“How do I operationalize Kubernetes at scale within my organization” is a question being wrestled with by many of VMware’s customers today – and there are few choices available that really offer you an Enterprise-ready way of delivering Kubernetes, leveraging the investments that you have already made as a new way to consume IT services […]
Ten years ago, the most common form of engagement we had with utilities companies was either the monthly bill or calling to report a fault. Neither promised to be particularly pleasant.
With the advent of smart home apps, and the remote control of thermostats, lights and security cameras, that engagement has been transformed. And there is a good chance it will be pleasant.
Hive, launched in 2012 by UK utilities giant Centrica, says it never expected its app to be taken up so quickly, and to be used so often. It thought users might use the app once a week, at best. Instead, within months Hive found many were opening the app several times a day. Today, Hive has one million users in the UK, and is planning for five million across France, Italy and the U.S. by 2020.
The challenge for Hive, and others like it, is not just to scale to accommodate this growth (it has been cloud-native since day one), but to act positively on user engagement. How can it make sense of these millions of weekly interactions, and then create new services to delight its users?
Hive found the answer in, on the face of it, a wildly different industry: online social gaming. Online social gaming, it turns out, shares many similar traits. The best games need to prepare to go global within days, and they are desperate to understand consumer trends. The better they understand trends and behaviours, the more informed their developers will be.
Hive has relied on Wavefront™ by VMware® for several years to monitor its cloud infrastructure. Wavefront allows Hive to set up alerts, troubleshoot problems with automated anomaly detection, see the real-time impact of production codes, and create meaningful dashboards to monitor the overall system health. Ultimately, it frees up Hive to focus on what the data means.
Hive is throwing up to 100,000 data points per second to Wavefront, a figure it knows will rise (and Wavefront will handle) as user numbers rise to five million. Yet its developers are able to make sense of the insight generated, and never feel overwhelmed by the deluge. Not only can they use this insight to create major new services – and keep Hive ahead of the competition, but they can test and tweak minor changes on a daily basis.
The effect has been transformative. Previously, often the first time Hive would understand there was an issue with a service was when a customer made a call to the Hive call center. Today, more often than not they know when a problem might start before the customer is even impacted.
The really warm glow from Hive’s adoption of Wavefront? As Hive is better able to manage its cloud capacity, it has now cut its monthly cloud bills by 25%. And who doesn’t like to get a smaller monthly bill?
Learn more about scaling SaaS with analytics in this blog post from VMworld Europe 2018.