Subscribe

Why BI projects fail to deliver value

By Gustav Piater, Sales and Marketing Director, AIGS (Yellowfin South Africa)
Johannesburg, 08 Mar 2017

ITWeb Business Intelligence Summit 2017

Meet Gustav Piater and his colleagues from AIGS (YellowFin South Africa) at the 12th Business Intelligence Summit in March 2017. AIGS is one of the proud sponsors of this annual event. Registration is open and the agenda is live. Click here for the most updated agenda.

Over the years, business intelligence (BI) has helped tens of thousands of enterprises gain a better understanding of the gaps and opportunities in their markets and organisational make-up. Nevertheless, it has been slow to permeate throughout organisations.

This is because few tools offer ease of use AND technology leadership. As such, BI remains strong among data scientists and analysts, but has not convinced business users - who comprise 90%-plus of the enterprise.

However, it's a brave new world, and technology is catching up to user expectations, as can be seen in the higher adoption rates of the more successful tools.

BI selection criteria

But has the 'adoption' pendulum swung too far? In the search for accessibility, have vendors neglected technological innovation? What else should organisations look for in selecting a BI platform?

The best tools focus strongly on user enablement, offering ease of use, self-service BI, collaboration and mobility. On the other hand, they strive unrelentingly to be at the cutting-edge of innovation, with support for multiple data sources and robust data governance that supports both user accessibility and data integrity.

Finally, while a tool might tick all the right boxes for most companies, it may not be right for you. BI that is powerful, accessible and fit for purpose in your organisation should tick all the following boxes.

Accessibility

First, there is accessibility. It begins at the front-end, with a good mix of presentation modules including reporting, dashboards and storytelling.

Originally, BI was seen as a means to create reports for decision-makers. But many tools did not do reporting well, so companies needed data and IT experts to do data extraction and manipulation. The more modern BI platforms possess a strong, easy-to-use reporting component, offering business users decision support without having to be great at Excel.

Users also demand great visual presentation in the form of dashboards. While some tools offer attractive visualisations, they are based indirectly on reports. The best ones tap directly into data sources, turning slick visualisation into something more - decision support at the middle layer.

At senior management level, tools should offer storyboarding, again with the same set of data.

Technology leadership

Then there is innovation, starting with governance. With the abovementioned tools, non-technical users can easily perform BI self-service, but if not supervised, they can present a great risk to organisational data integrity.

The way leading tools get past this is by allowing view-only access to core data. Users can manipulate data into different views, but they cannot touch (alter) the underlying data itself.

Another issue is cross-platform support. BI is no longer about understanding only your CRM, ERP or other business-critical application. It needs to span multiple data sources - from traditional applications through to unstructured social and machine data.

To let information workers, management and the executive work together seamlessly, tools must further have a strong collaborative focus, offering task management and social media-style annotations (comments).

And of course, if the tool isn't mobile, it may as well not exist.

Right for you

When BI tools are both accessible and cutting-edge, backed by control, there's a good chance they will find favour with everyone in your organisation. But how do you know a specific tool is right for you?

The best tools combine vertical industry credentials with cross-industry success, and a global track record with local development, infrastructure and support.

Also, look for differentiators like cloud subscription models, speed of development, and a low resource-per-user requirement.

With all these boxes ticked, your choice of BI tool should enjoy very high uptake in the organisation. If penetration is 15%, it is poor. If it's more like 30-40%, sign on the bottom line.

Share