August 16, 2012
How many times have you seen a chart that looks something like this?
If you are an analyst, odds are you spend a great deal of time trying to figure out what is causing significant dips or spikes in key metrics. Drilling into the data might provide clues as to what happened but usually the answer isn’t in the data at all. That is because external events are usually the cause of significant changes in key metrics. Maybe a big spike is due to a new marketing campaign or a dip is due to a payment systems outage.
While the marketing people would know about a new campaign and the Ops team just resolved a recent systems outage, unless you know about these events, the chart doesn’t tell you anything useful. Without proper context, the chart can’t provide any meaningful insight.
In most organizations, countless hours are spent asking and answering “why?” questions around key metrics. Even when an analyst finally uncovers what impacted a metric, most business intelligence tools don’t allow effective capture of that information. As a result, insights are typically shared via. e-mail and are not accessible to anyone who didn’t participate in the original conversation.
The information in any data source is only the first piece of the puzzle. It is the center of a knowledge ecosystem that includes external events and individual insights.
Only by capturing and presenting trends in context we can transform data into insight. In doing so, we can save countless hours of time for the analyst while making the discovery of insights an effortless process for the business user.