Fundamental best practices for BI/analytics: Self-Service

My last two posts have focused on two best practices for business intelligence and analytics strategy.  Last week, we looked at how formalizing the approach taken to analytics can help your scarce IT skills be most productive.  This post looks at a very complementary strategy, enabling end-users to be more self sufficient in their use of analytics tools.  Applied correctly, self-service analytics can both help users find timely information more frequently, and free up the time of skilled IT professionals for more strategic work.

One of the reasons that users often become disenchanted with a BI solution is simply because they can't find the information they need, when they need it.  When that happens, one of two things happens next.  Either the manager goes to an IT professional for help, or they just go ahead and make a decision without any relevant data.  And unfortunately in practice, seeking the help of an analytics specialist isn't always fruitful either.  Often, their technical skills are in such demand that they aren't always able to help the user in time anyway.  So much better then if a user is able to find the data they need, unaided.  That can best be achieved by a combination of things:

• The right tool:  Static presentations of information can only take you so far.  They're great as long as the information needs of the user are perfectly anticipated, and those needs never change.  In other words, static displays of information never really provide everything that the user needs.  On the other hand, an interactive view of data that allows a manager to drill down to detail, and explore data can be much more powerful when unforeseen needs arise.  Interacting with data can be a great way for users to start to explore cause-and-effect and really start to drive necessary change in business processes and workflow.


• The right training:  for many users, a highly interactive analytics tool might be completely foreign to them.  For that reason, training is a must-have for many uses.  And, that's not just training on the features, functions, and quirks of a particular tool.  It might also need, for example, to encompass training on basic analytics practices, good decision-making skills, and education on the uses and limitations of data.


• The right support:  One of the benefits of self-service analytics is that it can reduce the burden of mundane needs on skilled IT staff.  Unfortunately though, that doesn't mean the IT staff are off the hook completely for support.  Certainly in the early stages, a lot of handholding may be required.  And specifically, it may be more valuable to have those support resources embedded within a particular business unit during the transition.  It's important to remember that so many people have had bad experiences with new IT systems that they have little patience for new roll-outs.  Much better then to make sure they are fully and completely supported during the initial part of the roll-out in order to make the project successful overall.


• The right culture:  In many cases, all of the above is also critically dependent on the organizational culture.  That is, more junior managers may need to feel truly empowered to make decisions which were previously above their pay grade.  And that of course, means that more senior managers have to be willing to delegate those decisions too.  Without this cultural shift, all the technology, all the training, and all the support in the world will not make the project a success.

Done well, introducing a more self-service style of analytics can be a win-win.  Users are more likely to find the information they need when they need it, independently without help from skilled IT staff.  And, those skilled IT staff in turn are able to focus on more strategic things - like spreading analytics and best practices throughout the company.