Most brands are producing more content than ever before. Even those still operating predominantly in campaign mode are finding social media demands an always-on content pipeline.
For any marketer struggling under the weight of the data generated by CRM and other marketing systems, the last thing you may want to hear is that there is another class of data just waiting to be exploited.
Machine data, such as that created by servers and mobile devices, is proving to be a rich source of intelligence into customer activity, especially when mashed up with existing structured data sources such as customer records.
Interest in unstructured machine data has been driving the growth of US-based data analytics toolmaker, Splunk, which at its last quarterly earnings announcement in February reported revenue growth of 53 per cent over the same period a year earlier.
What makes machine data interesting is the range of applications it can be applied to. According to Splunk’s vice-president of business analytics, Tapan Bhatt, the pizza chain Domino’s is using machine data from its mobile applications to make better targeting decisions for its discount offers.
“They are taking data off the mobile app, as well as data on which coupons have been redeemed for different kinds of pizzas, and correlating that with geo-location data,” Bhatt says. “So essentially they are tweaking coupons and offers to drive higher conversions based on customers’ preferences.
“This is the type of data people have not relied upon in the past for campaign execution and refining their campaigns, but that is what we are seeing more and more. People are looking beyond what’s possible in structured data, to look at machine data to drive new kinds of analytics.”
Another example is a “large coffee company in Seattle” taking data off mobile apps to understand what coffee varieties are being ordered at different times of day by different user profiles, Bhatt explains.
Other marketing clients include Tesco, which is using Splunk to analyse the path taken by clients through its website prior to purchasing, and the online music service Cricket, which uses it to determine what music may be popular that is not in the company’s library.
According to Bhatt, Splunk differs from traditional business intelligence tools is in its ability to mash up unstructured machine data with structured data from business applications, and do so in real time to provide immediate feedback to marketers.
The full range of uses of machine data are incredibly varied. National Australia bank for instance is using it to analyse server data from its online banking channels, such as cookie downloads and changes in IP addresses for individual users, to identify potential fraud.
In Japan an elevator maintenance company is using Splunk to analyse the records from sensors on the lifts that it services to determine usage patterns and create better service contracts for its clients.
“A lot of traditional manufacturing companies are looking to mine the data that whatever they are making is collecting,” Bhatt says. “If you are making medical devices you are collecting so much information, but a lot of that information is thrown away because there is no way to process the data.
“And so now they see what is possible.”