Fivetran and Snowflake enable data-driven decision making for Sleeping Duck

“We are an e-commerce company, but we’re not a tech company and we don’t have an army of engineers. Being able to rely on Fivetran and Snowflake to pull in and manage the data we need frees up resources for other, more critical projects that are our core business.”
— Chao Li, data scientist, Sleeping Duck

Competing in a Crowded Marketplace

Sleeping Duck is a pioneer in the crowded online mattress space. Founded in 2014, it delivers high-quality, customizable sleep experiences to customers. The company’s innovative business model and focus on customer service captured the attention of consumers when the Sleeping Duck founders went on Shark Tank Australia. The founders ended up not taking a deal that day, but instead decided to grow organically. Sleeping Duck is now one of Australia’s most recognizable brands and a leader in a highly contested  market. In order to keep its competitive edge, data drives every decision that the company makes. Streamlining the customer journey, optimizing advertising spend, pinpointing new markets for growth and ensuring superior customer service are just several ways Sleeping Duck uses data to drive results.

Data Complexity and Tech Debt

Sleeping Duck found that harnessing the power of data is actually quite hard — even if you build a digital business from the ground up. According to Chao Li, a data scientist with the company, Sleeping Duck’s data lives in disparate sources in Software as a Service (SaaS) platforms, web apps, marketing platforms such as Facebook and Google Ads, and in the company’s own product. Pulling data from these disparate sources in the cloud, sifting through all of it to identify relevant information and deriving actionable insights was a complex, manual process. To create scripts manually to extract data from these sources would have required the company’s engineers to write custom scripts — and then they would have to maintain them in perpetuity. It just wasn’t a scalable or sustainable business practice.

“Our business was growing very rapidly,” Chao explains. “We were a small team with extremely high volumes, and it became clear that we needed to streamline critical processes so we could continue to expand. It was time to rethink how we were using data to drive our business.”

Snowflake + Fivetran

Chao’s first step was to migrate to a reliable data warehouse in the cloud. He chose Snowflake because the platform needed to be SQL-based and be able to scale compute and storage separately and quickly. This was important, according to Chao, because he knew that Sleeping Duck’s data analytics capabilities wouldn’t necessarily scale in predictable ways. He might need to process a small amount of data very quickly, for example, or a large amount of non-mission-critical data overnight. The flexibility to scale compute and storage separately in Snowflake would save the company a lot of money as its data analytics needs continued to evolve.

Chao also needed a data pipeline solution that could pull in terabytes of data from various sources without much coding or configuration. Fortunately, Snowflake has several data integrator options already baked in — one of which, Fivetran, caught his eye.

Chao now uses Fivetran to pull critical business data from dozens of sources in the cloud to the company’s Snowflake data warehouse. From there, the data can be manipulated as needed and sent to a business intelligence platform for analysis. Chao can then present the findings to internal users or build a custom dashboard that allows them to access the analysis themselves. Data can be synced monthly, weekly, daily or even every 15 minutes, providing near real-time business intelligence.

“Fivetran is super easy to use, and that is a critical business need from our perspective,” he said. “We are an e-commerce company, but we’re not a tech company and we don’t have an army of engineers. Being able to rely on Fivetran and Snowflake to pull in and manage the data we need frees up resources for other, more critical projects that are our core business.”

Now, when Chao receives an analytics request from a business unit that requires connecting to a new data source, he is able to enter credentials into Fivetran, and, within minutes, start pulling data into Snowflake. In fact, within the first week of using Fivetran, Chao was able to start pulling data from 24 different connectors — from Google Analytics to Xero, the company’s accounting platform. Today, Sleeping Duck is using more than 70 connectors to power business decisions across the company.

[CTA_MODULE]

A Data Driven Business

Sleeping Duck’s new data pipeline, powered by Fivetran and Snowflake, enables the company to leverage data to make critical business decisions. Stakeholders from across the organization can present a question — What is preventing customers from completing transactions on the website? or How can we better identify individuals who are actively looking for a new mattress? — and Chao and his team can provide the analytics to answer it.

It doesn’t matter if it’s the advertising team wondering which TV spot would give them the most bang for their buck or the mattress delivery team wanting to narrow the delivery time ranges they give customers, data can provide the answer.

“Our data pipeline, powered by Snowflake and Fivetran, touches every department across the entire company,” Chao said. “We are truly a data-driven company that can take advantage of opportunities as they arise, quickly and with confidence.”

Here are three ways Sleeping Duck’s new data pipeline powered by Snowflake and Fivetran are able to transform the way it does business:

Optimize marketing spend: Sleeping Duck has unparalleled visibility into its marketing campaigns across channels. Going beyond counting clicks, the company’s data pipeline gives the company the ability to understand how leads turn into transactions and revenue. Marketers are able to better understand who is buying mattresses, what they are looking for, the pain points they are experiencing throughout their journey and why a target would pick Sleeping Duck over a competitor. This ability to pull in and easily compare disparate data sets allows marketers to see what campaigns work and which campaigns don’t — and make real-time, data-driven decisions about how best to spend future budget.

Streamline the customer journey: Data from SleepingDuck.com provides valuable insight into the purchasing journey, identifying bottlenecks on the site that lead to abandoned carts. Customer experience experts at the company can then use these insights to tweak the buying journey, ensuring that shoppers are ushered through the purchasing process to completion. The company can follow each customer through each purchasing phase — from first touch to mattress delivery — gaining valuable insights and optimizing each process along the way.

Put the customer first: Service reps have a 360-degree view of the customer when someone calls with an issue. They know the customer’s history with the company, if they have any open orders and how Sleeping Duck has resolved issues for them in the past. This eliminates the chance that reps offer the customer a solution that didn’t work in the past. It also allows the rep to reduce time to resolution, better control the customer experience and continue to drive brand loyalty.

New Ways to Use Data

According to Chao, Sleeping Duck is constantly looking at ways to improve service and customer experience. This means trying to improve marketing messages, address customers' concerns and reward loyal customers. The company is about to embark on a project that will require a deeper dive into the company’s customer demographic, in order to design a loyalty program that will lead to referrals and repeat business. This research will be heavily driven by the Fivetran/Snowflake pipeline.