Self Service BI, where is the demand for this?
Published:
Content Copyright © 2016 Bloor. All Rights Reserved.
Also posted on: The IM Blog
It appears that the big trend in BI tools over the last few years has been for self service BI tools, self service data blending, self service report generation, self service data mining, you name it then it is to be had with little more than a point and a click. I find this all very strange. I work most of my life delivering BI solutions to a wide variety of companies throughout the UK and Europe, and I see no demand from the business to be given the tools to enable the business to get into the BI DIY business.
I do see a lot of dissatisfaction with the service that the business receives from IT to its demands for more responsive solutions, delivered more quickly, that more closely match their real need, but that does not translate into “give me the tools and I will do it myself.” Often the business starts to recruit its own BI staff, and they start to build “quick and dirty solutions”, on the basis of they are “tactical” and the IT guys will replace it with a strategic solution, but that never happens, so the tactical solution remains in place with an ever increasing cost of ownership.
I run the majority of my projects as agile scrum projects where we ask that the business provide a Product Owner, someone who owns the problem, to help us deliver the right solution. That requires that they attend a daily stand up for 15 minutes, and every two weeks attend a one hour showcase of what will be the answer to their problems; and most of them cannot make those meetings regularly because of commitments to their day job, so how will they find the time to build a BI solution?
There is a significant group who do build their own solutions all of the time, and they are finance users. They blend data and create myriads of spreadsheets. They do this because they need to tackle the problem that there is no “single version of the truth” in a business, but always at least two, the data warehouse and the ledgers in the financial systems, and the two do not align. They love Excel because they can manipulate numbers to create the answers they need with minimal governance in the IT sense of data governance. The problem of Excel is another interesting topic and shows the potential dangers of too much DIY BI.
Despite all of this I like self-service tools. I give them to my BAs and developers as a means to work more closely with the business, to get rapid results, and to build prototypes to test out understanding, but from within IT so that we can govern what goes on and ensure that the total cost of ownership is managed. The other thing that is great about many of the DIY type tools is that they are visual, and they help the business to browse the data, and as they do that they come to take some sense of ownership of the data; and maybe one day if the business comes to own its data and not think that is an IT responsibility DIY BI will make sense.
Comments are closed.
I’ll add that without a self service strategy and roadmap, self service BI puts enterprises at risk of building departmental silo’s of data – all over again. The most successful use of self-service BI will be IT as the consumer of these solutions in the context of a larger BI framework – producing solutions in a broader but well governed way for the organization.