How to Adopt a BI Tool for Every Tier of the End-user Pyramid
Dec 5, 2018
Originally published with Innovation Enterprise.
In an age when everyone feels busy all the time, adopting a business intelligence (BI) tool can seem daunting. To be sure, each BI tool is different and higher levels of adoption require specialized training to master. The good news: There are multiple levels of end-user adoption, and only the heaviest users will require significant preparation.
BI adoption levels run parallel to BI user types. Think of it as a pyramid: At the top, there are data scientists, then analysts, then explorers, and at the base are consumers. The scientists and analysts at the apex are power users. They comprise about 10% of all knowledge workers in an organization and are masters of querying languages, statistical analysis, databases and modelling.
The casual users, about 90% of the workforce, make up the large base of the pyramid. These are either data consumers or data explorers who need information to complete their jobs but who aren't focused on analyzing data. Most of these more casual users are executives, managers and individual contributors with little database system experience.
If you fall into the latter category and are overwhelmed at the thought of taking your first foray into BI, don't worry: Follow these four steps and work your way up as you see fit.
1. Become a data consumer
Start simply by reading the reports already being published. Ask to be added to your company's listserv and sign up for scheduled reports. Run preexisting reports in your hub application, whatever that may be. If you're not sure which ones apply to this role, ask a colleague.
Start a mini report library for yourself, and access it easily via convenience tools like bookmarks and folders. Try changing the way the report displays using interactive filters, sorting options, conditional formatting and styling tools. These changes are superficial and do not alter the underlying report. Trying them, however, out can be a great first step toward making structural changes.
2. Act like a data explorer
Modify a premade or "canned" report, a convenient springboard from which to make context-specific modifications. If you always filter a particular canned report by country and then sort it by employee last name, for example, try duplicating that report and modifying the copy's definition to include your preferred sorts and filters by default. This saves you time while also familiarizing you with some of your BI application's reporting features.
Once you feel comfortable with making these kinds of modifications, progress to swapping out fields and changing the report structure. Once you can confidently do this, you're on your way to creating a brand new report from scratch.
3. Learn how to be a data analyst
The analyst role is different from those typically occupied by data consumers and explorers: Handling data is part of an analyst's job. If you find yourself taking on more analytical responsibilities, ensure that your system administrator has given you full access and start tinkering with the advanced options.
For maximum efficiency, begin this exploration process with a series of end goals in mind. Advanced tools might include such things as pivot tables, drilldowns, report linking, advanced sorting and filtering, conditional formatting, complex formulas and visualization designers. Which of these you utilize will depend on what you hope to accomplish.
4. Study how to serve as a data scientist
The difference between the scientist tier and the analyst tier in terms of BI usage is proficiency in understanding data models. As a data scientist, you might adjust a report's default join pathing logic or design a suite of reports around a data experiment you're running.
You might even request access to new data sources and have a hand in how they are aliased in the BI application. You should also have a deep database-level knowledge — that's what sets you apart.
Understanding a BI tool comes from the top of the pyramid down. Power users facilitate casual users by building and disseminating reports that casual users are too busy or specialized to build. Enabling end users at every level will make BI tool adoption far more manageable, even for the very busy.