Before beginning any analytics project it is advisable to consider these 7 common analytics avoidable shortcomings*. Percent represents % of respondents.
- 36% Software does not provide enough flexibility in accessing/analyzing/manipulating
Most of the top analytics platforms like Tableau, Qlik or Microsoft BI deliver the flexibility required to meet the majority of analytic requirements including Ad Hoc capabilities. Be cautious of proprietary analytics solutions that may come with restrictions limiting user manipulation, filtering, sorting, or system access.
Likewise, look for solutions that easily queue up commonly used filters and sort options that can be saved for later usage! See example below showing commonly used filters on right side of page.
- 31% User Interface is not Intuitive
Like flexibility, user interface issues have been addressed with most major analytic platforms. When assessing a user interface, make sure multiple users have the opportunity to “test drive” the system. Evaluate potential training time and training issues for less experienced users.
- 30% Not all relevant information is aggregated in a manner that is required
Accurate and complete data sets are often the most challenging aspects of an analytics strategy. This is particularly the case when a multi data sourced solution is required. Engage assistance from firms who are familiar with the data extraction process, and are familiar with your data sources.
In healthcare, for example, EHR/PM vendors may be well equipped to support their own data sets but provide no assistance for data outside their realm of expertise. Look beyond legacy vendors to provide an enterprise analytics solution, one with a broad breadth of experience in data extractions and data visualization from many data sources.
- 30% Data is not updated frequently enough
Determine how frequency you need data updates. Off the shelf analytics solutions will vary between daily, weekly, and monthly depending on the KPIs. Some may even require hourly updates. Don’t assume all frequencies can be easily supported. Data updates may also adversely impact system performance.
- 27% No support for predictive analytics
Today, improving Predictive analytics modeling capabilities is a major focus for analytics platform vendors and content providers. Organizations are turning to predictive analytics to help solve difficulty problems and uncover new opportunities. While predictive analytics is still maturing, look to ensure your analytics platform provider is committed to developing the tools and content to support your longer term business needs. The secret to effective predictive analytics models is the accuracy and confidence level of the underlying data (see #3)
- 25% Data is not granular or detailed to identify root cause
Drill down capabilities are essential to identify root cause and solve difficult problems. Ensure your analytics has the ability to easily drill down to a granular level of data directly from high level analyses. Solving and preventing issues requires a level of granularity not required in monitoring issues.
- 20% System Performance is too slow
Make sure system performance supports user needs. System performance is one of the biggest factors in analytics customer satisfaction (as most analytics solutions are designed to be interactive. ) Today’s analytics platforms, like Tableau, run at least 10 times faster than traditional legacy solutions.
*Source: IDC Global Technology and Research Organization