- Prior to using Fivetran, Papier’s CTO spent one full working day a week fixing ETL issues — time he now invests in strategy planning.
- Papier doubles the number of usable sources it regularly pulls data from and builds reports much faster.
- The company builds its own attribution model and improves product roadmap development.
- Before its modern data stack only one employee was using data; now roughly 60% of employees actively use Looker dashboards, reflecting a new data-driven culture.
- Pipeline: Fivetran
- Sources: Amazon S3, Apache Kafka on Heroku, Bing Ads, Facebook Ad Insights, Google Ads, Postgres SQL, Xero
- Warehouse: Amazon Redshift
- BI Tool: Looker
Papier is a design and personalisation business selling stationery, invitations, cards and photo books. With great success in the UK since its founding in 2015, the ecommerce company has started to expand into the US, Australia, France and Germany. Papier also plans on broadening its product portfolio in the coming year, moving from paper goods to other personalised products. While the company appreciates the tradition of sending and receiving cards, its feelings about the tradition of manually extracting, transforming and loading data are a different story.
Seeking a Solution for Jumbled ETL Scripts
Papier uses paid marketing as part of its acquisition strategy, so its ability to analyse spend across Facebook, Google, Bing and other ad providers is critical to its success. Prior to engaging Fivetran, Papier was attempting to centralise its ad, clickstream and transactional data into Redshift using ETL scripts and code written in-house.
This solution was not optimal and the business was spending too much time working on ETL-related issues. While syncing historical data and matching data across campaigns, CTO Joe Robertson found his team would find inaccuracies and inconsistencies that would often require re-syncing.
Looking to save time, Robertson sought out an automated ETL solution. He was impressed with the wide range of connectors that Fivetran offers, including “quite esoteric ones like Bing and Pinterest” that were important for Papier. He started with the Facebook Ad Insights connector and describes the implementation process as “really straightforward.” He was able to replace his previous table in Redshift and start getting better insights that very day.
Since engaging Fivetran, Robertson has doubled the number of usable data sources that he looks at regularly, and he looks forward to adding more. “Fivetran has improved the quality of our ETL,” he says. “It's been fantastic to just let it run.”
Papier is also able to combine key transaction tables straight from its production database into its data warehouse with Fivetran. This gives Robertson the ability to cross-reference transactional data with clickstream and event data to have a single source of truth on how many people are viewing each product, placing orders, etc. His team is able to tie the transactional data together with site behaviour and other data to work out customer lifetime value.
Prior to Fivetran, Papier could only sync its data once a day. “It would take too long to pull reports so we weren't doing it very often,” Robertson says. Now data from Papier's production database syncs every five minutes and Facebook data syncs every hour. He describes the difference as "immeasurable."
Creating Attribution Models
With its data centralised in Redshift, Papier is able to use Looker to derive insights from all of its data sources, and even hired a BI analyst to work strictly with the company’s data in Looker. “We use Looker for all our core data reporting,” Robertson says. “We got up to speed with it very quickly. I’ve connected as many different sources as possible so we can get real insights out of it. It has had a really positive impact on the business.”
One of Papier’s biggest accomplishments has been building its own attribution model using the data from ad providers. “We can link the ad data with our customer records to show ROA and ROI on advertising based on the lifetime value of the customers we acquire,” Robertson says. This has shifted how Papier thinks about acquiring new customers: “It makes such a difference to be able to see if people come back again and place new orders. It’s been incredibly helpful.”
Transforming the Organisation Through Data
Robertson has definitely noticed a shift in the culture since employing his modern data stack: “It’s really nice when I walk around the office and see people with Looker up on their screens, looking at dashboards, most of which they have created themselves.”
People with limited technical knowledge can get up to speed quickly, and they’re able to easily answer questions that would have taken BI teams a long time to answer. “That has changed the culture of the company and how we focus on measurable outcomes,” Robertson says. In the past, he was the only employee actively using the data. Now he estimates that two-thirds of the company actively uses Looker.
Papier’s Customer Service team can now get its Zendesk data directly and analyze how quickly tickets are sold. The Operations team has faster access to information like stock prices, where items are sold, and how things are set up thanks to a Fivetran connector to the company’s replica production database. Previously, the team had been working solely off an admin interface.
“It has had such a transformational effect on how we look at data that it’s almost difficult to imagine what we were doing before, to be honest,” Robertson says. “The company has become so much more data-focused since we’ve begun using these tools.”
Want to see how an automated ETL solution can centralise your data and transform your organisation? Schedule time with a product specialist for a demo of our service or get started today with a free trial.
About Fivetran: Our standardised technology delivers data into your warehouse the right way. Shaped by the real-world needs of data analysts, Fivetran technology supports agile analytics, enabling data-backed decisions across organisations. After a five-minute setup, Fivetran replicates all your applications, databases, events and files into a high-performance data warehouse. Our standardised cloud pipelines are fully managed and zero-maintenance.
About Amazon Redshift: Amazon Redshift is a fast, fully managed, petabyte-scale data warehouse solution that makes it simple and cost-effective to efficiently analyse all of your data using your existing business intelligence tools.
About Looker: Looker is a modern platform for data that offers data analytics and business insights to every department at scale, and easily integrates into applications to deliver data directly into the decision-making process.