What is “self-service” analytics all about? Consider the analogy of the self-service gas pump. The driver gets out of her vehicle, grabs the fuel dispenser, hits a button and pumps her gas. Voila: self-service.
We should note from the beginning that “self-service” doesn’t mean doing everything for yourself. The driver didn’t load the station’s underground gas tanks. She didn’t process the credit card payment. She didn’t design the eco-friendly engine. But at the moment where it made the most sense for her to step in and take over the process for herself, she did so.
Likewise, when we talk about a self-service analytics analytics environment, we aren’t describing an environment where end-users create all the infrastructure or manage all the data. We’re talking about an environment where end users step in when the time is right and manage the process for themselves without a lot of hand-holding.
In a recent webinar, Claudia Imhoff, President and Founder of the Boulder BI Brain Trust and Intelligent Solutions, outlined 4 objectives for self-service BI / Analytics:
- Make BI results easy to consume and enhance – the solution should be easy to understand and not require a PhD to figure out.
- Make BI tool easy to use – the tools must be intuitive and obvious. You shouldn’t have to be a database admin to understand the data.
- Make the solutions fast to deploy and easy to manage – data must be readily available; infrastructure must be stable and easily updated
- Make it easy to access the data
Of the four, Claudia believes that last objective is the most important. “Make it easy to access the data… After all, if you can’t access the data, the other three are kind of moot. Who cares?”
I’ve been in the BI space for over 15 years. Throughout that time, easy data access has been an ongoing topic of discussion, and a primary goal of many of the solution providers within the space. (It’s worth noting that the goal has proved very difficult to achieve.) And now here we are in the era of big data. What does self-service mean to us today?
I think Claudia’s analysis holds up very well. Hadoop has done a great job of bringing together all the data, including so many different kinds of data from so many different sources. But delivering that data to end users in an accessible and easy-to-manage way has only become more complicated. Traditional BI vendors are making various attempts to retrofit their solutions in the era of Big Data, but these fixes have done little to make access any easier.
Many of these approaches simply add to the complexity by requiring one or more intermediate layers between the source data and the user. If the new layer is in place, you can go ahead and use your good ole’’ BI tool. Problems arise as you want more or new data; now you have to go back and tweak that intermediate layer. But this kind of “tweaking” is well outside of the scope of what most business end users of BI or analytics solutions can do. It’s kind of like asking our driver to fill the gas station’s fuel tanks or redesign her car. Now IT has to step in and help, and we’re no longer talking about self-service.
Platfora takes a different approach. Platfora decided early on that the big data era requires a new approach to data access: direct access to raw data via an intuitive, easy-to-use, browser-based analytics environment. In such a model, any business user can query the data without SQL or programming or other specialized Big Data skills. And there is no need for “tweaking” as new data is introduced, because there is no intermediate layer that needs to be updated.
A direct approach can help keep us on track when thinking about self-service in the age of big data. This is what sets a true Big Data Analytics solution apart.