Stitch Labs: Giving retailers stability and scalability to manage business
About Stitch Labs
Founded in 2011, San Francisco-based Stitch Labs provides a commerce operations platform that centralizes inventory, sales, purchasing, and fulfillment. The software is designed to give retailers greater visibility, efficiency, insight, and control across their business.
Tell us your challenge. We're here to help.
Contact usBy migrating its e-commerce platform to Google Cloud Platform, Stitch Labs improves reliability, scalability, and security and is growing its business by attracting larger retailers.
Google Cloud Results
- Scales automatically as needed with no headcount increase
- Focuses on development versus maintenance
- Troubleshoots in real time, fixing issues quickly and proactively
- Attracts larger customers due to reliability, scalability, and security
Scale from 28 servers to 300+ with no headcount increase
Black Friday and Cyber Monday can be scary times for many retailers. At Stitch Labs, an e-commerce operations platform provider, the approach was to overprovision its infrastructure to help ensure it could handle the surge in traffic on those busy shopping days. “The shock-and-awe approach to scaling was expensive and still didn’t allay engineers’ fears,” says Robert Navarro, Director of Platform Engineering at Stitch Labs. “We’d go into Black Friday with everyone really hoping for the best, not knowing exactly what would happen.”
The rest of the year, Stitch Labs’ rented infrastructure caused other concerns, too. Its infrastructure provider simply supplied the hardware, nothing more. Stitch Labs’ developers built or handled most everything else themselves—logging, database development and administration, web servers, search capabilities, and more. Also, holiday peaks aside, Stitch Labs needed its infrastructure to reliably and cost-effectively scale, as its plans for growth include targeting larger retailers with higher transaction volumes.
In early 2016, Stitch Labs began migrating to Google Cloud Platform to address these and other challenges. “Google Cloud Platform provides us the stability and the scalability we needed,” says Robert. “Google had a lot more services to offer than our previous infrastructure provider, so we no longer have to do so much of the work ourselves. And because it’s cost-effective, we don't have to pass along the costs of the stability and scalability to our customers.”
“It’s clear Google wants to go above and beyond with its infrastructure. Google is taking the extra step to solve customers’ problems, rather than making customers work around those problems.”
—Robert Navarro, Director of Platform Engineering, Stitch LabsMaintenance without downtime
Compared to another cloud-based managed service provider that Stitch Labs considered, Google Cloud Platform offered live migration to its servers, minimizing downtime.
Maintenance on Google Cloud Platform also proved to be painless for Stitch Labs. On the other provider, Stitch Labs would have had to shut down during maintenance. “On Google Cloud Platform, our virtual machine instances are moved around during maintenance without us even noticing,” Robert says. “From the beginning, that showed me how forward-looking Google is. It’s clear Google wants to go above and beyond with its infrastructure. Google is taking the extra step to solve customers’ problems, rather than making customers work around those problems.”
The Google support team proactively reached out to help ensure Stitch Labs’ transition to Google Cloud Platform ran smoothly. Representatives at a competitive managed services cloud provider, for instance, simply told Stitch Labs’ developers where to find documentation to answer their questions. In contrast, Google teams were in frequent contact with Stitch Labs’ developers and even helped architect much of the migration.
The wide array of managed services available via Google Cloud Platform enables Stitch Labs’ developers to focus on developing. “Why do we want to spend our engineering resources on logging or search?” Robert asks. “We’re not a logging or a search company. We’re an inventory management company.”
At the same time, using Google Stackdriver Logging and other technologies, Stitch Labs is able to scale its logging to roughly 2 million logs an hour, or about 1 terabyte of log data every day.
Google Cloud Platform allows Stitch Labs to scale its infrastructure without growing its team. A few years ago, before Google Cloud Platform, Stitch Labs had 28 servers hosting its database. Today, more than 300 servers host its database, with no increase in headcount. Without Google Cloud Platform, Stitch Labs would likely have needed three or four extra DevOps people.
