How Globe is driving data analytics with Snowflake
The largest telco in the Philippines has moved its on-premise data warehouse to Snowflake to address scalability challenges and improve customer experience
In the Philippines, which has a highly competitive telecoms market where consumers switch telcos based on prices alone, operators have had to think out of the box to retain and attract customers through innovative services and a superior customer experience.
For Globe Telecom, the largest telco in the country, the use of data to deliver personalised and customised services for customers is key to staying ahead of rivals.
“We want to have a relationship with customers, and one of our most important initiatives is to build a customer portrait so we can truly understand our customers and know their interests, sentiments and satisfaction,” said Dan Natindim, vice-president at Globe’s enterprise data office.
These initiatives range from product recommendations based on a customer’s usage patterns, to airtime loans for prepaid customers that ran out of airtime, to network planning. “At the end of the day, it’s about being able to handle data well and enable every organisation at Globe to function better,” said Natindim.
Underpinning Globe’s data analytics initiatives is its data warehouse, an on-premise system that it migrated to Snowflake in just 53 days, thanks to its diligence in governing, cleansing and cataloguing data, which organisations often struggle with in their data analytics journeys.
But despite having well-established data management practices, Globe still faced scalability challenges with an on-premise data warehouse, prompting it to move to a cloud-based environment.
“There was more data being produced and we were getting more granular with the use of data, such as using log files to predict certain things,” said Natindim. “And with 5G, there’s just more and more data, so how can we scale and launch more applications and solutions?”
“One of our most important initiatives is to build a customer portrait so we can truly understand our customers and know their interests, sentiments and satisfaction”
Dan Natindim, Globe Telecom
After defining its data analytics roadmap and testing different tools, Globe ran a proof of concept of Snowflake across its organisation. What sealed the telco’s decision to go with Snowflake was the software’s ability to take complexity out of data.
“We have the data, but the challenge is with last-mile analytics that makes data tangible and usable for users without technical skills or having to learn Python,” he said.
“Snowflake makes data very accessible with just SQL skills. It allows users to create their own reports and answer their own questions. We’re able to do things at speed and it is highly performant.
“Because of our big data, we do very heavy report processing that sometimes takes hours, and in some cases the reporting load conflicts with critical workloads. With a cloud solution like Snowflake, we are able to address that.”
Globe has a very active data analytics community, with over 1,000 members from finance, mobile broadband, networks and other business units. Natindim said they have been keen to learn more about Snowflake since a launch event which was attended by over 200 people.
But rather than take a big-bang approach towards its Snowflake deployment, Globe is rolling out the tool to different business units over time. The first group to adopt it was its customer relationship management teams, which can now use the tool to personalise services for customers, said Natindim.
Strong leadership support
The move to Snowflake, which will be completed in early 2023, is just one example of how Globe is living up to its cloud-first mantra, which has received strong support from its leadership team and CEO Ernest Cu.
“One of the questions from Ernest when I presented this was, ‘Why only now?’ or, ‘Why did we start on-premise?’.” And the story behind that was that six to eight years ago, there weren’t really good cloud solutions for big data.
“We are doing very well in terms of engagement, with our net promoter score reaching an all-time high during the pandemic. The trend continues, and I believe the work we’re doing in personalisation and network planning are big factors”
Dan Natindim, Globe Telecom
“At the time, we started with Hortonworks and MapReduce, but now cloud technology providers have caught up, people have really good options to do data analytics. It was easy to get the go-ahead with the migration for the benefits of being on cloud.”
From a cyber security perspective, Natindim said Globe went through the process of ensuring that Snowflake met its information security and data privacy requirements. It also checked out Snowflake’s customer list and was assured that financial institutions and stock exchanges were also using the platform.
Business outcomes related to data analytics, and in Globe’s case a shift to a cloud-based data warehouse, can be hard to quantify as they are often a result of multiple factors beyond technology.
“It’s not just about the data – it’s also about the product, the methods, marketing – a combination of all of those things,” said Natindim.
“But I can broadly share that we are doing very well in terms of engagement, with our net promoter score reaching an all-time high during the pandemic. The trend continues, and I believe the work we’re doing in personalisation and network planning are big factors.”
Snowflake recently expanded to the Philippines from Singapore in a bid to tap more opportunities in Southeast Asia.
Sanjay Deshmukh, senior regional vice-president for Snowflake in ASEAN and India, said the company was attracted to the Philippines’ pro-cloud regulations and the growing adoption of cloud services by enterprises that have been aggressive in digital transformation.
Deshmukh, a former VMware executive, believed the business from Globe would help Snowflake extend its reach into other enterprises. “Globe is a big brand in the Philippines and when they go out and say that Snowflake has helped them accelerate transformation and become more data-driven, that’s going to have a big impact for us,” he said.
Databricks has been expanding its footprint across the APAC region amid growing interest among traditional enterprises and digital native companies in the data lakehouse architecture.
