We mentioned last week that 70-80% of BI projects fail and provided 5 key tips for making sure yours isn’t one of them. Now that you’ve had a little time to let those recommendations sink in, I have the next 6 suggestions—keep them in mind to use business intelligence to your advantage and get the most out of your BI investment.
Identify meaningful slices as a starting point
Do your data segments need to be redefined? Have they been through a recent review or validation? Nailing down the segments and data filters by the most meaningful cuts is one way to ensure a starting range of business analysis that you’ll be happy with.
Wrangle your raw data
While there are infinitely more data sources, formats and feeds sprouting from the cosmos of vendors out there, you can only analyze what you have. And to analyze it, you’ve got to corral it in and get it into the right format. Need help with this one? Some of the most amazing advances in the BI sphere deal with tool sets that can help you wrangle data to draw insight.
Consider an analytical data model
Gone are the days of kitchen sink analysis. Bringing in entire relational or transactional data sources and looking at record level data is tempting to many analytical minds, but unlimited data access has a price – slow processing speed, longer report development time and a larger learning curve when it comes to digesting the information. Your data model should be optimized for analysis to ensure quicker processing, faster speed to insight and business decisions.
Streamline data views
What data is most important? Consider listing out all of the metrics you want to measure, ranking them in their order of how actionable they are and using that prioritized list to inform the universe and structure of the sources of data you choose to analyze. Timing is also an extremely important consideration when it comes to shaping your analyzable universe. How often will you refresh your reporting? How far back will you look? All of that informs potentially large amounts of work on the back end via ETL (Extract Transfer Load) processes and real time engines.
Trending is Trending
This is a big one and one that people often lose sight of while hacking through the thick of the data jungle. Have your trends in mind from the get-go to make your conclusions and correlations painless. Make sure that those trends tie to notifications sent to key decision makers to enable action when needed. Not sure where to start? Some fundamental trend windows include month-over-month, quarter-over-quarter, and year-over-year cadence.
Make accessibility a priority
Who is reviewing the data and through what interface? Users today expect intuitive data accessibility anywhere, anytime. Make sure that your data and tools align with this expectation and that capabilities are in place to share relevant data on the go via web browsers, mobile devices and tablets. When designing an output, factor the interface into the design. Ask yourself the question – would I rate this 5 stars in the app store? If you’re still not clear, try opening a big Excel spreadsheet on your smart phone. Ew, right?
It’s a wrap
There you have it. Once you’ve addressed these 6 tips (and 5 tips from last time), you’ll be well on your way to creating an effective ecosystem and successfully using business intelligence. So remember, keep it real and find your sweet spot.