Stitch helps Envoy form a central 'source of truth'
Envoy is transforming modern workplaces with products that make office life easier and work more meaningful. Envoy's workplace platform has redefined how offices welcome visitors, keep employees safe, book desks and conference rooms, and manage deliveries in over 14,000 locations around the globe while building products for a new era of workplace experience. Companies like Slack, Pinterest, and Warby Parker rely on Envoy to create an unrivaled first impression and keep their offices safe and compliant.
All of these functions generate a lot of data. Envoy Data Scientist Arvind Ramesh says, "We wanted to centralize our data to form a 'source of truth' which we could use to make business decisions. In addition, combining different data sources allowed us to answer deeper questions and do more advanced analysis. For instance, we were able to create a multitouch attribution model, calculate our fully loaded LTV / CAC ratio, and develop an account health system that identified at risk accounts before they churned."
Envoy reviewed a couple of ETL solution and decided on Stitch thanks in part to its row-based pricing model. Ramesh says, "We had a lot of different data sources that we needed to get into our data warehouse, but many of these sources had a relatively small number of rows. This made Stitch a much better fit than tools that charged based on the number of data sources used."
It proved to be a good choice. "At my previous company, we had one ETL job that would take eight hours to run, and it would break at least once every two weeks. This caused all sorts of problems for the data team; we couldn't build confidence within the company when our reports were sometimes outdated and inaccurate. With Stitch, we have no such problems. Stitch has saved us countless hours across our data team and has allowed us to scale the team to five people without needing a single data engineer. Stitch gives us more time to focus on the things that matter; instead of building and maintaining custom ETL pipelines we can focus on analysis."
Ramesh has also been impressed with the way Stitch gives Envoy "the ability to quickly replicate new data sources to our data warehouse as we add them. For example, we started using a new net promoter score (NPS) tool and instantly had access to that data because of Stitch."
Stitch gives us more time to focus on the things that matter – instead of building and maintaining custom ETL pipelines we can focus on analysis.
As Envoy continued to grow, so did the numbers of tools and services from which it required data. For some of these tools, no ETL integrations were available from any providers, so the data team had to build them in-house. After evaluating options, Envoy decided to build these custom ETL pipelines using the Singer open-source framework. Singer abstracts away much of the heavy lifting when it comes to processing and formatting data from various APIs, and Singer scripts can be run with Stitch. Ramesh says, "Using Singer in conjunction with Stitch's Import API has made building our own ETL incredibly simple and scalable."
Using Singer in conjunction with Stitch's Import API has made building our own ETL incredibly simple and scalable.