Data coming in real-time is becoming easier to collect. Very large amounts of data are created every day from sensors, on-board GPS devices, radars, data streams from social networks. But is not real-time counterproductive?
Real-time data applications
Companies rely on real-time social networking feeds like Twitter, to measure opinions and monitor public opinion on topics of interest to them. This is how disciplines such as the analysis of feeling were born, also called opinion drilling.
Real-time dashboards are widely used in companies that are required to handle a large volume of information, such as in the online sales industry, the financial sector, or large franchises etc. The flow of information that scrolls on the screens of the trading rooms is a good illustration of real-time dashboard.
They can be updated every second, minute or hour. The notion of real-time is relative to the company that uses it, and according to its business needs.
Operational dashboards allow you to view and display in real-time key transaction information.
Why use dashboard with real-time data?
Generally speaking, the primary purpose of a dashboard is to provide information that enables timely decision-making. In the case of social network analysis, for example, what is the number of visits to the website at this time?
However, an activity report over a longer period will have more effect.
When analyzing cyclical data, a slight decrease may be normal. And the analysis of data over a longer time interval may reveal an overall increase, despite some downward movements.
It is therefore important to know how to correctly choose the right frequency for reporting, and for this several elements can be considered:
- Change the calendar: Instead of watching the data in real-time or the last day, watch the week or last month. Perhaps there is a greater seasonal trend that will allow us to contextualize today’s data.
- Moving upstream: Instead of reporting something like daily income, which is the output of each step of your funnel, report on the underlying real factors on which one can actually have a concrete impact.
- Contextualize the data: Instead of showing an absolute metric, show a percentage change or comparison to last week or last month.
- Add alert functions to dashboards: you cannot rely on checking a dashboard in real-time to detect something that is wrong. A dashboard should allow the user to define its criteria and automate alerts by sending an email or notification, to signal the user when it needs to take a deeper look.
So, we should use dashboard in real-time or not?
The real-time dashboards unfortunately not detect gaps or send alerts. They just present the information. A good alternative is the near real-time dashboard (close to real time). The company will need to determine the need to use a dashboard in real-time before choosing this alternative.