more about it. Learn more—schedule a time to speak with us or read the new “Self-Service Analytics: Fact or Fiction?”