Making data usable to end-users has been a challenge for many years. Many public organizations have invested in data collection and analysis with the goal of improving decision making, opening their data up for outside use, and streamlining internal processes.
However, usability remains a barrier to wider adoption of information assets across an organization. In this interview, Donald MacCormick — Chief Data Strategy Officer at Squirrel365 talks about:
- How self-service analytics has evolved
- What is missing today when it comes to serving data to the users
- Tips for boosting user adoption of dashboards
- As well as the future of dashboards.
How Self-Service Analytics Has Evolved
Donald has been trying to find the right interface between data and end-users for more than 20 years, so he has seen how self-service has evolved and he argues that it may not have been for the better.
A lot has changed, but indeed — a lot has stayed the same when it comes to self-service analytics. If you look at Executive Information Systems (EIS) , for example, the output that users see from the system do not look very different from the results they get today, save for the vastness of the data that is now available.
One of the big things that has changed in self-service analytics are the access to data and the tools for analysts and data scientists. The problem, Donald argues, is too many users see this as the only-go to solution, causing problems such as reporting backlogs. A much better approach is to ensure that users can do their own analytics as part of a self-service solution.
What Is Missing Today When it Comes to Serving Data to the Users
The problem with serving up information to the end-users nowadays is that organizations are unable to package up data and information in a way that is as simple as how modern-day applications work.
In the modern-day applications, it is very easy for them to integrate and extend their base data model via relatively simple means. They don’t even require extensive programming; instead, they offer a more declarative configuration.
The end-user, who should be driven by data, has been left behind because of self-service in this regard, because they are unable to use data efficiently for their roles — be it in sales, marketing, administration, or something else.
The main thing that is missing is a clear separation of responsibilities. Wherein it is clear which end-users are the ones who actually perform advanced analytics, and which are the ones who simply need to use the insights of the data that they have access to without the need for acuity and training.
How Organizations Can Boost User-Adoption With Dashboards
“Do not forget that you are in the marketing game, there is always the option for using a dashboard.”, Donald says. What he is alluding to is that you have to entice people to use your Business Intelligence tools. You can use the analogy of a website that offers whatever service. It is not important how impressive the solution the website is selling is — if the interface is not appealing, users will have very little interest in it.
Companies should not underestimate the power of the design of their dashboards. This, according to Donald, is because having a professional touch to how your data is presented is what really matters to people in the beginning. It doesn’t matter what the capabilities are, end-users simply will not flock to a dashboard that is densely packed and boring.
Future of Dashboards
Donald believes that there will be a convergence between the delivery of the data to end-users, as well as specific applications in the workplace. He believes that in the future, dashboards (and business intelligence in general) will simply work the way apps on your phone do. There is data that has a specific purpose, an-end user that is not technical and knows nothing about the actual preparation of that data would only have to click a button and get the insights needed for a certain task.
The same way functionality and advanced algorithms are packed up in applications, that is how information should be packed up as well. Users do not have to know how an algorithm works in an application before they can use it, they do not have to be technical or be trained in machine learning to benefit from machine learning optimization, so why should it be different for data? This is what Donald is alluding to — the future of dashboards would be one that is purpose-oriented and packaged.
