I believe in the power of big dreams. I really do.
Want to end world hunger? Awesome. Want to make sure that everyone in your city has a warm/dry place to sleep at night? Fantastic. Trying to build schools in areas devastated by global conflict? Go for it!
But a dream dashboard?
There is another name for that…
The problem with dream dashboards.
So this is how it starts.
Someone asks this question, “what do we want to see?” Then it’s followed by conversation after conversation, draft after draft, and dream after dream.
Then ultimately, usually months or years later, you figure something out. Nobody really had the motivation, or the money, to make the dream dashboard actually work.
This isn’t to say the “what do we want to see?” conversation shouldn’t take place. It’s a great thought exercise and can help you identify the limitations or gaps in your data collection strategy.
It’s just not a very efficient way to start building a dashboard.
What to do instead.
To make something successful efficiently, you need to start with what you have, not what you wish you had.
To make a useful dashboard you need two things.
- Access to data.
- People who could benefit from the data if it were made more accessible.
Just having one of these two is not enough for success. There is no value in making data accessible that nobody wants. And there is nothing more frustrating than a dashboard that promises something beneficial but doesn’t have the underlying data to deliver.
An example and a challenge.
When I was in the nonprofit world as an early childhood data specialist in North Carolina, one of our most important datasets held child care facility star ratings. These star ratings were the basis for many of our performance measures.
Our local agencies representing NC’s 100 counties would provide direct and indirect technical assistance (TA) to child care centers and homes. The goal was to improve the quality of child care. And the star ratings could play an important role in helping TA providers target their services.
There were different ways to get this data but most local organizations would pull the data directly from the NC Division of Child Development Search Site. At the state level we had direct access to the data warehouse behind the scenes. But for the individual counties, it was a pretty manual process of searching and copying/pasting data that would be updated intermittently.
The dataset is not overly large but not small either. There are close to seven thousand licensed child care facilities that can be found through the search site. For TA providers, the search site isn’t really a user friendly way to find the information they need quickly and efficiently.
So here is our challenge, can we make the data available through the search site more accessible for local agencies and TA providers?
Basic Zip Code Dashboard.
I made the dashboard above using the NC Division of Child Development Search Site, Excel, and Tableau Public. If you click on the image you’ll be taken to Tableau Public to see the interactive version. I didn’t use any other prepared data, fancy tools or plugins.
Up for a challenge? Can you get to something similar starting with the search site?
A few tips.
- What happens if you hit the search button at the bottom of the search site without adding any additional parameters?
- Audience focus is important in dashboards. We’re creating this one for TA providers, they’re going to be more interested in centers and homes with 3 stars or less. 4 and 5 star centers are already considered high quality.
- If you take the time to clean your data properly, Tableau becomes much easier to use.
- I think most locals would choose to see their data at the county level, not the zip level. But zip was easier to isolate. Start with what you have. If you need to iterate or adapt a dataset, you can always do that later.
I’ll have a way for you to see all my steps from “before to after” next week. Until then, let me know in the comments if you decide to give the challenge a try.