How data drives decisions at GoCardless: Interview with BI Analyst, Liz
Last editedJan 20202 min read
The GoCardless data team is on a mission to help everyone here at GoCardless make better, faster, more data-driven decisions.
To find out how this works in practice, from tech stack and data sources to projects - we spoke to GoCardless BI Analyst, Liz.
What was it about the data team that attracted you to GoCardless?
It’s a new team that is evolving all the time. You get a real opportunity to be part of a growth journey, not just within the data team, but in shaping processes and culture for the whole company.
My previous positions have been in larger businesses, so I found the opportunity to work for a fast growing start-up really exciting. It’s such a great culture here - we have good fun, but we work hard too.
What interesting projects are you working on at the minute?
Our focus at the moment is building our data foundations - working with stakeholders to define exec level KPIs, to get a deeper understanding of business growth.
With these in place, we can focus on use cases, building out analysis and visualisations in Tableau and enabling teams across the business to self serve.
How does the data team work with other departments?
We are working with other teams to review department-level KPIs and identify where we may need more data sources. This is really exciting as it means there are lots of opportunities to work with new data - and different business challenges.
We just launched data team ‘office hours’ - we use this time to answer other teams’ analytics questions and support ad-hoc data needs.
Over time, we’ll use these sessions to identify trends and common questions, so we can communicate the answers to the wider team.
What types of data are you working with and how large are the datasets?
We work with a wide variety of datasets and there’s a lot of it - last month we processed ~4.1 million payments!
We look at data on payments, revenue and direct debit mandates, defining dimensions and segments within the data.
We’ve recently started working with Salesforce data, looking at the relationships between the sources and using it to map a merchant through the funnel from lead to activation.
Tell us about your tech stack
Google BigQuery to build sql views to deliver data sources that can feed into our BI tool
Airflow to build our ETL processes
Github for great version control; and open source terminal and text editors alongside this to support the build and testing of processes between development, staging and production
PostgreSQL via Postico for some data processes
Tableau for business intelligence
Python is also an area we have potential to upskill and develop into as it can be integrated with Tableau to enable more advanced analytics.
I didn’t have all these tech skills when I first joined, but within 6 months I have learned them, which has been great for my personal development.
As a data team, we continually review our tools with a view to helping the company in making better, faster, and more data driven decisions.
How is your time split between ETL and other technical tasks, vs analysis work?
No working day in the data team at GoCardless is the same, which is brilliant! It’s a real mix between ETL, technical and analysis.
Recently our focus has been on building foundations and domain models - understanding stakeholder requirements, aligning metric definitions, understanding the data we need and how to source it, building SQL queries to join the data together and the end ETL through Airflow.
The more foundations that we build, the more we will be able to focus on the analytical side of things. For example, we’re now looking at how to visualise data in Tableau, understanding the business use cases for the data and how we can enable others to understand and self serve.