Did you ever hear of Google Flu Trends?
The idea was pretty simple ... track changes in web activity over time & make predictions.

For example, suppose Google saw a growth in searches for "flu" during February. They could make a prediction (along with other insights) that a flu outbreak was on its way.

Although the Flu Trends project has now ended, the concept remains powerful. You can do something similar by tracking data on your own website

Change in web activity / Time = Prediction

Last year a busy year for foreign passport inquiries.

After the Brexit vote & the election of Donald Trump, traffic to citizenship and immigration websites in Canada, New Zealand, Australia, Ireland and more soared.

It was an easy bet, therefore, that passport applications would start to increase a short time after. Indeed, in the 2nd half of 2016 applications for citizenship by UK & US citizens in many countries grew dramatically.

This demonstrates how web can act as an early warning system

Changes in business activity (originating from customers) will almost always be detected first via changes in online activity. The insight generated may be used in several ways, for example to:

• Plan resource allocation
• Prepare public communications
• Alert staff to coming work

Of course, the type the predictions you can make is limited by the sophistication of your web management tools. Nevertheless, the idea of tracking online engagement & correlating it to offline activity can serve you well.