Vocus Communications

We delivered our data-analytics and customer-insights platform in three months by using AWS and BryteFlow, reducing deployment time by at least 75%.”

Hamish Morgan – ‎General Manager of Digital & Analytics

BryteFlow on AWS:

  • Increased market share of nbn broadband to 7.7%
  • Delivered analytics and BI platform 75% faster
  • Helped overcome capacity limitations of existing infrastructure
  • Avoided capital costs of on-premises platform
  • Gained a platform that delivers increased business value over time

About Vocus

Australia-based Vocus Communications (Vocus) is a telecommunications company. The company owns an extensive infrastructure network of metro and backhaul fiber, and connects both capital and regional cities across Australia. As of its 2016 annual report, Vocus had a market capitalization of AU$3.5 billion (US$2.6 billion).

The Challenge

The Australian government is investing in the country’s digital future with the rollout of nbn—Australia’s new broadband network. The rollout—scheduled for completion by 2020—will mean internet speeds of up to 100 MB per second across the country. As the network is rolled out, internet service providers (ISPs) can plug into the nbn network and offer high-speed connectivity to residents.

Vocus saw the value of using data analytics to both acquire customers who have newly connected to the nbn, and reduce the rate of churn for existing customers. Vocus wanted to model data on newly connected households and businesses with existing customer information to generate sales leads. At its simplest, the idea was that a newly connected household or business that was an existing Vocus customer would make a strong sales lead for nbn-related services. What is more, Vocus could reduce customer churn by offering existing customers a far superior customer experience on the nbn network versus the ADSL/copper network.

To deliver this analytics platform, Vocus faced several challenges with its on-premises infrastructure. Because Vocus had largely grown through acquisition, the IT infrastructure was an amalgamation of separate, fully utilized platforms. According to Hamish Mogan, general manager of digital and analytics at Vocus Group, the infrastructure lacked the processing and storage capacity for a high-performance analytics solution. The option to scale the company’s IT infrastructure wasn’t available because of restrictive capital costs and the time it would take to deploy and manage the solution. “It would have taken us at least 12 months to develop the platform internally,” says Mogan.

Why BryteFlow on AWS

Vocus looked to the cloud to run its analytics platform and address the challenges it had with its on-premises infrastructure. Mogan had successfully engaged with Amazon Web Services (AWS) in previous jobs and had personal experience working with AWS solutions architects, whom he contacted with his ideas for an analytics platform running in the AWS Cloud. “The AWS team I worked with really helped me demonstrate to key stakeholders at Vocus how secure and highly available our data would be in the AWS Cloud,” says Mogan. “We highlighted that many financial services, telcos, and pure-play digital companies were already working with AWS, and that gave stakeholders a lot of confidence.”

Vocus engaged AWS Partner Network (APN) Advanced Partner Bryte Systems to create the analytics platform using their BryteFlow software. The data is extracted by BryteFlow near real-time with zero load on the sources and comes from multiple systems across Vocus brands as well as the internet. It gets deposited in Amazon Simple Storage Service (Amazon S3), at which point the BryteFlow software processes this Amazon Elastic MapReduce (Amazon EMR) to create replicas as on source – this high-volume processing is all done automatically with the BryteFlow software with zero coding. The processed data is deposited on the Amazon S3 Data Lake and is used by data scientists for their models and machine learning and is also transferred to an Amazon Redshift data warehouse for complex and recurring SQL queries for Business Intelligence and reporting.

Says Mogan, “You have to understand the amount of work it takes to build a fully integrated data-analytics platform like ours. But through the power of BryteFlow and the modularity of AWS, we have already developed several data models that deliver valuable customer insight to the business.”

The Benefits of using BryteFlow

Vocus is successfully modelling data on newly connected households and businesses with existing customer information to generate strong sales leads. Says Mogan, “We have reached a 7.7 percent market share of nbn with the support of our data-analytics platform running on the AWS Cloud, and we expect that number to continue growing.”

By moving to AWS, Vocus overcame the lack of processing power and storage capacity available through its old on-premises infrastructure. “The processing and storage capacity of the AWS Cloud make it a great choice for any data-analytics platform,” says Mogan. “Eventually, we expect the platform to be ingesting about 10 TB of data each day, and the cost of doing that with an on-premises solution would be prohibitive, let alone the time it would take to curate and manage the data with an on-premises solution.” The company also avoided the capital costs that were another barrier to creating an on-premises solution for its analytics platform.

By using the BryteFlow software, Vocus was able to speed up the implementation of its data-analytics platform. “We delivered our data-analytics and customer insights platform in three months by using AWS and BryteFlow – reducing deployment time by at least 75 percent.”

Vocus has created a platform to drive nbn market share and deliver further value to all Vocus brands over time. Says Mogan, “We are just putting our customer churn model into production. Looking ahead, as we build on our AWS platform and add real-time customer experience feeds, for example, we’ll be able to enhance our strategies around customer servicing and acquisitions.”


Start Your Free Trial