Jump to Content
Data Analytics

Wayfair delights suppliers and customers with help from Google Cloud

November 25, 2020
https://storage.googleapis.com/gweb-cloudblog-publish/images/GCP_x_Wayfair.max-2600x2600.jpg
Ethan Dickinson

Associate Director, Analytics Infrastructure, Wayfair

At Wayfair, we use data to advance our business processes and help our suppliers work more efficiently, all with the end goal of delivering great customer experiences. As one of the world’s largest online destinations for the home, our massive scale allows us to use data to delight our customers and help our thousands of suppliers to identify opportunities and bottlenecks. We had previously worked with Google Cloud for our storefront expansion and relied on them to help us scale our web service that was supporting the buyer experience. As we continue to rapidly grow, this partnership will give us more flexibility to handle surges in customer web traffic and unlock more ways to improve the shopping experience. Being able to help scale operations, while providing a richer experience for our customers, employees, and suppliers, gave us the confidence to continue to work with Google Cloud for our analytics needs. 

Improving our customer and supplier experience

With over 18 million products from more than 12,000 suppliers, the process of helping customers find the exact right item for their needs across the vast supplier ecosystem presents exciting challenges, from managing our online catalog and inventory to building a strong logistics network that includes aspects like route optimization and bin packing, while also making it easier to share product data with our suppliers. 

At Wayfair, we work hand-in-hand with our suppliers so that we can help them grow their businesses and create offerings that are a win-win for both the supplier and for customers. Thanks to this partnership mindset, our suppliers benefit from a steady stream of recommendations that are informed by data. For example, we might let a supplier know that there is an opportunity to capitalize on demand within a certain category by making some merchandising adjustments, such as creating more robust product descriptions. We might also work with a supplier to identify ways to incorporate product tags that allow us to surface a more personalized offering for customers whose aesthetic preferences lean toward a certain style. We are in constant dialogue with our supplier partners, sharing insights like “We know there’s a growing demand for this category and you could surface your products better if you made these adjustments to your merchandising decisions,” or working with them on questions such as, “If we have tens of thousands of sofas, how do we offer personalized recommendations to our end buyers?” Obviously, providing this level of analysis at scale requires a platform that is able to process massive amounts of data across multiple systems.

Why we chose Google Cloud 

We chose Google Cloud because we knew they could scale to meet our needs. Google Cloud helped us effectively centralize our data on a platform with low operational overhead, enabling our data analysts and data scientists to run business-critical analytics. With Google Cloud, we were able to move our application datastores, data movement, and analytics and data science tools all into one place, which gave our developers and analysts the ability to store, secure, enrich, and present data that our teams could take action on. 

Google Cloud’s flexibility and embracing of open-source solutions in products like Dataproc and Composer was proof to us that they are investing in a platform without too much proprietary technology, which made it easier for our teams to adopt and use those tools. The team also liked how easy it was to move data in from different sources into Google Cloud. Plus, Google Cloud’s consistent data access model improved data governance for Wayfair. The standardization on Cloud Identity and Access Management (Cloud IAM) controls makes sure that our data is accessible to the right people and always secure.

Google Cloud’s fully managed platform has well-defined services, which made it easy for us to use and adopt products across the portfolio. For example, the Cloud DLP API can be composed together with other Google Cloud tools such as BigQuery and Pub/Sub to build integrated applications for data security, and the BigQuery Storage API and managed metastore offerings enable smooth integration of open source products with Google’s data platform offerings. 

How we modernized our data stack

We needed a way to get our streaming and batch data available quickly for insights. In our previous environment, we maintained data warehouse systems that required multiple copies of data to scale and required complex data synchronization routines. This had resulted in long lead times for our team.

Now, we can ingest event data from Pub/Sub and Dataflow as the data pipeline for real-time insights and centralize our data using Dataproc, Cloud Storage, and BigQuery storage to help overcome data silos, and derive actionable insights. Because BigQuery decouples compute and storage, we’re able to operate with more agility. Unstructured data lives in Dataproc while structured data lives in BigQuery. Our Dataproc instance is used as a single managed cluster with autoscaling for Hive, Presto, and Spark jobs that read data from BigQuery and Cloud Storage-based tables. We visualize our data in Looker to develop curated dashboards to offer a high-level summary with the ability to drill into diagnostics on what’s driving a particular business metric. We also use Data Studio for operational reporting, which is straightforward to spin up on BigQuery.

By analyzing data from our operational SQL stores data as our applications in BigQuery, we were able to improve our inventory and demand forecasting to help our suppliers make better decisions and generate more revenue, faster. Using BigQuery’s flat-rate pricing option, we were able to ensure price predictability for our business.

Video Thumbnail

Enjoying the results of a cloud data platform

At Wayfair, we have always believed in the value of data and recognize the importance of maintaining volume, velocity, and agility as we continue to grow. Google Cloud’s powerful and accessible infrastructure has let our data teams reallocate their time and effort from moving and managing the data to instead innovating on what’s next. 

BigQuery and Dataproc give us high-performance, low-maintenance access to our data at scale. Google Cloud’s analytics product offerings support the full set of requirements of our internal and external users—from descriptive analytics to proscriptive alerting and ML—in a platform that effectively blends Google’s internal technology and open-source standards and technologies. 

In addition to enjoying the scalability and power these tools bring, we also value the performance. In production, we are seeing a greater than 90% reduction in the number of analytic queries that take more than one minute to run. The combination of scale and speed is generating impressive adoption. 

Less than a year into our transition, the migration has had tangible benefits—users on cloud tooling report 30% higher NPS with the platform offerings over existing alternatives with significantly lower investment in support. We get more business done with less effort and more satisfied users with Google Cloud.

We’re looking forward to our continued work with Google Cloud in improving our overall customer and supplier experience.

Want to learn more about Wayfair?

Check out all the exciting things happening at Wayfair engineering on our blog, and if you want to work on these kinds of challenges with a talented, global team, check out our Engineering and ML roles.

Posted in