Self-service BI is a myth

Face it.  It’s not a thing.  Or, at least the way the concept is spun, it’s not.  I’ve sold BI technology for six or seven vendors (depending on how you define BI) over the last thirty years.  I’ve sold for the big stack vendors, the sexy new start up types, the niche players, the data warehouse guys, and a third tier wanna’ be.   I’ve focused on the back end, the front end, and the full solution.  I’ve sold the software and the consulting services to implement it…. I’ve been around.

Today there are a lot of folks who seem to think that self-service BI is the way to go.  And it’s a wonderful idea, I know.  Hell, Gartner even created a whole new market segment just for those vendors who were able to convince them that was what they delivered. 

Well forgive me if I’m a bit slow, or may have missed something along the way, but from my perspective the promise of self-service BI (as typically defined) will not be fulfilled by a specific vendor or product.  Why?  Because the audience these tools are meant for (generally business people) is completely diverse in terms of technical ability and experience. 

No one is going to make a tool which you can give to a middle aged clerical person who doesn’t own a smart phone and is barely competent with Excel and expect them to whip up their own dashboard (that yields reliable or useful information).  It just will not happen.  It doesn’t exist now and no one is going to deliver it in my lifetime.  I’m speaking from experience.  I’ve seen lots of very frightened non-technical folks in classes trying to learn a reporting tool or dashboard tool and knowing no more when they left than when they walked in three days earlier.  The dread on their faces as they leave knowing they are expected to be able to handle this on Monday morning is horrible to behold.

And by the way, the tools which claim to be able to do self-service BI still require a good deal of effort from IT or very technically savvy power users to access, model, and integrate the data sources, set up security, and create some preliminary elements which can be reused by those wishing to create their own reports, dashboards, etc.  This work needs to be done nearly every time a new use case is defined or new data sources need to be folded in.  It just is not as easy as giving the business people a tool and telling them to go get whatever data they need themselves and do the analytics which yield spectacular insights while they’re at it.

Yet this is what some vendors want you to believe.   Try that and the average business person hits a wall the moment you tell them they will need to join two data sources.  This is NOT a knock on business people.  Business people know their jobs well, but most have modest technical abilities at best, so telling them to join two data sources is like telling them they will need to build a rocket ship and fly to the moon.  Where do they begin?  This is not what they’re trained for.  And we haven’t even broached the subject of predictive analytics yet!

I think “self-service” needs to be redefined.  It’s a fluid term which means different things to the various kinds of users in your company.  There ARE some technically savvy folks who know data and databases and are adept enough with technology to handle these sexy self-service tools, but they represent a narrow slice of your employee pie.   For the majority they will need some combination of pre-built reports, interactive reports (with prompts or filters), some type of cube technology which allows them to navigate within a pre-defined multidimensional data base, or dashboard/scorecard type tools to serve up interactive data visually.  Giving the right people the right combination of these offerings is the best way to achieve self-service.  And make no mistake, all of these require investment of time and resources from IT to design, build and maintain.  That’s just reality.

New Call-to-action

One final but very important point: user adoption of these tools is unpredictable at best.  Many will balk at learning a new tool which is unfamiliar and often intimidating.  Some will not grasp the underlying concepts necessary for using them effectively.  Others will see them as too time consuming and take away too much time to do the job they’re being paid for (often sales people like me). 

The solution here very often is to embed the information and analytics into their existing applications and daily workflow.  Don’t make them learn a new tool or open up another browser session.  Don’t give them more work to do.  Instead, give them the relevant data they need at the moment they need it wrapped within the applications they use…and give them the ability to interact with it as need be.  This kind of embedded analytics takes the idea of self-service BI and turns it into….”Intravenous BI”, if you will.  This feeds them the specific data and insights  they need without them needing to do any additional steps and does not require them to disrupt their workflow or learn new technologies.  For the business this maximizes the impact of the analytics because it removes the user adoption barriers and mainlines the powerful insights directly into the bloodstream of the company.

So make sure you understand what you’re signing up for if you decide to chase the self-service BI dream.  If you’d like to discuss further what I’ve seen and what I know about all these competing solutions and how to embed BI into your business you can reach out to me here and I’d be happy to help in any way I can.  

Let me know your thoughts! 

Browse

Article by channel:

Read more articles tagged: Analytics, Featured, Predictive Analytics

Data & Analytics