Treasure hunting for data
Recently, I found myself in one of those meetings that go around in circles.
We were considering adding a feature to a dashboard that would show the difference between when an event happened, and when the business was informed. I felt like I was repeating a simple question. ‘Do we have the data?’
Data means different things to different people. When asked ‘Do we have the data?’ business users might answer ‘Yes!’ because there’s a form that they fill out or an excel spreadsheet they save. But where is it being captured?
On the other hand, the BI developer might say ‘No’. That data isn’t in the warehouse we’ve been given access to, or perhaps it’s hidden in plain sight, in one of the many columns with almost the same name. Or maybe it’s all free text, filled in by hand by different people and too inconsistent to analyse.
Before you can start development, you need people from different roles to ask, ‘do we have the data?’
Talking to team members one at a time often results in multiple answers, so facilitating a group discussion is more likely to reach a definitive answer.
The goal of the conversation is to get a clear understanding of one of the following scenarios.
- We don’t have the data
- We have something like data, but it is messy, incomplete or unreliable
- The data looks great but isn’t in the database
- The data is already connected to the dashboard, or easy to import
An analyst can move forward with any of those outcomes. It could mean helping the business collect data or addressing the business goals with a different data source. It could mean adding something to a database or writing scripts to clean data.
I often find myself describing what makes data machine-readable, which is essential for data to be used in a dashboard. Often, this means more numbers, more consistency and less text than business users expect. Sometimes we find a business expert who can immediately point us to a treasure chest of golden data, and other times we need to dust off our digging gear and prepare an expedition force.
Either way, it is crucial to get stakeholders and the project team together to ponder.
‘Do we have the data?’
Get in touch with Christine Dixon
I came to analytics via biomedical research, including drug screening, biophysics and genetics, with a heavy focus on science communication. Data isn’t interesting by itself; you need creativity to build the transformations, metrics and visualisations to turn it into meaning. Show me your data and let’s build a story together.Get in Touch