Troubleshooting in real time
With Google Cloud Platform, Stitch Labs can troubleshoot infrastructure issues in real time to reactively and proactively deliver a more reliable service to customers. If a customer experiences problems in order processing, for instance, Stitch Labs is better able to identify the problem’s root cause and fix it quickly. Proactively, Google Cloud Platform helps Stitch Labs detect problems in the system or with a customer’s account before the customer notices.
Stitch Labs provides Application Program Interfaces (APIs) to customers, enabling them to integrate their systems with Stitch Labs’ platform. “If you're going to succeed in this landscape, you need to take responsibility for how your customers use your APIs,” says Owen Wendland, Chief Technology Officer at Stitch Labs. “And if you have systems that can troubleshoot proactively and show you potential issues in real time, you can make sure customers have a successful experience.”
“There’s a lot of talk that you can't host a database in Kubernetes or in containers. Well, we do it, and we do it well. We’re pioneers in running MySQL in a shared infrastructure on Kubernetes.”
—Robert Navarro, Director of Platform EngineeringThe robust Google network security also gives Owen and Robert peace of mind, knowing they can provide their platform to customers without spending time managing security.
By delivering an enhanced customer experience with increased stability, reliability, scalability, and security, Stitch Labs is attracting larger retailers as customers. “It’s given us a lot of confidence to move forward with more and bigger customers,” says Owen.
Stitch Labs’ entire database runs in Google Kubernetes Engine containers. “There’s a lot of talk that you can't host a database in Kubernetes or in containers,” Robert says. “Well, we do it, and we do it well. We’re pioneers in running MySQL in a shared infrastructure on Kubernetes.”
“Google BigQuery is one of the most transformative technologies I’ve used in years. When someone in the company needs a report, I can slice and dice data any way they want it.”
—Owen Wendland, Chief Technology Officer, Stitch LabsA transformational technology
“Google BigQuery is one of the most transformative technologies I’ve used in years,” says Owen, allowing Stitch Labs to use data and analytics in ways that would have been cost prohibitive in the past. “When someone in the company needs a report, I can slice and dice data any way they want it. That opens up a lot of possibilities. And it has us imagining how we can use Google BigQuery for even more advanced reporting and analytics that we can provide our customers.”
Stitch Labs has many different application servers, with an enormous amount of API requests that travel across its network to give customers the data they need. Because Google BigQuery is fully integrated into other Google services that Stitch Labs uses, all the network performance logs are sent to Google BigQuery. As a result, it’s easy to identify and troubleshoot issues quickly. “A single request used to take an engineer almost two days to get all that data so we could troubleshoot,” Owen says. “Now, we can see it in minutes.”
Managing huge traffic spikes
Historically, traffic on the Stitch Labs e-commerce platform doubles on Black Fridays and triples on Cyber Mondays. One way of planning for those spikes is to run load-test experiments in controlled settings. Even so, building an infrastructure robust enough to accommodate big, occasional spikes is cost prohibitive. And the higher costs would make it difficult to provide the Stitch Labs platform to retailers at competitive prices.
After migrating to Google Cloud Platform, Stitch Labs has sailed through the holidays without outages or issues. Robert and Owen point to Google Kubernetes Engine as a major factor in that performance, by supporting the Stitch Labs platform with an infrastructure that autoscales in real time.
At the same time, the ability to provision containers for customers helps Stitch Labs understand its actual costs per customer, which in turn enables the company to more accurately price its services based on infrastructure use.
In the future, Stitch Labs plans to move its entire application stack into Docker containers running on Google Kubernetes Engine. The company hopes to serve as a complete operational hub for its customers, with the ability to track all inventories, orders, shipments, deployments, notifications, and more, as well as provide metrics and analytics to help customers be more efficient and competitive.
“Without a system like Google Cloud Platform, it would be cost prohibitive for us to collect all those different types of data and make sense of it for customers in a meaningful way,” Owen says.
Tell us your challenge. We're here to help.
Contact us
About Stitch Labs
Founded in 2011, San Francisco-based Stitch Labs provides a commerce operations platform that centralizes inventory, sales, purchasing, and fulfillment. The software is designed to give retailers greater visibility, efficiency, insight, and control across their business.