Collecting data is easy. Getting people to use the data is the hard part.
Most data tools require specific skills and experience to be used properly. This places access to data in the hands of a small group of specialists, and out of the direct reach of the teams and decision makers who need it most.
As the demand for data-informed decision making grows, so does the pressure on the data specialists to support their colleagues.
To reduce the pressure, we need to democratise the data. This means making it easier for people to use data in a self-service way.
By making it easier for more people in the organisation to access and experiment with data, you’re able to build a more diverse user group, which increases creativity and (in theory) enables better, more efficient decision making.
Meanwhile the data specialists, freed from the daily churn of data requests, have more time to focus on improving data quality and reliability.
The question then, is how do you guide a non-data specialist through the process of writing a query, validating the results and interpreting the insights?
One approach we’re looking into is using AI chatbots to help non-techie users define what they need to know and query the data sets. It’s a similar approach to GitHub Copilot, which is used by software engineers to write code more efficiently.
There’s lots of potential with this approach, but also lots of challenges. Privacy and data ownership (especially when using 3rd party technology) are a major consideration. As are the quality of the data insights and the user experience, particularly with the chatbot and the visualisation. Another factor is keeping an audit trail and validating the data queries that are used to make key decisions.