How far along are organizations on the path to self-service business intelligence? Most organizations want to provide self-service to provide more flexibility to business users, as well as free up their IT departments from having to build and crank out reports. This is especially important in making hay out of Big Data. But there's a lot of work to be done before this becomes a reality.
Recently, I had the opportunity to deliver a Webcast, in conjunction Suzanne Hoffman of Tableau Software, on the state of self-service BI. My presentation is based on a new survey I helped develop and publish as part of my work with Unisphere Research, a division of Information Today, Inc. The survey included the responses of 250 IT managers and professionals, subscribers to Database Trends & Applications.
In the survey, close to half of the respondents, 49%, say they have some form of self-service in their organizations, in which end users are able to access approved enterprise data sources and create their personalized reports and queries with no intervention from IT or other departments. However, drill down a bit further, and we found that only 10% report this capability is offered “extensively,” compared to 39% who say it is offered in limited situations. So organizations have our work cut out for them.
Only a very small handful report having tried it and discontinued a self-service attempt, mainly due to data quality issues encountered. But data quality is a very important challenge to the self-service concept. In fact, among those organizations that do have self-service BI, 41% are struggling with data quality issues, making this their leading challenge.
Respondents also told us they need more budget support to move such initiatives forward, as well as building a solid business cases -- pointing to issues around organizational buy-in. And a third of respondents say that it is a matter of getting the end users themselves to be accepting of a self-service BI approach, versus doing their own workarounds with spreadsheets and the like.
You could argue that self-service BI has been with us for a very long time. The spreadsheets that popped up in the 1980s were a classic example of self-service apps created and run by business users. And the reports delivered by IT departments are no longer on reams of paper -- much of it now can be delivered as PDF attachments to email. Organizations may be saving trees, but the problem is they're not providing any additional value with rigid, per-configured reports that don't tell business users all they need to know.
But when we talk about having a good self-service BI tool in this context, we're talking about more than just an interface that someone throws up for some type of one-off analysis or activity. Effective self-service BI tools need to provide consistent, verifiable views of enterprise data. They need to be interactive and personalized – delivering exactly the information an end user wants. Ideally, they should also enable collaboration – that is, others can view the same information that the user originally created.
We know business users are very frustrated with the glacial pace that's involved with sending a change request for a report to IT, and waiting and waiting for the results in a future report delivery. It shouldn't be difficult to make changes to a data report. Businesses are moving way too fast, and what is needed is the ability to do rapid and flexible data querying across multiple data sources.
Many users these days are accustomed to the speed of delivery of Google Analytics, for example. Just set it up on the Google site, and away it goes.
Here are issues identified in the survey:
- A majority of respondents, 52%, cite issues with the latency of information—which they say needs to be delivered faster or be more timely to make a difference.
- Another two-fifths of the group report that members of their organizations are experiencing difficulty with the complexity of the BI tools and solutions they currently have in place.
- About two-fifths also report that they and their end users have difficulty finding the right information needed from data stores across their organizations. Thus, the need for organizational support rears its head. Along these same lines, we see 37% reporting that they face budget headwinds in attempting to improve their delivery of BI information.
Decision makers or information workers may be under-served by many existing BI environments, in which access to data is out of reach to all except IT and selected power users. Reliance on a few information gatekeepers to answer critical business questions results in decision makers’ frustration with time delays, while burdening already stressed IT departments.
Of course, the data and information seen in the enterprise is more complex, and buried in a lot more proprietary systems than the simply available data on traffic and trends across the web that is captured in Google Analytics. The challenge is working with the owners of the various systems which house the data needed for analysis, and providing it in a format that can be digested by self-service BI systems.
Ultimately, what is needed is a unified view of the enterprise. There tend to be many different types of BI and analytic solutions and tools scattered across enterprises. Since many critical decisions may come out of self-service BI dashboards, it's especially important that the data being fed into these systems is verified and of the highest possible quality.
This requires working closely with data owners to ensure that the data they are contributing is timely and of high quality. This requires support from across the organization, especially among the leadership of organizations. Current traditional BI systems and infrastructures may be clunky and non-responsive, so it's no surprise that we see a high level of support toward eventually enabling end users to build their own interfaces to BI and analytic data.
(Disclosure: The survey and Webcast were sponsored by Tableau Software. I was compensated for my participation in both projects. I am alos a contributor to Database Trends & Applications.)
(Thumbnail photo: Joe McKendrick.)