Yoco Simplifies Workflows with Data Warehouse
Cape Town-based Yoco is tackling one of the biggest issues that small and medium-sized business face across Africa: how to accept payment in the increasingly digital economy. They’ve built a mobile card reader and other valuable tools that enable more than 18,000 businesses to affordably process card payments at the point of sale.
In just a few short years since launching in 2015, Yoco has grown to process more than $170 million in credit card payments. With a massive volume of transactional data at their fingertips, the Yoco team wanted to improve its data analysis to understand merchants’ pain points and develop services that help their businesses grow.
“In order to be a highly data-driven company, thinking through both how we use the data we have, and then how we keep it secure and private has been really important. Our belief and practice is that our merchants’ data should always be used to their benefit and with their permission and Sisense for Cloud Data Teams helps us keep on top of this,” says Bradley Wattrus, Co-Founder and CFO of Yoco.
Taking Data From Operational to Critical
When the business was still very small with only a few hundred vendors, an operational database solution without optimization for running analytical queries worked well enough, but query runtime became a major problem as the company kept growing.
“We recorded transactions in a log file, but it wasn’t really in a way that you could use for data science,” says André Cloete, data scientist at Yoco. “My focus has been to create more historical records so we can resurface data that reflects the state of the business at any given time.”
With about 100 employees on 15 domain teams in three different locations, the need arose to move to a system where analytical queries and reports could be generated quickly. There was no more room for error and they couldn’t afford to question whether data had been recently refreshed. They needed a data platform that provided them with a single source of truth, at scale.
As they explored their options, Yoco found issues with other vendors with cross-database joins, forcing them to replicate data on another database just to be able to complete the joins. “We had problems where queries would start timing out and take way too long to run,” says Cloete.
That led them to move their data into Amazon Redshift and seek out a cloud-based vendor that offered ability to intuitively bring together multiple data sources. With the launch of its Unified Data Platform and its warehouse solution integrated closely with Amazon Web Services, Sisense for Cloud Data Teams became a natural fit.
“Moving all of our data to one place just made sense – it simplifies the pipeline,” says Cloete. “Any table locating issues we had before aren’t an issue anymore with everything on the same warehouse in Sisense for Cloud Data Teams.”
The adoption of Sisense for Cloud Data Teams meant bringing together Yoco’s warehouse and visualization platforms, which has dramatically simplified their data pipeline by bringing all data sources and segments together uniquely in one place.
“Our goal in unifying our warehouse and visualization platforms is to make sure that even as the company continues to scale 2x to 3x each year, we can still go back and answer questions and monitor performance reliably from a single source of data,” says James Kramer, Yoco’s VP of Risk & Data. “Sisense for Cloud Data Teams makes it a lot easier by having a complete solution that addresses both aspects of that directly for us.”
Uniting the Team Around Data
As a payment processing service, Yoco has access to loads of transactional data from its merchants that it uses to discover insights that will improve their businesses. Nearly all of Yoco’s 100 employees are regularly checking the Sisense for Cloud Data Teams platform to get broad views of what’s happening in the business or dive deeper into data within various customer segments.
“As a business, we’re very data informed – every team is making business decisions constantly, requesting data and making decisions through logical, data-informed metrics,” says Kramer. “It’s incredibly important for us to show that we have consistent, up-to-date, real-time information that’s easy for our teams to consume.”
The Data Team at Yoco also gets numerous data requests from other teams within Yoco to aid with projects or for company-wide metrics so management knows where the company is heading as a whole.
“Our goal in unifying our warehouse and visualization platforms is to make sure that even as the company continues to scale 2x to 3x each year, we can still go back and answer questions and monitor performance reliably from a single source of data. Sisense for Cloud Data Teams makes it a lot easier by having a complete solution that addresses both aspects of that directly for us.”
James Kramer, VP of Risk & Data, Yoco
“You can look at analytics from some platforms like Facebook and Google on their own, but we find it helpful to have a high-level view of the entire marketing funnel – from spending, to engagement on various platforms, to signups and conversions – it all comes together in Sisense for Cloud Data Teams for us,” says Kramer.
When any type of data is requested, whether for an internal team or an executive meeting, Yoco’s data team will set up a dashboard or chart on the Sisense for Cloud Data Teams platform, enabling their colleagues to log on quickly and download the data they need.
“It definitely saves us a lot of time,” says Cloete. “With the constant flow of new data we have, we’d have to devote an entire engineering team to building a data pipeline if we didn’t have Sisense for Cloud Data Teams.”
A Complete Package for Data Teams
Moving forward, Yoco plans to share even more of the data and insights they collect directly with merchants to explore data themselves in Sisense for Cloud Data Teams. Instead of investing valuable engineering resources to build visualizations for its end-users, they will embed dashboards from Sisense for Cloud Data Teams to visualize insights directly. They are also excited to explore Sisense for Cloud Data Teams’s new Python functionality for machine learning and other complex analyses.
As their data team continues to grow, Sisense for Cloud Data Teams makes it easy for Yoco to bring on new team members to a platform that’s very intuitive, even for those who are fairly new to data science.
“As we’ve picked up the pace with our caching and full-on warehouse solution, our team has also really valued the documentation and blog content Sisense for Cloud Data Teams provides us with nifty ways of complete data analysis,” says Cloete. “These are really helpful for our more junior data scientists to use.”
“It’s so easy to use – without any SQL familiarity, our new team members can go from using Excel to writing queries and produce great visualizations within a week,” says Kramer. “And yet, as simple as it is, Sisense for Cloud Data Teams also has the power to solve our more complex problems with deep technical analysis. It really is a great all-around package.”