In a previous post, I outlined my framework for tackling any new analytics project or problem:
Step 1: Clarify your questions
Step 2: Catalogue your data sources
Step 3: Correct your data
Step 4: Chart your data
Step 5: Communicate your findings
Now, knowing that I live and breathe Tableau Software, it won’t come as any surprise to you that this framework is dotted with references to the Tableau suite of products. For those of you still feeling your way through the fog of data visualization and business intelligence tools, simply replace ‘Tableau’ with your product of choice (although you could save yourself a whole lot of heart-ache and just get on board with Tableau…. but we’ll leave that for another post!)
Let’s have a deeper look at the first three steps of my framework:
1. Clarify your questions
You have to clarify the questions that the business needs answered. Think about your organizational Key Performance Indicators (KPIs) — what are your company’s measures of success? Turn these into questions.
If a KPI is to improve performance of “X Service”, then what questions could we ask to measure performance? What questions could we use to understand the factors contributing to that performance and identify areas for improvement or areas of success to be replicated across the business?
These are the real questions that need to be answered and once we have clarified these questions, it is time to think about how we’re going to answer them. If business problems and questions are clearly defined, this will make identifying the data required (and therefore the data sources required) a lot easier.
2. Catalogue your data sources
With a clear set of questions, the task of defining and then sourcing data becomes a lot easier. I would encourage everyone at this stage to stop, think and then think again. Think past the data sets that you know already exist and think about the range of systems and processes in your business that generate and collect data.
Does your Human Resources system capture data that could be used in conjunction with data from your Sales system to identify key characteristics of sales people who outperform others?
Does your Accounting system capture data that could be blended with your customer data to identify trends in activity based on geographic locations?
If you are part of a team or department of analysts, I strongly recommend taking the time — as a group — to collectively think through your key enterprise systems and data sets and invest some time in preparing a suite or range of data sources which bring together the extensive content and data knowledge of the team. These new and expanded combinations of data sources will allow for some fascinating analysis and exploration later on!
3. Correct your data
A key step in working with Tableau Desktop is the creation of data sources. This process is discussed in detail throughout this book, however, it is worth noting here that this step will take some time. The majority of clients I work with underestimate the amount of time this step will take. Yes, Tableau has some fantastic features and functions to make this step easier, but there are still a number of things we need to do to prepare our data to make our analysis and exploration in Tableau Desktop as smooth as possible.
However, most data preparation tasks are largely a “once off” activity that can be reused by different Tableau worksheets and dashboards if done properly. These data sources can then be shared across the team, indeed the entire business, allowing fast connection to well-prepared data to support meaningful analysis.
As more and more data sources are created and shared, working with Tableau Desktop becomes much faster and smoother, data blending opportunities become greater, and the analysts’ role shifts from “data dude” to “insight guru”.
Stay tuned for the final two steps – Chart and Communicate – in an upcoming post.
Join my next masterclass!
Register now to join my next Tableau for Business Masterclass - a free webinar packed with tips to take your Tableau skills to the next level!