New Relic is expanding its IT monitoring service to offer metrics that could help organizations better understand how effectively they are serving their customers.
"We built a cloud-based service that allows you to ask questions of your software in real time," said New Relic CEO Lew Cirne.
The service, called New Relic Insights, provides metrics on how an organization's Web applications and websites are being used by their customers, using operational data already collected by New Relic's software agents. Currently, the company has over 80,000 customers who generate over 1.1 trillion event metrics each month.
New Relic's speciality has been in the field of application performance management (APM), which involves collecting vast amounts of operational metrics potentially useful in monitoring system performance and debugging issues. Such data could also be useful for understanding others aspects of business as well, Cirne said.
Offered as a service, New Relic Insights, previously code-named Project Rubicon, repackages the data already collected by New Relic agents in ways that can be understand and reused by business managers.
Typically, New Relic customers will install small agents on their software, which transmits usage information to New Relic. The software then aggregates, parses and otherwise packages the data so it can be understood by humans. Metrics might include each time a Web application completes a transaction, or the time it takes for a database to return a response.
The new service offers a SQL-like command set, called the New Relic Query Language (NRQL), for querying data sets in detail. Results can be presented in visualization and dashboards forms, or exported for further analysis in other software programs, such as a spreadsheet.
Such data could be useful for an organization in a number of ways, according to New Relic. A software development team can see, in near real-time, how well a new feature within a software product is being used by customers. The marketing department can track how effectively a campaign is working. Sales teams could check to ensure their top customers are getting premium performance from the organization's service.
Of course, there is no shortage of products available for analyzing data, such as the increasingly popular Apache Hadoop data processing framework.
"Hadoop is overkill for a lot of companies and a lot of use cases," Cirne said. Though Hadoop is a powerful tool, it requires considerable programming and data science expertise, he noted. It also requires considerable IT effort to install and maintain.
"Most of the time, a lot of people are just installing Hadoop or other big data solutions to make sense of their click streams, to better understand their customers," Cirne said. "Well, all that raw clickstream data is coming into New Relic anyway."
New Relic Insights is also slightly different than Web analytic software, such as Adobe Omniture. "The previous generation of Web analytic tools have been focused on websites, and concentrate more on the content of the page rather than the application function," Cirne said.
"A software company like Uber wants to know how many people are booking cabs, not which Web page people are visiting," Cirne said. "Those previous generation Web analytic applications were all about increasing traffic, not understanding what your customers are doing with your software."
New Relic is not alone in expanding its IT operations data for additional uses. Machine data search engine provider Splunk recently augmented its software to provide information for business analysts as well.
Splunk also partnered with business intelligence software provider Tableau software to allow its customers to pair Splunk data with other sources of data in the organization, for additional insights.
New Relic has an advantage over Splunk, as well as similar competitors, in that New Relic offers a pure cloud service, rather than software that has to be installed on premises, Cirne argued.
"New Relic has a massive cluster that all its customers benefit from. So when a customer runs a query, it runs across all of those servers," Cirne said. "If you're a single tenant, it would be incredibly expensive to get to that speed."