Thinklytics

Texas A&M University System · Higher Education · College Station, TX · 24 weeks

11 campus data warehouses consolidated into one

Texas A&M University System managed 11 separate campus data warehouses, each with duplicate infrastructure and ETL processes, making cross-campus reporting impossible. We merged all 11 into a single Snowflake environment, cutting infrastructure costs by $2.3 million annually and delivering same-day reports across campuses.

Challenge

Each of the system’s 11 campuses maintained its own data warehouse, developed separately over 15 years. These warehouses didn’t share metrics or support reporting across campuses. Infrastructure costs overlapped heavily, totaling $2.3 million annually. The chancellor ordered a consolidation to cut costs and simplify technology.

Approach

We spent six weeks mapping 11 warehouses, their data models, and reporting dependencies. Then, we built a single Snowflake environment with a shared semantic layer. We migrated the warehouses in groups of three over 18 weeks. Each migration had a four-week overlap where both old and new systems ran in parallel before shutting down the legacy warehouse.

Outcome

By week 24, we consolidated all 11 campuses into a single system. This cut annual infrastructure costs from $2.3 million to $380,000. We built a reporting platform that delivers cross-campus data on the same day. Within the first month after consolidation, the chancellor's office launched three analytics projects using this new system.

When we dug into the audit, we spotted three warehouses with zero active users.

In just six weeks, we found three warehouses with fewer than five active users each. But get this, they were costing $420K a year in licenses and maintenance. We didn’t mess around. We shut them down immediately to stop the bleeding before moving on to the bigger consolidation stuff.

Tracking cohort progress with a little help from our campus champs

Before we started the migration, we made sure to put a data lead in charge at each campus. The first three campuses dove right in and set the pace. Watching them make headway really lit a fire under the others. By the time the third group came around, some were even pushing to speed things up beyond our original plan.

Diving into cross-campus data to fix real, everyday problems

We put together a dashboard to track enrollment and retention across all 11 campuses. The chancellor’s office had been struggling with data gaps for 15 years, this finally put an end to that. But it wasn’t just about filling in missing pieces. We gave them clear, actionable insights that even the biggest skeptics couldn’t argue with.

Results

  • $2.3M Annual infrastructure savings
  • 11 Warehouses consolidated to one
  • 5 days to 2 hrs Cross-campus report time
  • 24 wks Full system migration

We were running 11 warehouses at $2.3 million a year, and none of them were connected. Thinklytics pulled it all together in 24 weeks. Now, the chancellor’s office can see cross-campus analytics they didn’t have before.

Chief Information Officer, Texas A&M University System

In the client's terms

The question

What breaks when these systems become one?

The situation

Eleven campuses had built their own warehouses independently over fifteen years. No shared metrics, no cross-campus reporting, and heavily overlapping infrastructure at $2.3M a year. The chancellor had ordered consolidation.

What we found

Fifteen years of independent build meant eleven sets of reporting dependencies nobody had catalogued. The risk was not the migration, it was the reports that would quietly stop working on a campus that had not been asked.

What it produced

Eleven became one by week 24. Infrastructure cost fell from $2.3M a year to $380K. Cross-campus reporting went from five days to two hours. The chancellor's office started three analytics projects in the first month.

The roadmap

Six weeks mapping all eleven warehouses, their models and their reporting dependencies before any migration. One Snowflake environment with a shared semantic layer. Migration in groups of three across 18 weeks, each with a four-week parallel run so nothing was switched off until its replacement was proved.

The work behind this result

Related case studies

Frequently asked questions

How do you consolidate data warehouses across multiple campuses?

Map the dependencies before moving anything. Texas A&M University System spent six weeks cataloguing eleven warehouses and their reporting dependencies, then migrated in groups of three with a four-week parallel run each, finishing in 24 weeks.

What does duplicated campus infrastructure cost?

$2.3M a year here across eleven campuses built independently over fifteen years, falling to $380K. Three warehouses with fewer than five active users each, costing $420K between them, were retired before the consolidation rather than migrated into it.

How do you get campuses to cooperate on a consolidation?

Let the early ones set the pace. Each campus named a data lead before migration, and the first three moving successfully did more to convince the remaining eight than the chancellor's mandate did. By the third group campuses were asking to go earlier.