The British obsession with the weather is well-founded – Celebrus data enrichmentPublished: Monday, 12 December 2016 by Ant Phillips, Senior Developer
Perhaps the most exciting new feature in our latest release, Version 8 Update 16, is what we call Data Enrichment. The concept is really simple; it allows you to call out from Celebrus to retrieve extra information. The information you get is added to the Celebrus Event Stream as one or more new events. Exactly what information you get, when you get it, and where you get it from, are entirely up to you.
Let’s look at an example of data enrichment to better understand this feature. Many retailers understand how influential weather is in purchasing decisions – for example:
“Weather has had a greater effect [on sales] than economic numbers, we’ve known that forever” – Andy Street, John Lewis Managing Director
Now consider what information you have when a visitor arrives on your website and starts a new browsing session. Sessions are automatically contextualized with data such as how they arrived on the site and their browsing agent. For website users we also convert the browser’s IP address into a geo-location. The accuracy of this reverse mapping varies depending on location but in general it is pretty good, especially in urban areas. This physical location is fabulously useful and with this information in hand you can set up a data enrichment which calls out to a weather service and determines the weather in their location (rainfall, temperature, humidity, wind speed, sunrise and sunset).
The final step is for the data enrichment to create a new event which contains the weather information. This event is part of the event stream just as if it were one of the built-in events provided by Celebrus such as a session start, page load or click. This means that all the goodness of the event stream is at your disposal. You can, for example, send the event to a decisioning system and use it to personalize offers to your visitors based on their local weather. Equally, you can store the weather events in a table in your data warehouse and use it to correlate purchasing propensity between products and weather conditions to enhance future personalization, inform online and in-store merchandising, optimize inventory control and more
As a brief aside, we’ve been working with Open Weather Map for a while now and have had excellent results, in fact there are a host of weather services with varying features, pricing and service levels.
This weather example is only the tip of the iceberg, here are some more examples:
Looking through the examples listed above I hope you can see the potential in this feature, creating a rich and valuable event stream which helps you both understand and communicate with your customers better than ever.