My non-techie spouse just built his own small business website for a fledgling custom guitar-making business. All by himself. It did not take him long – some dragging and dropping, a little typing and a picture upload later – and it looks good. Ten years ago he would not have even attempted it and instead […]
My non-techie spouse just built his own small business website for a fledgling custom guitar-making business. All by himself.
It did not take him long – some dragging and dropping, a little typing and a picture upload later – and it looks good. Ten years ago he would not have even attempted it and instead pleaded with anyone he knew that was remotely technical to create it for him. Today the tools to build a basic website are inexpensive or free, intuitive, and produce decent looking results quickly. I’d say the same about modern Business Intelligence (BI) tools.
Earlier this year, Gartner officially separated their evaluation of BI platforms into two groups for their 2016 BI Magic Quadrant. They used the terms Modern BI and IT-Centric BI to differentiate the two and explained their rationale as:
“… (the) multiyear shift of focus from IT-led reporting to business-led self-service analytics has reached a tipping point. Modern BI platforms support organizational needs for greater accessibility, agility and analytical insight from a diverse range of data sources.” — Gartner BI Magic Quadrant, February 2016
The big news was that now BI platforms had to qualify under Gartner’s new Modern BI criteria the previous, all but one of the traditional big BI vendors dropped into other quadrants, leaving only Tableau, Qlik, and Microsoft to bask in the glow of Leader Quadrant glory.
According to Gartner, modern BI platforms are:
“…easy-to-use tools that support a full range of analytic workflow capabilities and do not require significant involvement from IT to predefine data models upfront as a prerequisite to analysis”.
As with any change to the usual way of doing things, there are pros and cons with Modern BI tools. Yes, Tableau, Qlik, and others are enabling non-techies and techies alike to confidently create impressive reports and dashboards. It’s empowering, it’s efficient, and it’s driven the market-changing popularity of these types of tools. I agree there’s a lot to love about the drag and drop onto a map option in Tableau!
Now remember the saying “with great power comes great responsibility”? In this case, the “great responsibility” is a de-centralized management of the underlying data, especially when mixing/blending data from multiple sources. In the now named “IT-Centric” BI tools like Cognos and Business Objects there is typically a semantic layer (usually created and curated by IT) that creates a data map of all the data elements to be made available for analysis and how they connect. The semantic layer can prevent incorrect data joins that would cause queries to fail or produce erroneous results. The semantic layer is typically not editable by BI users who want to ask questions of the data and most prefer that it is pre-mapped for them.
There are two advantages to a single, centralized data mapping layer:
The limitation with this model is that adding a new data source for use in reports and dashboards usually means IT-involvement.
In modern BI platforms, the report/dashboard creator usually has total control over the data they want to use and how the data is joined or blended together. If you are simply pointing a tool like Tableau to an existing datamart or a single data source in Excel the data quality risk in these scenarios is low.
If you are joining data from multiple sources (the enterprise warehouse, datamarts, a big data repository, an Excel spreadsheet) the data quality risk increases. To map data for reports and dashboards successfully you need some data integration skills in order to identify the relationships between the different data sources and to understand which data elements should be used as a common key to link them.
BI Power Users welcome this ability because it gives them more control and flexibility to mix and match sources. There’s still a risk that one power user can join the same data differently and enable a different result. Establishing and communicating standards can help here.
Less technical BI users may not want to blend their own data sources or if they do they should spend some time learning how to appropriately map the data together to avoid inaccurate query results. Getting assistance from IT or a BI Power User experienced in using this dataset to understand the dos and don’ts are recommended.
If you are interested in learning more about finding the right BI tools for your organization, contact me at [email protected]
Look for our upcoming post on free BI tools – we put 5 tools through their paces, see which ones triumphed and which fell short.