Over the years, I have seen several organizations begin to lose trust in the data behind their carefully crafted Tableau dashboards. It can be easy to get to that place, but it’s completely unnecessary. Why get to a point where you start jeopardizing sustained levels of user adoption because data source usage is out of control? Maybe you’re at that place now or perhaps you want to know how you can avoid becoming desperate. There’s a better way to document your data sources and bring established methods to the madness.
Workbook Performance Deadline
Let me run through the details of a scenario that played out recently with one of my clients. The client was looking to outside consultants to provide insight on getting better performance in their workbooks. Sadly, they kept coming up empty-handed. Their deadline was getting closer and they needed someone who was willing to take on a seemingly impossible situation. So, they turned to InterWorks, and I was the one tasked with helping them get a better understanding of how they could turn things around.
I immediately realized that we needed to focus on how they were using calculations. Their particular set of workbooks had over a thousand calculations and the logic was very complex. Figuring out a way to reduce the number of calculations was a big deal in and of itself. But we also needed a way to monitor how calculations were changing on a regular basis. I could see why they hadn’t set out to do this sort of thing before. It’s a daunting task to try and build out documentation for that many calculations, plus keep it updated on a regular basis.
Calculation Logic with Data Source Audit
The good news is that I was able to leverage a tool that InterWorks had already built. It’s called the Data Source Audit tool and it can be found in Power Tools for Tableau: Desktop. If you haven’t checked out a free trial yet, you can do so here.
In less than a minute, we were able to use this tool to export a list of calculation logic in their workbooks on Tableau Server. We were able to quickly determine where the calculations were coming from. In this case, it was either going to be coming from an Excel worksheet, the Salesforce system or a data warehouse.
The Excel export from our Data Source Audit provided a list of all their calculations. We distributed the list to their finance team and asked them if they agreed with the calculation logic. After receiving feedback, we were able to make adjustments and reduce the number of calculations. They could easily run the exports on their own whenever they needed to go back through the process of verifying their calculations. This meant that documentation would be visible to the end users on a regular basis.
There is also a great Tableau workbook export that is probably my favorite option in the Data Source Audit tool. It’s easy to publish to Tableau Server where everyone can see it. I also like to use the workbook extract and make some of my own custom views that show calculated fields along with some additional notes about the logic. I’ll typically add a tab to that before I publish it as a glossary.
Discovering Volumes of Information
The Data Source Discovery tool, also in Power Tools: Desktop, goes hand-in-hand with Data Source Audit. It helps boil down a large volume of information about your data sources. In this project, we ended up using it to answer some other critical questions. There were over 100 fields in one of the data sources. We were trying to find out in which field we could find this piece of information. Instead of having to make more calculations, we had several instances where we were able to attain the necessary information from an existing field. The Data Source Discovery tool helped narrow that down.
Above: The initial data source selection screen in the Data Source Discovery tool.
This project ended up having a drastic impact on performance, and we were able to resolve their issues before the looming deadline. You may be in a similar situation where you’re trying to avoid some really expensive, potential band-aids. Perhaps you’re in a better position where you’re choosing to invest the time up front in data architecting. These tools will help you in either situation and we would be glad to walk you through some examples of how they could be used in your environment.
If you want to learn more about these features and more, head over to the Power Tools for Tableau: Desktop page and try a free trial for yourself.