How Welcome Homes established an SSoT while saving 10 hours a week managing data pipelines with Y42
How a one-person data team built out a scalable data infrastructure from scratch that provided a single source of truth with an accelerated time-to-insight and significantly reduced management efforts.
industry
PropTech
location
New York, USA
First end-to-end use case set up in 2 days.
Time spent maintaining the pipeline and transforming data reduced by 10 hours per week after setting up Y42.
Solid infrastructure is paving the way for additional use cases and new data hires
Implementing a data setup without a data team
Welcome Homes is a next generation home builder that offers beautifully designed home collections, customized to your lifestyle and delivered to you, turnkey, at a guaranteed price.
As such, the company is managing multiple sales funnels dealing with buyers, land sellers, and real estate agents which increases complexity for data-driven decision making, especially when combining sales funnel with product data. So, when the company started scaling at the end of 2021, the leadership team decided to build up a professional data infrastructure. The catch? Welcome Homes did not have a data team!
Starting a company-wide data initiative
When Joey Shampain joined the team 6 months prior, Welcome Homes did neither have a Product nor a Data team. As the newly appointed Senior Director of Product, he had to wear both hats and was tasked with building out a robust data infrastructure right after the funding round.
After struggling with repetitive ad hoc analyses, fragmented data sources, and no single source of truth, Joey started a big initiative to professionalize the data setup, as data was a key component for driving effective business outcomes at Welcome Homes.
“Data allows us to understand where our opportunities are, establish KPIs, and test specific hypotheses.”
He implemented a Postgres analytics database, installed Segment to pull data from Hubspot and Google Analytics to store them centrally, and set up Amplitude for quicker product analyses. After all that, he was looking for the missing piece of the puzzle — a data platform that would fit his needs to pull data from the pre-aggregated and additional sources, join them together, and create dashboards visualizing their core KPIs.
Finding the right data solution
In order to find the right data tool, Joey was looking for three attributes:
Ease of use - especially, joining multiple datasets in an easy fashion
Comprehensiveness - breadth of coverage of functional requirements
Price - Ability to grow with the platform
At first, Joey was looking for BI platforms such as PowerBI, Mode, and Metabase. Since those were mainly focused on data visualization, Joey realized that he would need to purchase an additional tool to ingest data from multiple sources and another one for data transformation.
On the other hand, having worked with the ‘Modern Data Stack’ before (the combination of point solutions such as Fivetran, dbt, and Metabase) he knew that it was an option that would require significant efforts in maintaining the stack as well as the addition of engineering resources sooner rather than later.
Looking for an alternative to the established but fragmented data tooling landscape, Joey found Y42 to be the most viable solution for his needs, after a final evaluation against Sisense and Mode.
“I knew the Modern Data Stack, how to write SQL, and understood the concept of having an ingestion layer, a modeling layer, and a visualization layer. The big draw for Y42 was the ease of use in its transformation layer and that it was all in the same place.”
A scalable data architecture
After choosing to go with Y42, the implementation was very quick. Joey set up Google Bigquery as the data warehouse of his choice and used Y42 to connect both the Postgres analytics database — which pre-aggregated data from Hubspot, Google Ads, Facebook Ads and the product backend via Segment — and the MySQL production database.
With this setup, Welcome Homes managed to ingest, transform, and visualize data for a variety of use cases:
Visualize sales funnels of buyers, land owners, and real estate agents.
Combine product-and sales conversion data to inform targeted sales activities.
Track web behavior and conversion attribution.
The result
Overall, Y42 has helped Welcome Homes to realize efficiency gains in terms of time-to-insight, maintenance effort, and operational improvements.
As for time-to-insight, Joey managed to build out his first end-to-end sales funnels within just two days of working with the platform. Joey also estimates that working with the no-code transformation feature and being able to debug and adjust models rapidly through the end-to-end lineage has saved him up to 10 hours per week. Above all, Welcome Homes has managed to establish a truly data-driven culture, powered by Y42.
“Zooming all the way out, Y42 has allowed us to establish a baseline, see how we’re improving, and ultimately transform into a data-driven business. Through that, we have seen efficiency gains across the product-, marketing- and sales funnel. ”
Leave us a comment