Recently, it seems that every article published about the future of Business Intelligence expounds the power and benefit of dashboards vs traditional reporting. It seems that the methods of building complex queries and data cubes is beyond the reach of the average business user and the only solution is to “dumb down” the output and interaction methods. The answer – so far – appears to be dashboards.
Dashboards conceal the queries and technology enabling a business manager to quickly understand the current state of the organization. The user can drill down into details and calculations that were fabricated and assembled by a more technical user. However, if the user has a question about the data beyond what’s present in the dashboard, BI promptly falls back into the world of a technical person constructing a picture of the truth at the request of a business user.
There may be another way. Rather than seeing dashboards as an end, they must be re-established as a means. The very nature of a dashboard is retrospective or real-time at best; a car does not predict the speed of your vehicle five miles down the road, but rather informs the driver of his/her current velocity. Dashboards capture the decisions that were already made and fail to provide any recommendations to the user as to what decisions to make in the future. Dashboards are “easy to use” simply because they provide little functionality, not because they have clearly articulated a specific business intelligence need.
Dashboards are a stepping stone to the future of business intelligence. A necessary step, but a costly one. Companies who see dashboards as the end game will spend exorbitant amounts of money creating or implementing a “solution” only to find out in the months and years to come that dashboards provide the illusion of intelligence. Information delivered more quickly and in a graphical format is not intelligence. Dashboards still require the human user to interpret, build consensus and make decisions.
What if a truly intelligent business system could provide recommendations around decisions? We’re not talking about notifications, but actually recommending specific actions around a business condition. Instead of generating a dashboard of sales trends, a smart system could provide the recommendation for how to act on the trend. If this is the case, viewing the dashboard is only necessary to understand why the recommendation was made. The user no longer needs to interpret the data, but simply build consensus around the recommendation (do we act or not?) and take action. The idea of digging through dashboard after dashboard seeking the truth will seem arduous and unfulfilling.
Dashboards will continue to be an integral aspect of BI software as a means of quickly conveying complex information in a simple format. Why a user will need to view a dashboard is an evolving question.
Tags: Business Intelligence, Dashboards, Decision Support, Ease of Use