Top results
Weeks to <1 day for new-source onboarding: The team reduced the time to ingest a new cloud source and model it to the silver layer from weeks or months to less than a day using Fivetran, Snowflake, and Coalesce Transform.
65+ EnterpriseOne objects relabeled in under an hour: Using Coalesce macros and the Oracle EnterpriseOne data dictionary, the team bulk relabeled cryptic fields across more than 65 objects in under an hour.
Higher reliability for 100M+ row ERP tables: Truncate-and-reload patterns were replaced with incremental processing on tables exceeding 100 million rows, preventing failed runs from clearing tables and breaking downstream reporting.
Standardized medallion pipelines and repeatable development: The team standardized bronze–silver–gold transformation patterns and reduced one-off pipeline styles, enabling a four-person engineering team to deliver consistent SQL models faster.
Shared definitions and lineage through an integrated catalog: Coalesce Catalog replaced a manual SharePoint inventory with connected lineage and a business glossary tied to Snowflake and BI assets, improving discoverability and trust.
Pain points
Toll Brothers ran analytics across siloed databases with limited ownership, lineage, and operational consistency: Teams created databases, spreadsheets, and desktop tools independently, making it difficult to manage analytics at enterprise scale.
Legacy ETL workflows bottlenecked delivery behind DBAs and manual SSIS development: New data requests often required custom packages and stored procedures, extending timelines from days into weeks or months.
High-volume ERP tables relied on truncate-and-reload patterns, increasing failure risk: A single failed run could clear tables exceeding 100 million rows and disrupt downstream reporting.
The lack of standardized transformation patterns and shared definitions increased operational overhead and made analytics difficult to maintain as demand grew.
“The biggest problem the company had was that there were so many data silos. There were databases everywhere. No one knew where things came from, how data got ingested, or the timing of data. It was essentially just a data dumping ground.”
The challenge
David Le, Data & BI team lead, inherited a reporting environment where analysts and business users often selected datasets based on naming alone. Without clear lineage or definition context, different teams reported conflicting metrics from the same systems.
“There wasn’t a lot of trust in the data that we had. Some people would find data just based on the naming of it. They thought it was a good use for their needs, but they didn’t realize what the purpose of that dataset was for.”
Incorrect metrics reached leadership dashboards, creating credibility risk for the data team. Toll Brothers needed a cloud-first architecture that made definitions explicit, lineage visible, and analytics delivery predictable—without requiring months of bespoke development for each new request.
The evaluation
Toll Brothers evaluated transformation and catalog platforms that could accelerate delivery while preserving engineering control and improving data discoverability for business users.
Key criteria included:
Standard SQL with transparency: The platform needed to generate readable SQL while allowing engineers to inspect, debug, and extend logic when required.
Low-code speed without lock-in: The team wanted a visual, low-code experience to reduce development time and onboarding friction without obscuring core logic behind proprietary abstractions.
Medallion modeling support: First-class support for bronze–silver–gold patterns was required, including incremental processing suitable for large ERP tables.
Governance and discoverability: The solution needed connected lineage and shared definitions tied directly to Snowflake objects and BI assets, eliminating guesswork around metrics and intent.
“We started looking at tools like Coalesce, where it was no-code, low-code solutions. And we also liked how Coalesce also allowed you to view the SQL that it generated, as well as the ability to write custom SQL.”
Toll Brothers chose Coalesce Transform and Coalesce Catalog on Snowflake because the approach aligned with how the team wanted to build, document, and scale analytics across the enterprise.
The migration
Toll Brothers took a greenfield approach in Snowflake and defined a medallion model that moved data from bronze ingestion to silver conformed models and gold analytics-ready datasets. The team standardized transformation work in Coalesce Transform, aligned ingestion through connectors such as Fivetran, and implemented environments and testing patterns that matched its release process.
“The standardization Coalesce provides is the key to our scalability. With Coalesce, we aren’t wrestling with maintaining hairy CTEs and stored procedures anymore; we’re working with highly efficient and optimized code. This ensures that our development is predictable, transparent, and easy to maintain, making the tool an extension of our engineering team’s existing expertise.”
Once repeatable node patterns were established, the team shifted high-risk workloads away from full reloads toward incremental processing for large operational sources, including ERP tables exceeding 100M+ rows. It also automated repetitive preparation work by applying consistent labels across more than 65 Oracle EnterpriseOne objects using macros and data-dictionary-driven scripting.
“When I saw Coalesce Catalog, I was like, wow. This is what I had envisioned years ago, and now I feel like this is the right time. Because we have the data in Snowflake. Now people can understand what the source data is, how it gets moved from staging bronze, silver, and gold.”
The impact
With Snowflake, Coalesce Transform, and Coalesce Catalog, Toll Brothers replaced slow, fragile pipelines with a standardized bronze–silver–gold foundation built on readable SQL, incremental processing, and shared definitions, reducing delivery time while improving reliability and clarity across analytics workflows.
Faster engineering cycles with standardized incremental patterns
Toll Brothers shifted from a DBA-centered delivery model to a
repeatable engineering workflow built on standard SQL and consistent transformation patterns. Incremental loading replaced full reloads for large operational sources, significantly reducing pipeline blast radius and operational risk.
“One big thing that we are doing with Coalesce that we didn’t do historically was incremental loading, meaning we’re only processing changed data, new data, or deleted data. But if there is a failure for whatever reason, that data is still intact.”
The team also eliminated manual cleanup work by automating the relabeling of cryptic Oracle EnterpriseOne fields. Using macros and the EnterpriseOne data dictionary, engineers standardized field names across more than 65 objects in under an hour—reducing ambiguity and accelerating downstream modeling.
“Relabeling all of those fields with a click of a button—that was massive. That was huge.”
Higher trust in reporting through shared definitions and lineage
Toll Brothers replaced a manual SharePoint inventory with Coalesce Catalog, directly connected to Snowflake objects and bronze–silver–gold lineage. Business users could trace where data originated, how it was transformed, and how definitions applied across reporting tools.
The catalog connected analytics across BI surfaces (Power BI, Sigma) reducing confusion and restoring confidence in reporting.
A modern data foundation for long-term analytics growth
The modernization program gave Toll Brothers a scalable foundation for expanding analytics as the business grows. With standardized transformation patterns and a shared operating model, the data team shifted from maintaining pipelines to enabling analytics across the organization.
As new use cases emerged across construction operations, marketing, and customer experience, Toll Brothers added sources and delivered new datasets without re-engineering pipelines or disrupting production reporting. The result was a data platform designed to support ongoing growth—without increasing delivery risk or operational overhead.